As a non-web dev, I have a question about this part:
> There was a sad coda; as is the way of contract work, I moved on. I explained what I had built to my replacement, that it always worked even without javascript. He was appalled and said, “but that’s a lot more work for us.”
Why is it more work? The approach described in the article seems honestly reasonably simple: just write the standard <input> components for the form, have a submit button at the bottom. When I was making my own websites many years ago now, that's how it worked, and it wasn't that hard. Maybe it's reflecting my ignorance in this field, but doing fancy front-ends seems much harder to me.
Starting a few years ago, I realized some junior and medior engineers never once considered the possibility of building a website (app, experience, etc.) in anything other than a heavy SPA framework. But they're not stupid people! If you directly asked "Can you build a website without React?" they know the answer is obviously "Yes." However, if you asked them to build a new website, they would unthinkingly start a new React project, mostly out of familiarity and a desire to get the job done.
A few of them would outright not know how to do anything else. No knowledge of how to stand up a boring HTTP server to send pure HTML. No experience building a form that validates or submits without JavaScript. These are not the people who post here on HN. They are not engaged in online discussions of new tools and skills (or old tools and skills!). These are people who learned just enough from a bootcamp, or their uni's single "web apps" course, to get a job. Since then, they have just-in-time learned whatever their employer required, or whatever particular tools someone else on their team chose for a project.
As an old, it took me a while to recognize/realize it, but I understand them now. Depending on their career path, someone will encounter the simplest aspects of HTML, CSS and vanilla JavaScript after they learn the complex, framework-specific aspects of each. It feels (to them) like more esoteric, advanced, or tertiary knowledge.
Tying it back to to the quote "that’s a lot more work for us", that's not necessarily an intentionally false claim. It probably does feel like a lot more work to perform a task using unfamiliar tools, even if they are less-complex tools.
You are far too empathetic to them. They should not hold the jobs they have.
These are the people writing React monstrosities for government benefit websites, and testing them on fast iPhones and fast 4G, without realizing that every page load for actual users will take 30 seconds on their old $200 Android on 3G, and users won’t complete the form.
It’s a culture of not giving a shit, that’s the deeper issue.
I had a contract once to save a government website that had serious performance issues, it was so unusable that people preferred to go in-person and wait 4h in a queue rather than try to fill the forms online.
The frontend was in React because the company that got the contract initially used React for everything. The frontend was a 5MB SPA, but it could've been (mostly static) HTML files with some interactivity for forms like TFA. Everyone working on the project agreed React didn't make sense, but we couldn't do anything about it because someone from the government IT department would have to admit they made a mistake. There was no budget for rewrites in the contract. The few times a developer attempted to remove any "React monstrosity" they got in trouble.
Sometimes developers care, but the people in charge don't, and in government environments every change must go through them first.
I’m curious if - and when - LLMs change this. They’re very good at web apps. And they’re great at rewriting existing stuff. Just give them a well scoped /goal and go get coffee.
Theres lots of open questions about the future of our profession in the age of AI. But, playing with opus and fable, I think the future will be bright for our users. There is no reason any more for teams to put out junk that’s worse than what an LLM can do.
Unfortunately the LLMs are trained on what we've made, and there's going to be a ton more React garbage[1] in the training set than there are carefully-crafted websites like the article describes, so I don't expect a decrease in overengineered, bloated junk. If anything, I predict that the fact that you can shit one out in less time than before will have a different effect: A modest increase in bloat since an LLM won't mind adding a half dozen redundant and competing ways to do the same things in a large codebase, combined with a shorter mean-time-between-full-rewrites.
I think most of us have seen incredibly creaky codebases that are too buggy to be maintained any longer, where we make the hard choice to wipe the slate clean and build a new one.
We might find those rewrites happening every 12-24 months instead of after a decade.
[1] Frontend people, I mean no disrespect -- just that React & friends are (ab)used for nearly every website now, even those which map perfectly onto the "Simple document viewing with occasional submission of incredibly simple form data" model that plain HTML has always been perfect for.
> Sometimes developers care, but the people in charge don't, and in government environments every change must go through them first.
To be fair, the same thing happens in private companies. How many UI changes have people gone through that didn't actually make anything better and just made everybody relearn everything? We would have been better of scrapping many of those and let people continue to use what's already familiar, but that too would have to involve someone admitting failure, which is a hard thing to do for some people.
I've used many a government website in the Navy, and they were almost invariably bad, but it had nothing to do with React per se.
A very slow website I can think of had something like 200 GET requests required to load the landing page, and it used Liferay with Material Design Bootstrap. That was closer to the "style at the time". React is the style of this time, but you can write very slow websites in anything, I'm convinced.
In Canada you can't call yourself an engineer unless you have some kind of association behind it; the title holds meaning including partially accountability. Something that is lacking in the tech world. I'm not saying I want to live in that world but also I worked hard for the knowledge I have starting in the IE days of web dev; it was hard earned experience making things work across the web without loosing performance. The idea that we have developers out there now getting paid higher than me that are clueless on how auth works, how the browser works, why css and browsers maintain backwards comparability for a reason.. well it's sad; but good for them I guess?
The behaviours of developers as well being beholden to their managers rather than the craft; meaning not saying No we will not move forward without proper unit tests, or pushing back when business demands quick corner cutting solutions.
Anyway, decades of bitterness. I wish we had associations to uphold some level of accountability on developers as much as protect developers. I think things would be a lot more expensive and slow if we did that though.
Fundamentally I agree with your take, not just on dev side but just the web/dev/produce' a culture of not giving a shit.
Junior and midlevel devs aren't decision makers for government benefit websites. The culture of not giving a shit is real, but the responsibility goes far beyond these roles.
If we're talking a government site, chances are you don't have the budget to be able to hire much above junior or midlevel devs. And the project manager probably has a small budget [^1] and little experience with what the web design choices really mean (and what the trade off are).
I think you'd be surprised who ends up making those decisions.
Which goes back to the original point (that's valid for any project) - keep your user in mind. If your users will be using recent-ish iOS or Android devices, use as much flair as you'd like. If your users will be using mass-market low-end devices or used devices from 4+ years ago, then maybe dial down the interface.
Knowing your user is important, no matter what level you're at.
[1] Unless we're talking about some kind of large system that's being redesigned by a consulting company on a cost-plus contract. Who knows how those decisions are made.
Even if this were the case, and I wouldn't be surprised, it's still misplaced blame.
> Knowing your user is important, no matter what level you're at.
I agree, but it's absolutely ridiculous to expect a junior dev to make excellent decisions on this. Software development is a massive industry with no prescribed methods. It's not like these folks are going through a residency before getting the job. Even if they went to uni for CS those programs don't teach these skills.
You'd be surprised, then. Some managers don't know squat. I rolled onto a project once and found that an entire application was being delivered as a 300MB ActiveX control, to run in a browser because that was cool and "cutting-edge" at the time.
Looking at the code, I found it was using UI elements for data storage and other such nonsense. A colleague and I had to tell the manager that the entire thing had to be rewritten. I'm not sure he actually went pale, but that's how I remember it.
It is EXACTLY the type of people that are hired to make decisions, because of either nepotism or impressing with portfolio filled with overcomplicated, 3.js frontpages.
The tech stack is almost always decided by someone in leadership that has no developer experience. Or by the consultant company that will chose the most complicated and difficult to maintain stack because then they can invoice more and will win all future contracts. The trick is to hire someone that is not corrupted by money, someone like the author of this post, who cares more for the users then how much he gets paid.
New cheap android phones are just as slow as old cheap android phones. The bottom of the market has been stuck in performance limbo for years, and modern web dev frameworks are ill designed to meet them where they are at.
I use to have an old pentium 2 computer for testing websites. Sometimes you cant make things fast enough for the old box. A fun trick is/was to have <script>elm.textContent="loading images"</script> between each "heavy" section, all targeting the same elm. If the computer, network or server is truly extremely slow you will get a nice message at the top describing what they are waiting for. On a normal slow computer you won't see the messages unless something went wrong.
I just had one of these people, a contractor working for a state government, argue vocally with me in a meeting stating that "500 JavaScript requests is not a problem" for a single page. Un-cached, of course, despite there being a CDN in front of the site.
You can't win against cargo-cult coders because they just assume you're from a different, competing cult.
They have no concept of engineering or science, they have never encountered it.
Heh this is one nice thing about doing engineering work in Australia. Our round-trip time to US data centers is often about 200ms. There’s no hiding from sloppy choices in the performance panel.
I had an argument a few weeks ago because our page took 4 serial requests before content appeared. I argued - with solid data - that it should be 1. If we could manage that, cold load time would ~ halve.
I see no reason not to be empathetic. The frustration is fair, but it's aimed at the wrong layer. These people were guided into this spot by bootcamps and curricula that start at React and never go down the stack.
My experience was the reverse. I learned HTML and CSS first, then Rails in college to serve templated pages. I understood the client/server boundary fine as a concept, what I couldn't see was where it actually sat in a web context. I sort of knew JavaScript ran in the browser, but then I'd see ERB templates stamping values directly into script tags, so the server was writing the JavaScript that ran on the client, and my mental model fell apart. Where does my code actually execute? Why does this variable exist here but not there? Why does the page have data the network tab never fetched? Nobody ever sat me down and explained the request/response lifecycle as its own thing. I had to assemble it from fragments over years. This was around 2017 for context.
How you learn something shapes how you keep learning. If your mental model is misaligned, everything downstream is friction. The thing that finally made it click for me was reading the actual HTTP RFCs, which is apparently a weird thing to do, because HTTP itself is absent from nearly every guide and curriculum. Tutorials teach you the framework, maybe the language, and just assume the protocol underneath. These days I make newbies read the MDN docs like a book and skim the HTTP wiki page, learn the history of the protocol. It's short! It's not even a book! That gives you a firm foundation. But if your foundation starts at React, drilling down is like digging past bedrock. People don't know where to start, and Googling only shows them wrong answers because they don't yet know how to ask the question.
Are you sympathetic to a doctor who specialized in surgery and now always recommends surgery, even for a common cold? Or would you say they are in the wrong job, if they are anywhere but surgery?
Ridiculous example that does nothing to argue the original, fair point. Obviously health interventions demand more finely tuned solutions than information technology
FWIW, maintaining at least a moderate degree of empathy even in systemically frustrating situations is good for the empathizer and thus in one’s interest
Yup. It’s exactly like the dismal state of CSS frameworks we're mired in.
All these new kids walk in and learn the CSS framework du jour first, then find themselves stranded when things move on. If they had just learned CSS the first time, they'd be set for life.
Nobody should learn React before learning HTML and vanilla javascript.
Kinda sorta analogous to the cloud engineers who can standup complex monstrosities in AWS-land, but don’t know the first thing about how to troubleshoot say a connectivity or simple problem where they have to ssh to an ec2 box and do the needful
well... just because you know how to ssh into the prod DNS host and manually update the prod zone files in vim to remove orphan A records + duplicate CNAME records, in order to fix an ip address exhaustion issue that is blocking new VM's from spinning up for your customers... it doesnt mean that you should, lol.
that was 8 years ago in my first gig. now i kinda wonder... having those skills made it easier to put off implementing a robust long term solution. it was playing with fire, sure, but i was a rookie
I’ve long thought frontend web developers are the ones most threatened by LLM-assisted programming for a bunch of reasons and now I can add “many don’t understand web fundamentals” to the list.
>Don't you think they have a legit skill issue here and should they be better off upskilling themselves?
Absolutely agree! Just because I understand how they got there doesn't mean I think it's a good state of affairs ;)
My post was already quite long, and I didn't want to append a treatise on what one should do when encountering those engineers. It depends on many details. Avoid hiring them, if that's a power you have. If you are stuck working with them, depending on your authority, encourage them to learn or force them to learn. If you're coming in to clean up after them... well, hopefully your comp is worth the annoyance.
We are all simultaneously in the position of encountering "the world as it is", understanding it, and doing what we can to improve it.
Yep. It’s also an attitude problem. A lot of devs are able to up-skill just fine, but some are downright demeaning towards anything they don’t understand, or towards anything that doesn’t come from a FAANG.
In the olden days, people wouldn't take office jobs or factory job necessarily because they thought: "Yes! That's my passion! That's exactly what I've always wanted to do." Passion isn't your first and foremost thought when you have a family to feed.
A few decades ago, IT jobs were for the most part done only by people who were in it for the kick they got out of working with computers. They already hacked at their dad's computer in their early teens (or sometimes even younger), and just could just never let go. It was for people who loved it because it was a niche.
But today, IT is no longer that. It's the backbone of much of our society. And so the field no longer attracts just the die hard fans, the nerds. It attracts ordinary "career people", who just need to have a job to feed the family. Who turn the machine off after 8 hours. Who don't go on coding all through the evening on their hobby project. Who don't try out new tech just for the heck of it.
I think it's hard to understand if you belong to the first group, the nerds, that anyone working in the field isn't like you. Because they all used to be! But those days are gone. We live in the times of enshitification for a reason. If you have the hacker spirit, you don't enshitify because you simply can't. You know what is the right way to do it. Sometimes that's a React app but sometimes it's just an HTML page.
You're not just in it for the money. You care. Not necessarily for the end user, although that would be nice. You care for the tech. And when - like in this article - both come together, sweet things can happen.
I'm not a web developer. I built a few websites in high school, but these days I write safety-critical real-time code for robots.
A few years ago I was back in grad school and I took a class with undergrad senior CS students. We had to write a fairly simple web service, and I was blown away by how complicated they were making it. Based on the requirements we easily could have written 90% of it in plain HTML, but everyone else insisted it should be 100% react. Part of that is honors students wanting to do everything the most complex way possible to impress teacher, and part of it is them simply not knowing that other options exist.
A few of them would outright not know how to do anything else.
It's like how a lot of people these days reach for an electric drill/driver for even the most simple projects like tightening a screw. It never occurs to them to use a screwdriver, or even a butter knife.
There is a reason we use React today - if you want dynamic and complex site, HTML is PITA to work with. Either you reload everything on any change (not good), or end up with a maze of partial reloads in event handlers (not good either). It was done, it was bad, React (and similar frameworks) was the solution. If you don't need complexity of React, fine, use HTML. But if you do, who can guarantee that 5 min later someone will not start requesting site not to reload here and there, but update this bit here. So I don't blame people for just using React. And in this case the issue wasn't that react was used, but that it was used poorly.
As someone who has built both react based frontends and html based ones (with htmx), there is a law of diminishing returns at play.
To start off, writing a basic crud website with forms is much easier with htmx.
But when you start building more complex components, and integrate with other systems (OAuth for e.g.) there are tons of libraries and SDKs for the react ecosystem, but not many for pure html components.
At this point, it's much easier to use off the shelf components than it is to manually write html to handle all the bizarre UI edge cases.
Two years ago, I started a new company, and decided at the outset to avoid using any heavy JavaScript SPA framework. We stuck to simple server-rendered html and only use progressive-enhancement style JavaScript.
Our app was fast, and simple, but it also came at a cost: we were limited in our ability to take rich UI elements off the shelf with an npm package. We had to do a lot more work to provide a rich user experience. Everything took longer, and the user experience was worse as a result. We cared, but sometimes you don't have time to carry through.
The company failed, and I don't think react would have saved it. But I can tell you first hand that righteous adherence to "simplicity" didn't help either. It's always a trade-off.
As someone that reads a lot of code written by others, I'm confident that "learning a new way to do something" is perceived by many as the hardest thing in the world.
It probably has to do with what technology people are used to. There has been a couple of generations of web developers who have only known javascript and its ecosystem for building webapps, and so anything other than pure javascript solution looks foreign.
A lot of developers have made or just perceive very strong silos between frontend and backend today. Any coordination that needs to happen between frontend and backend is potentially a communication challenge.
It seems like a lot more work because you have to keep the backend and frontend in closer sync. The backend has to be aware of and able to store every sub-form in the full process (which sounds like a "wizard form" with a multiple sub-forms to get through the full "form" process), not just accept a "finished" or "complete" submission. If a sub-form needs a change the backend, the backend's storage, and the frontend all need to change. The backend and frontend have to agree on validation logic for each small piece of the form. The backend and the frontend need to both validate every small piece of the form, and maybe can't share that validation logic (depending on what language the backend is written in), especially if one of the goals is to do as much of the frontend validation as possible with Browser native validation tools (`<input required` and `<input type="email"` and so forth) so that you get the most benefits of progressive enhancement.
The original ways of making websites were "full stack" and from a full stack perspective it shouldn't seem that hard to have a coordinated frontend and backend, especially when a progressive enhancement approach likely means a smaller more agile frontend, but current siloed world where frontend and backend are different teams with different goals and alignment makes that seem like way too much work.
That's generous. I always heard people espouse that ideal, but I rarely saw them actually do it. And I never saw it at work.
There were always certain UX requirements that required JS, and that meant the company wasn't interested in testing to make sure it worked without JS. None of their customers were going to use it that way.
Angular, React, etc helped force it further, but they didn't cause it.
I know I always did. CakePHP and Rails made it really easy to determine if a request came from AJAX or direct and you could slightly tweak the response to match the medium.
Agree that most people didn't, but I was always an advocate.
Someone is saying that they delivered a very reasonable solution that's simpler than most would come up. Person taking over was not happy.
Do we know if the code being handed over was high quality? Were they reacting to the fact that it was "not React"? Maybe they have a template they enforce in the company about how apps are built?
As an application becomes more stateful it gets harder to keep that state aligned across the frontend and backend, especially if they're in different languages.
It is harder to do more with less. There is a reason React and other is used so much. Makes it easier to make interactive websites. It’s like asking why backend engineers think it’s harder to code an api using C instead of using Django.
I used to think that was true but I now think it’s only true for very interaction-heavy apps: if you have hundreds of interactions on a page over many minutes, using an SPA is amortized across a lot of time, but if it’s something you could do with e.g. a simple Django app you’ll not only be done faster but will spend an order of magnitude less time on maintenance and accessibility work.
I was confused by that too, but then I thought maybe the new replacement was commenting that using javascript in the app would provide more work for contractors, which he perceives as a good thing. So not using javascript provides less work. I could be wrong though.
As a web dev a lot of this is simply ongoing maintenance of a largely unknown quantity. Most web devs know React and use it extensively; Astro is something they'll have to learn on the job or hire for specifically.
It's akin to writing a backend in Haskell. Chances are you could write something performant that leverages FP in a way that serves as a magic bullet for your domain. But now everyone after you needs to learn Haskell and how to model all future problems in a way that conforms with it - or rewrite things again.
Not a web-dev myself and I was wondering if, apart from unfamiliarity with astro or HTML being treated as unknown technology, it also has to do with having to handle fallback cases, eg the 3 point validation (web component, browser default, server), esp when one is used to have react (libraries) just handling it all without any more considerations.
Before LLMs, learning on the job looked like reading documentation. Now it’s a guided tour with verification. When I produce things in this way, I’m not just blindly accepting it. The goal is that by the end of it I have learned more about the codebase and architecture, not less. I feel that’s important.
Many people don't understand this, even big tech engineers. They see LLMs as a bottleneck. It's more that they don't understand how to use it to multiply their skills, just basics and code gen.
I use multiple Claudes at a time, daily. It's precisely because of that experience that I wrote:
LLM + framework you don't understand goes in ... unmaintainable garbage comes out.
Claude follows code patterns and structure. If you setup that structure and those patterns properly, it will produce great code. If not, it will follow ... whatever it feels like, with each commit.
If you just have it built something with a framework you don't understand, it will do so just fine! But over time every "vibe coded" change you make will drift it further and further, until you are left with a mess of vibe-coded spaghetti.
I agree that's fine in that at least it's doesn't cause unmaintainable garbage. And might even get you up to speed quicker that reading the docs old school.
But the GP point, that you're better off finding people that already, truly understand and are familiar with the tech (ie. Astro), imo still stands.
I read a fun comment the other day from a frustrated windows user who failed to configure linux in previous attempts but now with LLMs it was very easy for him.
By this point people don't appear to have any real clue how to write HTML anymore. Writing semantic HTML isn't significantly harder than say writing Markdown. You copy some HTMl skeleton and you literally just stack your elements into the body. I managed to do that as a 13 year old on MySpace without any deep instruction. Sure you have to close elements as well so the syntax is slightly harder than markdown, but that allows you to differenciate between for example <article>, <section> and <aside>.
I am convinced the one single thing that made HTML unusable over the time was that people wanted or needed a way to re-use parts of the page across multiple pages, like headers, navigational elements and footers.
This meant people used frames, PHP, templating engines or any other new technology mainly for the purpose of creating shared elements, simply because HTML failed (and to this day: fails) to offer a way to include one HTML file in another without having it suck (like frames definitely did, since the browser treated each subpart of the page like its own entity including caching).
The `<template>` tag has gotten very close. Right now you still need JS (optionally wrapped as a Web Component) to load a `<template>` from an outside HTML file (at which case, yeah, it's so easy to just use a JS-based HTML renderer instead of a template today), but discussions are ongoing about closing that loop for simpler "JS-free Web Components".
I don't know when that will be accepted into the web platform, but it still feels more like a matter of time that it may happen eventually.
I've found at least some of my static page generation has moved to just dumbly appending `<template>` tags to the bottom of a page rather than use some other template language, so it feels like we are closer every day to finally having "HTML-native" simple part reuse.
Large websites resorted to PHP and server side includes to get headers and footers. Smaller websites resorted to frames and copy/paste. It wasn't perfect, but it also wasn't horrible or unusable either.
I think it was... SHTML? that allowed for server includes. My recollection from... 25ish years ago was that it was generally quite well supported and worked quite well (and was dead simple to implement). Not sure why, if that was the issue, the fix didn't quite catch on (but it's totally possible I'm mis-remembering the state of browser support).
Server-side includes worked fine but weren't enabled by default in any of the mainstream web servers. I think the lack of default-enabled status hampered their adoption. Joe User couldn't just FTP a bunch of ".shtml" files up to their shared web space and expect it to work right.
I certainly used the heck out of them in the late 90s, though.
It would have been very cool if HTML had been created with the ability to do client-side includes without having to resort to using a Turing-complete VM in the client to do it.
yeah editing all the footers and navigation parts in html is too annoying to me so i've resorted for my websites to just a back button to a page that has links to everything else lmao
I suspect they are both more familiar with client-side rendering, and also thinking of things being able to share components, reuse existing libraries, and so on. So re-implementing everything with vanilla HTML and forms feels like reinventing the wheel to a team used to an SPA component library, it's not that it's intrinsically harder, it's that they don't have the existing building blocks they'd normally reach for.
Modern frameworks such as Astro allow for a similar development experience (and can optionally use JS, React and other client-side libraries) while still being able to generate a static site if desired.
I think server-side rendering and static site generation are less familiar to many web devs who came up in the React/Vue/Svelte era. The patterns and mental model are just different. In an ideal world we combine both: fast static HTML that works everywhere, with progressive enhancement for interactivity. Astro does this well; Next.js supports it too, though the SSR parts have a learning curve.
I think Facebook with their money and Vercel with their VC funding tried hard to push the React and then the Next.js everywhere. So it arrived in time for AIs to all train on it. And now it’s the one true way :)
But do we really need all that stuff? Build steps, bundling, tree shaking, all for what? And is it really simpler… hmm
Simpler doesn't mean easier. Consider a chef who at their previous job started using a wood-burning stove. This is an objectively simpler tool than a gas or electric stove, yet it would be very difficult (even impossible, depending on local architecture and regulations) for a new kitchen to add one.
I'm interested to hear what architecture and regulations prevent the use of something that is foundational to web develpment and backwards compatible by design? Which also, by the way, comes with the advantage of not incinerating other parts of the restaurant (accessibility, user experience...), forcing expensive countermeasures or total rebuilds of the things destroyed every time you turn it on.
90% of the SPAs I use could be Django/Rails/Flask apps with no noticeable difference, other than that they'd be many times more responsive on slower devices.
"Old" SSR apps are mature, not obsolete. It's ridiculous when they're not considered even when they'd be the right tool for the job. I wouldn't try to clone Google Docs in Django, but something like Linear could 100% be modeled as a CRUD app with new page loads on click without a substantial loss of functionality from the user's POV. Not to mention built-in, automatic support for pervasive "deep linking", aka "linking", because you access a URL's contents by GETting and rendering that URL is just how it works.
I agree but you and I aren't the audience and I think experts in general should be a little more holistic when critiquing other people's choices. For any given problem there always exists a perspective from which your solution is over-engineered. People who (like us, I presume) understand processes, files, the command line, compilers, a computer language or two, a bit about computation theory (e.g. Turing, big-O, Knuth..) can get to a very broad swath of places along many different (often shorter!) paths. This is not where most people are starting from.
Speaking of Knuth, imagine being asked to "write a program to add two numbers" and using something like Python instead of assembler, or because really that's complex too, machine code. Do you think that the amount of housekeeping and computer activity is justified for adding two numbers? Objectively, it is not, its just that steady-state dominates the transient over time.
The message you just wrote involved how many complex systems, from your keyboard switches and firmware to your BIOS and OS interrupts, to your browser, the internet and middle boxes, just to say one sentence to someone. It would be much simpler (and more secure!) if you just told me with your mouth, but you didn't do that.
Computers are very good at repeating a known "recipe". They can add numbers billions of time per second. Yes, billions with a bee.
The hard part is coming up with a recipe that solves your problem and that the machine can run without breaking things when it runs around with a few billion steps per second. You have to think ahead for it and handle edge cases in the recipe.
This takes me back close to 15 years ago: using backend session management in Grails and the html forms that were enhanced with “some” JS, using responsive CSS. The difference at that time was browser tech not being as advanced as now, we had to care for different browsers and deal with IE7 and even IE6, it was difficult and we needed extensive QA (Browserstack would appear later). There is a reason why we had JS library evolution. Dude there was no npm, not even bower. Then we had Backbone.js - loved it, then AngularJS - amazing, then Angular version which had huge breaking changes then React, Polymer etc. Native browsers can do a lot these days, it is easy to enhance the functionality as well. But it was not always the same, the decisions to use React made sense for a variety of reasons at the time, maybe it was the case here as well.
I haven't heard much about in a while, but the HTML Triptych proposal [0] is still something I hope to eventually land in browsers. HTML forms speaking to REST endpoints are a good pattern. (meaning user-aiding validation is handled via the input attributes, real validation is handled on the far side of the request, and the flow is GET /form => POST /thing => GET /thing/1) It would be a great pattern with the triptych features implemented!
Most of my apps are now simply HTMX + Go + SQLite.
I've found it's enough for most projects.
One of my sites is image heavy and serves 10 TB of traffic per month. For this, I use the following setup:
1. S3 (I wanted reliable data storage)
2. In front of it, I have Cloudflare (with Tiered Cache enabled, which makes POPs prefer pulling from Cloudflare rather than the origin). I've set rules to cache everything on both the browser and Cloudflare for 1 year, ignore origin cache policies, ignore query strings, etc., and I simply use immutable objects that require revisioning.
3. BunnyCDN in front
Cloudflare will not let you run an image heavy site on its own, so I use this approach to massively cut the bills. Their policy says you cannot use it primarily for images; it must be used for HTML, CSS, JavaScript, and other site content.
And if you run only S3, the bills will be huge.
But yes, lately I’ve been building mobile apps. PWAs are limited; the OS can evict IndexedDB storage, so I cannot offer people reliable data storage in the app without sign up or involving a backend.
What can I do? So I was forced to switch to Flutter on Android, but I ran into another pain point: app updates sometimes spend a lot of time "under review," which is frustrating. For the same app, I maintain a web app that is very quick to update by comparison.
I wonder why there isn't a mobile OS that simply lets you build apps with JavaScript, HTML, and CSS and gives you reliable storage without all this effort.
> I wonder why there isn't a mobile OS that simply lets you build apps with JavaScript, HTML, and CSS and gives you reliable storage without all this effort.
There is! You just have to time travel all the way back to 2009 when webOS was launched by Palm. Time travel is the easy part, you then also need to somehow prevent Palms demise and webOS fading into obscurity as a smartphone OS.
If 2009 is too far back you can try your luck in 2012 with Firefox OS.
Joking aside, people and companies have given it a go. But a combination of bad timing and various other events never made that reality happen in our timeline.
The amount of effort that goes into keeping Termux barely functional, has a lot to do with Android and the platform making it harder and harder, to access a dev environment on a phone.
Running `npm install` on Android isn't so easy.
(Caveat: The new Android Terminal that only works on a handful of models.)
No that's not correct, they both support PWAs with a large feature set. I built and have been maintaining a client's PWA (an internal tool for their employees) for over 10 years now.
Not my site, but someone on HN (can't remember who) built it: https://pwascore.com
Go is so awesome for server apps. I should have discovered it much sooner. It somehow sits in the exact optimal point having no bullshit overhead like C, yet also getting out of your way so you can focus on the business logic like Java (not Rust).
It's not great for every task - in particular the lack of abstraction-building capabilities - but it's great for business-logic-heavy server apps. It feels like it's specialized for that and not trying to be a jack of all trades.
>One of my sites is image heavy and serves 10 TB of traffic per month
I can't imagine this kind of traffic without acting as a CDN, advertising broker, pornographer, or part of a massive ecommerce site. I have to wonder, what are you doing that generates 10TB of traffic per month?
Are you me? I've been on a tear building stuff with HTMX + Go + SQLite. It's like the trifecta of boring technology that jives with me. Stuff gets deployed to a colo server using a generic bash script.
I created a couple libraries to abstract the SQL and HTMX/web/OAuth bits; my apps are now very similar and easy to copy features between.
> I wonder why there isn't a mobile OS that simply lets you build apps with JavaScript, HTML, and CSS and gives you reliable storage without all this effort.
Because there isn't a 30% walled garden you can create with that.
I've learned about several frameworks that are system language + web frontend. The general approach is, ship a small binary compiled for the specific platform, and use the platform's native browser (and whatever html frontend tech you like) for frontend. The whole binary can be 1mb or less.
There's a couple of Rust libraries like this; right now I'm building an app in Tauri+Svelte.
It looks like Wails might be a similar framework in Golang.
Can't vouch for it's effectiveness, but Hotwire Native might meet the needs for mobile paired with plain HTML. Despite the name, it doesn't seem that using Hotwire for the web UI is necessary to use Hotwire Native. I could be wrong about that though.
That's pretty much true of HTMX too (which is what GP mentions using). I mention Hotwire Native (different from plain Hotwire) because it makes it easy to wrap a web app as a mobile app. Then you can replace performance critical parts with native views, but keep everything else working through the web app portion. This is easier to maintain, albeit with the downside that everything requires network round trips. Depending on you use case, that may be an acceptable tradeoff.
> Cloudflare will not let you run an image heavy site on its own, so I use this approach to massively cut the bills. Their policy says you cannot use it primarily for images; it must be used for HTML, CSS, JavaScript, and other site content.
Pages has a 20k-100k limit on static files, but if they just guide you to R2 to offload it, which is still Cloudflare.
Did you mean the CDN? In which case, I'm not seeing that in the terms. [0] Though, I would have expected they'd have a similar thing. R2 resources don't generally count towards your cache limits.
10TB is nothing these days. All Hetzner virtual servers in Europe have 20TB/month traffic included (excess less than $2/TB) and all their dedicated servers have unlimited fair use (which is probably about 200TB/month averaged over many months).
Your Go server can have endpoints that render XML instead of HTML and basically get the same server-driven experience of your HTMX site. Fully skips the need for the app review process since you're not updating the actual client app code to make UI changes.
Since you're using HTMX, I have to ask: do you have any tips or idioms for composing complex forms and UI without things getting out of hand? I love the approach, but I'm having a bad time figuring out where the ideal balance is between too few or too many HTMX-replaced areas in a page. Thanks.
My #1 advice is not creating separate server endpoints for every HTMX fragment, unless you are 1000% sure that endpoint will be used in multiple different pages.
Working on a "simple html page" that is actually 5 different independent "subpages" (routes, views, templates) in the backend is awful. The UX was improved, but the DX was sacrificed.
I recommend having a single view function for each page/SPA and do sub-routing within that function to handle page fragments. In other words, use a GET/path/Header parameter that indicates which fragment is currently needed, defaulting to the full document as normal. Just make sure you are considering request logging and client-side caching in your solution.
This makes it very easy to add/remove async content from the page, since you are just editing the one view function/template and you can easily reason about the entire page as one logical unit.
It also means you don't need to duplicate security logic or other middlewares for the page, since it can be implemented once at the start of your multi-faceted view function.
Generally you don't even need to do the sub-routing in the handler. You can just render the entire page and have `hx-select` attributes pluck out the part that you want.
That is a good solution for reusing content across pages, but most of my HTMX usage is for fetching data that would otherwise delay first page load significantly, or for seamless interactivity. Very different use cases.
This post is good, and it's a great example of taking a problem and solving it with the appropriate tech with the right amount of depth. It really helps to have full domain knowledge of your customers as well.
However, I do not like how it is framed as "simple html is better than react" - because you could just as well have told the same story as a react developer.
(Nb. I could go on forever about the complexities and intricacies of storing things session based on a server vs browser based and etc - and lots of other things that were skimmed over in this article, but that would be too long)
All of those things that are simple in html are also simple in react.
It's literally the same code - there's nothing preventing you from using browser based html validation in react - all the same code that gets complicated in react (overly complicated validation logic) also ends up being complicated in astro - they have their own thing around schema validation etc and integrating it within an astro site means you have to integrate their client router etc etc.. so it's very easy to go overly-complicated there as well.
The comparison is also with an off-shore team doing development for you with probably incomplete knowledge and the way projects are structured they have an incentive to create the solution as fast as possible, in as little time as possible, with the biggest amount of complexity as possible.
The last point is devious - it's not necessarily that the contractor does this by design, but the incentive structure makes it so something that's overly complicated actually benefits them, so they don't have a direct incentive to go with something simple.
Anyways, a simple solution, directly addressing the problem at hand is always better - no matter what stack you pick.
(I'd like to say that I don't have anything against Astro's form validation, I was just trying to highlight how there's more to it than "native html browser validation")
Funny enough. I’m opening this on mobile internet connection and it stuck at loading spinner. I don’t know if the problem is with my internet (probably not) or support for mobile so I can’t even read the content.
Ages ago when writing.com was first modernizing its site, they started by hiding story content to display a spinner, waiting a second or two, then re-displaying the content. It was on the page the whole time, they just made it look like it was loading in the background.
Excellent article but I am always torn when I read inspirational articles like this - it makes perfect sense to me and I love the idea of simple, non-nonsense sites that work well, load quickly and don't rely on the latest browsers to function.
Then I start to wonder if that's just because I'm not smart enough to understand React or whatever the fancy technology of the day is.
Feels like I have a hard understanding threshold that cannot be breached - give me a simple editor like Sublime and ask me to make a web page - even with JavaScript - and it's my happy place. Give me VSCode or Zed, Claude/Copilot/ChatGPT plugins everywhere, React tutorials and my brain goes to mush.
I love the web. I hate what the React cretins have done to it.
Embrace Extend Extinguish is real, and the people going along with it deserve to be replaced by a LLM that lies and spits out garbage code just like they do but faster.
React has helped folks like vercel/convex/cloudflare to build fantastic dashboards. There's just as many examples of well done React as there are the opposite.
I generally prefer solidJS nowadays, but the react ecosystem has enabled lots of amazing user experiences (and developer experiences too if you don't fall into the trap of overcomplexity).
Sadly, this is the story of nearly every React project I've reluctantly inherited. In my experience, it's because React is not opinionated like its peers Vue and Svelte. So, a bunch of devs will use something for state management and and another team will use something else completely. Eventually both teams leave after making a mess. But, if you look at the graphs and numbers that MBAs chase, they will all look like everyone was productive until the very last minute. The ultimate casualty is unfortunately the user. Even Facebook hasn't figured out React across their properties. Just use Instagram / Facebook on the web. Bunch of spinners to load a static list of items in a drop down menu. Not even joking, click on the bhamburger menu on Instagram web. It makes a dozen requests, shows you a loading skeleton and takes 5 seconds before you can see a finite list of menu items. Ironically Facebook was super popular in the 2014s because they didnt have much React based BS going on. Everything was just good old hyperlinks.
If the creators of React haven't figured it out, what makes you think you can?
> Ironically Facebook was super popular in the 2014s because they didnt have much React based BS going on.
facebook was super popular in 2014s because they were just starting the transition from mostly being about people to mostly being algorithmic rage bait, and they tapped a new market of fox news/conservative talk radio listeners
This isn't "We replaced a React app with an HTML form and performance improved." It's "We replaced a bad web page with a good web page and performance improved."
Attributing this to the technology driving the browser experience is silly. You can make a brilliant user experience with React. You can make a terrible website with plain HTML.
The improvement comes from the change design, not tech.
> The improvement comes from the change design, not tech.
You could argue that the constraints of using HTML-first (as they call it) helped them stay away from the bad patterns they were using before.
But you’re right: The user change came from fixing the design, not the technology used.
This is a lot like those bad resume bullet points where someone tries to claim an increase in business was due to their code change. “Increased visitor count 100% by rewriting website to be HTML-first”. Then when you ask them about that point they concede that the entire site was redesigned to fix some design problems or add a feature and that’s what drove the visitor increase.
Plain old HTML with vanilla JS is not exactly a pit of success but React is much closer to a pit of despair. The former tends to be clunky but effective, while the latter requires a PhD in complexity avoidance to stand a chance.
JavaScript: The Good Parts by Douglas Crockford is a comically short book. React: The Good Parts would be shorter still.
The standard answer to that is that some technologies make one harder than the other. That's kind of true from first principles, but it requires making the case that e.g. React is actually harder to make good than a plain HTML page.
Fun thing, TFA describes a kind of multi-page wizard style form that I haven't seen a lot anymore in the last decade or so. But when I did see it, it's always some dogshit enterprise system. Some Oracle product for expensing expenses last time.
The problem with those things always seems to be that they are slow in the middle of doing your task. Every button is seconds of waiting. Doubly annoying if you have to go back a step or two. The badly coded SPAs seem to be slow at the start. It takes a while to load, but once it's loaded its performance is usually okay.
I don’t like multi-page forms when I’m not able to see at a glance all the info needed to fill out the form. Though I guess if my progress is durably saved that makes it not so bad. The worst is when I fill out a bunch of stuff and then realize it’s asking for info I don’t have, and after I get the info I have to fill everything out again from scratch!
Is it slow though? Like in practice? This demo [1] using Datastar (a streaming HTML framework) every action including scrolling roundtrips to the server. Even the checkboxes changing colour is a roundtrip.
The advantage of SPAs, like the checkboxes page, is that they can do the round-trip less visibly. The user can still continue the next thing. So even if it is slow, it's less of a deal than loading and rendering a page anew.
> Is it slow though? Like in practice?
The multi-page wizards? The ones I've seen were. Enterprise crap systems.
I'm personally of the mind that any .gov or utility website should work on old browsers, or at least have a lite fallback to ensure everyone can at least read how to do something the pen and paper way
I built apps like these on GOV.UK over 10 years ago for the Ministry of Justice. We built our own form wizard library that let us validate long forms in steps and break them out into multiple pages because Ruby on Rails didn't support doing that out of the box. It was a very important principle back then that everyone should be able to make use of these digital services regardless of whatever users were using to access them.
I've always liked basic HTML pages where one can upload a document without having to restart the entire application. That's a great practice you have there with general forms. With each session ID, it can cross reference a page in a multi page application with that session ID, so that the user can maybe type it in if necessary, but it should be able to determine that with enough information, like IP address, upload date, browser, OS and so on. But the most accurate session would be within the browser so that the cookies for a single application aren't mixed up with another applicant, like a relative, who might be using the Playstation Portable.
I'm curious how many web issues can be solved by having the people responsible for the relevant sites only be allowed to use them on a Windows 11 machine with only 4 GB of RAM using Firefox with the network throttled to 3G speeds.
Assuming the processor isn't horrible, I can still browse plenty of sites with those specs without much issue, and on the sites that do require more, it's very rarely because the sites actually needs it (i.e. I'm not running Windows XP in a VM in the browser or something). It could just be normal HTML and CSS and normal forms, sprinkled with some light JS to help out a bit. But the amount of sites made with that level of care and attention are sadly rare, since the people responsible rarely feel the pain or have the empathy to fix the problem.
> Windows 11 machine with only 4 GB of RAM using Firefox with the network throttled to 3G speeds
And their mobile device should be the cheapest Android phone from a monthly cell provider you've never heard of. This is what the real world is for a lot of people, and a huge number of developers simply don't know or care.
Don't tell me were going to rediscover progressive enhancement all over again after more than a decade. Back when we used to actually care about the end user whether you were programming frontend or backend.
Too much VC money and big tech influence in the JS ecosystem made the web worse in some ways.
I was going to comment on the Terence Eden excerpt quoted by the author about the woman researching housing benefits on an old PSP browser, when I noticed that you (the OP) are Terence himself. It's strikingly powerful, and a reminder of the duty we have in building our infrastructure.
> Of course, your javascript-based analytics package doesn’t see the users you are bouncing because of javascript failures.
It is frightening to think of how many people are alienated from critical systems every day because of this bias reinforcing the idea that they do not exist.
just use firefox with an adblocker like adnauseam and a fairly decent chunk of the internet stops working, including chase.com and several other massive corp sites.
I can't imagine trying to use links/lynx or a browser with less market share than FF that isn't based on chromium.
It doesn't work if you disable JavaScript...but it wasn't always this way!
They had a mobile version of their online banking service at https://m.chase.com that was EXTREMELY FAST and did 85% of what you need to do in an online banking portal (check balances, transfer funds). They scrapped it when they moved to their current bloated monstrosity of the portal that they have today.
It was a big reason why I moved to a credit union (who outsources their online banking services to Alkami, which maintains a very tight portal and supports 2FA AND passkeys!).
i am unclear on if it's because of the adblocker (specifically i use ad nauseam which does block some JS. some.) or because of firefox. I can load it on edge every time it fails on firefox. last week, chase.com worked fine on firefox. the previous 15 months where i needed to log in, it did not.
Someone at chase isn't checking their work on firefox.
FWIW I use Firefox with uBlock Origin and Enhanced Tracking Protection, and use Chase's website almost weekly. No issues that I've noticed on MacOS or Windows
There is one hard wall that stops very old clients from connecting: Not supporting a new enough version of TLS. TLS 1.2 is from 2008, and TLS 1.3 is from 2018. Web browsers older than 2008 can't connect to modern websites since TLS 1.0 and 1.1 were deprecated from web servers in 2021.
I still think it's worth it to provide connections over plain HTTP for this reason. It probably doesn't apply to many people, and you shouldn't allow anything really important to happen over plain HTTP (logins, payment), but normal viewing should still be possible.
Sadly, the internet as a whole seems to disagree. Even the most useful resource on the web one could use over plain HTTP, Wikipedia, only allows connections over HTTPS. I guess it kind of made sense as part of the campaign to push the internet as a whole over to HTTPS, but anyone who's connecting to any website over normal HTTP these days is doing so because they literally can't use HTTPS.
I agree; you should allow non-TLS connections as well as TLS. (At least, access that does not require authentication should not require TLS, but should still allow it if that is what the client wants.)
If you are concerned about accidental login or API keys without TLS, then you can consider supporting mutual TLS, which improves security (and flexibility) in other ways as well. (You do not necessarily, have to require mutual TLS, in case someone prefers to use a username/password login, or 2FA or something else like that instead.) (In the case of login forms, you can have the links to the login forms to always use HTTPS, in order to avoid the problem.)
I'm not convinced from the article that HTML-first was the thing that fixed the problem. What fixed the problem was 1) the person building it knew what they were doing and 2) it had design constraints from the get-go to be user-friendly. You can do that with React. It's arguable whether it's easier or better, but you can get there regardless of the approach you use.
1. Presumably the team that made the previous app also thought they knew what they were doing. Presumably they were not hired on the assumption that they couldn't make a good app.
2. The design constraints had always existed, the previous app just failed to meet them.
1a. They may have thought they knew what they were doing, but their work product shows otherwise.
1b. They may have thought they knew what they were doing and spoke confidently enough to convince whoever was doing the acquisition, likely non-technical, of the same, but the bad hire and the bad hire’s work product shows that neither was the case.
2. Ideas merely exist. To be constraints, they must be enforced.
JavaScript-heavy approaches are not compatible with long-term performance goals
178 points by luu 3 months ago | 237 comments
https://news.ycombinator.com/item?id=47029339 (should be called "Why React is not compatible with long-term performance goals")
I disagree. An HTML website which uses links, forms, buttons and inputs will by default:
* Have working back/forward buttons
* Have working progress indicator as provided by the browser
* Show errors to the user - even if they are ugly
* Be accessible to keyboard navigation
With SPAs these are all things the developer has to get right.
So often when using a SPA I'll click a button, you get a spinner and then nothing will happen. Is it still in progress? Don't know. Eventually I'll open developer console and trace the network requests to find the JSON HTTP request that returned "ERR_BAD_EMAIL" and fix what I've entered. With a normal form submission at least the user will see the error message and can press back and then fix it.
True. Crappy developers will build crappy websites irrespective of the tech.
The article is clearly aimed at non crappy developers or developers who want to do better for their users.
And it provides an anecdotal experience where an HTML first
option developed by a good developer was far superior to what a JS necessary option would have been, given the user base of this application.
It's definitely possible to make slow server-rendered website. Most of the slow client-side apps are slow because they're waiting on slow network requests.
(I still very much support fast, simple HTML websites. The good ones are a fantastic user experience)
but the host (the company) will need to pay the price in the form of server equipment. Not the user as is the case with client side rendering. If server side rendering becomes slow it will affect all users regardless of their hardware or connection, prompting earlier response from management and devteams.
I agree. Server-rendered React can also send down 100% HTML apps.
But I 100% see where the author's coming from, considering the massive fragmentation of react codebases/patterns and decision paralysis of React development in general. I really doubt most React apps, even the more accessible ones, are testing their multi-page form wizards with JS completely turned off.
HTML-first does seem highly underutilized in the commercial web, and I learnt a lot from reading this (as a solidJS/react dev).
This is patently untrue, give a craftsman terrible tools, and they'll still produce a decent end result. That said, defaults matter, and astro is going to be significantly more friendly out-of-the-box to low-end clients
> Just because something is a proverb, doesn't mean it's automatically true for all cases.
A professional woodcarver armed with only a metal spoon will still make a nicer woodcarving than I can given a full wood shop. Similarly, if you only give me notepad.exe, I can still make a pretty nice website.
Does using the right tools make our lives easier? For sure. Using mildly-wrong tools (react, in this case) isn't going to hold us back very much though.
How good are you at hammering nails with a screwdriver? Tools are important. We have a more apropos saying in our industry: use the right tool for the job. Don't let your brain get fried by the 'poor workman' adage, it's talking about the difference between an expensive tool and a cheap tool of the same kind, not two tools of different kinds entirely.
I find it's way easier to build crappy React apps than an HTML-first approach.
An "old school" Ruby on Rails/Symfony/Django app, with templates, usual get/post forms etc, frames you and pushes you in using the standards and relying on browser default behaviors.
In JS-heavy apps, it's as easy to code normal `button` elements as it is to code clickable `div` elements. But with the divs you just forget to handle keyboard nav, proper element roles, etc. It's easy to create fake links, not relying on `a` tags, using an internal JS router that doesn't expose URLs, doesn't handle middle click mouse, for no particular reasons.
In less JS-heavy contexts, the easiest way to do is to use proper HTML so you are less inclined to mess up.
Even on codebases that use a decent framework like Next.js that handles those for you on paper, it's often we see people not very aware of the benefits of using proper semantics and standard behaviors, and you easily end up with web apps with poor UX in the end.
Not really, no. Astro requires you to opt a component in to client-side rendering, React (with its server components etc) require you to opt out. Defaults matter in scenarios like this and I'd bet the average developer of crappy websites would have a much faster site with Astro than React for that reason alone.
Small correction: it's other frameworks that require you to opt-out, most notably Next.js. Although I've been seeing so many people confusing Next.js with React lately...
Astro itself works just fine with React, and it can still be HTML-only.
But you can also render React on the server yourself using renderToString, if you don't want a framework.
We have the "social media" companies to thank for this innovation. They introduced these stupid patterns primarily to keep users locked into their walled gardens and the rest of the industry followed them to the bad place like lemmings. Software is a fashion industry.
You ever tried to maintain or enhance an ASP form that was built 20 years ago and enhanced by random devs with no documentation? I was working with ASP forms 15 years ago that were hopelessly outdated, included tons of technical debt and were not in the least bit user friendly. You can accomplish so much more with things that ship in every browser now. It’s not a very modern technology and you miss out on things that are now very standardized and much, much easier and quicker to develop and maintain.
That this article even has to be written proves how stupid web development as a whole has become.
I loathe this area of software development, loads of unnecessary frameworks and spinning gears that add complexity while making less and less functional applications
I am not familiar with this astro framework they used. But having built some sites using Pure HTML/JS back in the day, React, Angular, Vue, Rails ERB, Rails Hotwire, and HTMX. I think HTML first websites are absolutely the way to go. Rails Hotwire with View Components makes rails sites super fast, faster to develop and easy to re-use components. HTMX more generally, but Ive used it with Spring boot and Thymeleaf. I really don't want to go back to SPAs. Development time is less and the website performance is better, and I haven't really seen any regressions in capability. With HTMX and some url parameters, I can make a pure HTML site that seems like a Single Page Application but without the excessive loading times.
I first tested Astro on my site and never went back. Now every new project defaults to Astro and I have to have a reason not to use it. So far no reasons. It's simple, fast and it kinda fits my desire to keep things minimal. For example, yes, page content matters, but all but one page on my site is under 10kb, most hovering in the 3-4kb range (100% of the downloadable content)
The full context of that quote makes it clear that it's meant more as a wry joke:
> A venerable web application pattern that has had a small modern renaissance thanks to Remix, form submissions and redirects took a while to explain to my colleagues, on account of everyone being used to heavily client-side web applications.
(Although it's not really a joke, it's pretty amazing how many professional web developers these days don't know how to use forms without JavaScript.)
The opposite is why I'd never be a good web developer. I grew up messing around with PHP and if I spent the time to learn the modern stack, I'd constantly be thinking it's stupid.
I think the author is suggesting that Remix was the inspiration for the renaissance, not that it's necessarily the most popular method for doing so.
I'd be curious to see the stats on how often Next.js users lean into the server component model that makes the frontend fast. My anecdotal experience is that it's an afterthought for many. By comparison, Astro (as mentioned by the author) makes you think about this stuff upfront via opt-in rather than opt-out. It's a wonderful framework.
HTML first, minimal JS, not using any frameworks unless absolutely justified is already part of my prompts to LLMs when I build stuff for myself. I’m not a frontend developer, and have not kept up to date with the ton of frameworks. That HTMl first approach already makes the LLM outputs faster/fresher for me.
I wish more people went this route for public facing services.
> this was a regulated monopoly, and if their customer satisfaction dropped below 96% (if I remember correctly) it could result in millions of pounds in fines.
OK, I'm still at the beginning and irrelevant to the article, but as a USA-ian, I am so jealous about that. Unheard of here.
That hit me, too, specifically thinking about my current gas/electricity provider. I have not heard one single piece of positive feedback from the public, and there's only ever problems. I feel like that's a pretty universal experience here. Even outside the scope of websites, it holds so very true.
Personal anecdote: Recently they were updating everyone to "smart meters" on the gas lines. They needed me to be home so they could enter my apartment and bleed the gas out of the line by turning on the stove prior to replacing the meter. I played phone tag with them for 6 months, setting up countless appointments, and nobody ever showed up, the meter remains un-upgraded. At the same time, I have received weekly phone calls and monthly physical letters stating that if I don't upgrade the meter, my gas will be shut off.
I just moved, so the new tenant will have to deal with it now.
The public sector, simple, no frills, accessible, no flashy graphics, websites were a massive eye-opener.
They just worked. They had a job. They did it. I wasn't going to buy more from them because of it, and they didn't care. It was great.
I've heard that recently they've dismantled the centralised team that wrote all the rules, enforced it, and started moving to decentralised hosting, but so far the whole still seems to hold to together really well. I think, I hope, they have embedded the expectation that the local council, the tax office, your visa status, etc, should just be utilitarian in nature, and work for everyone.
Hmm I cannot load the font on Firefox 151.0.3 Arch Linux. All I see is only a title and empty paragraphs. So I end up reading the article in the source code mode and felt pretty on-brand.
Back to HN comments it looks like this wasn't actually intentional?
This is such a great story. I am glad more people are sharing stories like this. I hope my article the other day inspired more to develop lightweight websites:
My biggest tip to reduce complexity of data validation if you are using React is to stop using React controlled components and switch to React uncontrolled components. They are an underused part of React. You usually don’t need React to handle every single keypress and every single character being entered by the user. In fact before React popularized it, it was unusual for form components to update on each key press; traditional desktop apps tend to validate when a field loses focus only, not on each key press. This has at least three benefits: (a) good for performance, (b) reduces unnecessary error UI when the user is in the middle of entering data, and (c) simplifies your own code by not having to deal with prefixes of valid input that’s not itself valid.
Also allows user scripts to interact with the forms, eg I can run a bookmarklet to fill out certain forms. With React controlled components all these changes are wiped out and reset with the state that React has in its app memory.
> When we launched, the number of people completing the form doubled. The analytics people didn’t even know where these users were coming from. Of course, your javascript-based analytics package doesn’t see the users you are bouncing because of javascript failures.
Yeah, reminds me of the b52 story re holes in wings on the planes that made it back from missions, leading them down the wrong path of strengthening wings. They weren't looking at the planes that never came back with holes in the fuel silages.
How many people are using browsers which don't support Javascript in 2026, and doing so out of necessity rather than out of choice? I can't imagine this number to be >1%.
How many such devices can still support modern TLS certificates anyway? By this logic, shouldn't we also use plain HTTP instead of TLS?
It may only be 1%, but that small fraction of users are also probably the people who sure as hell don't need even one more tiny thing going wrong in their life.
If you're using a decade old phone to sign up for a utility, you've got bigger problems in your life and no self-respecting person should be adding to them.
> By this logic, shouldn't we also use plain HTTP instead of TLS?
Better would be to use plain HTTP in addition to TLS, rather than instead of TLS. TLS does have benefits, but if it is optional then it can also be used on computers without TLS (as well as potentially other situations where you do not want TLS or where it is not useful).
I like JavaScript-light websites just as much as anyone (my own website works on the same principle). However, I do wonder how much of the increased traffic has to do with AI agents that now have an easier time working with the more standard web forms. My own contact form had a bunch of bots quoting Scorpion lyrics before I added a rate limiter.
> Adding a lot of pressure, this was a regulated monopoly
> Some requirements I derived:
> ...
> We had to meet WCAG accessibility (the team settled on AA rather than AAA)
The author doesn't doxx their employer by giving any dates, but if we take the story at face value and assume it took place in last few years, it is pretty shocking.
How does it take a single hero to be fighting for AA compliance as an afterthought for a project with this scope in the 2020s?
I've worked on much more niche projects that treated this with the respect it deserves as a quasi-legal requirement.
As a public policy specialist (and a tinkerer when it comes to tech), I read the article with pure joy.
This is the "wonder" that happens when you build for the client instead of for yourself. It takes being truly intentional: sometimes the client (especially that kind of client) does not know what he needs or want; sometimes you as a dev simply build something with good intentions but just out of habit.
Taking the time to ask true, relevant questions, and build backwards, is one of the ways you achieve this. It takes time, energy, intentionality (once again)... but it's worth it.
Everyone pays lip service to 'user experience' and 'putting the customer first', but, only with experiences such as acing customer service forms can one really claim to know what 'putting the customer first' means.
Sometimes it is just a form that you need, with the web page loading a new one after 'submit' has been clicked. Yes we can do fancy things to ajax-submit the data, but did the customer want that?
Getting to the form in its 'final form' takes work, particularly if asking the customer for more information, such as proof of purchase or important documents. Do you just start the conversation with a 'contact us' box to have customer service ask for the bits they need later, or do you ask the customer for all information up front, reducing the need for back and fore?
You need to actively test what your customers will do, with metrics such as time filling in the form and how big the ticket queue is.
There can also be internal problems that prevent getting the form right. For example, if there is some manager in charge of customer service that is an empire builder, with a vast team. If your form decimates the team because everyone is efficient and able to go home at the end of the day with an empty queue, then you have undermined the empire builder, so he will want you gone!
There are some huge benefits to getting customer service right. You can brazenly have phone numbers, email addresses and even physical addresses, all published on the website. If the forms work then the phone never rings, the inbox is empty. And nobody can accuse the company of hiding behind a wall of corporate website small print.
The word of mouth aspect is also not to be overlooked. You can harvest reviews from happy customers that should have been unhappy under the old system. If you fix their problem in an hour, or get a replacement product to them the next day, then they will write you a rave review, with that being great for the customer because they explain better than you can how dedicated your customer service team is.
I use the word customer lightly here, there is the term 'service user' that is used in the public/third sectors, but that doesn't sound good in front of the 'service user', probably because they have an actual name.
Getting to the form 'in the final form' means quite a few small changes that can be easily reverted and monitored. It could be just making an input box only show capital letters, or show a numeric rather than standard keyboard.
Ideally, a submitted form does something when completed to place the ball in either the customer service court or that of the customer. If the customer need to provide some information before anyone need look at the ticket, the form needs to send out that email, then park the ticket awaiting whatever the customer does next. There should be no need for someone on the team to do that step.
I know AI does everything awesomely under all circumstances, but the 3-6 month journey needed to deeply understand the customer, the product and the team is something that needs a human, simply because you are dealing with humans and their emotions.
What has proven to be a huge bonus is CSS grid styling. Inputs and labels can be written without the div and span cruft, with everything lining up nicely with a few align 'center' CSS things.
What a fun time to be doing forms that actually work!
Not even that old. 60 year people can't user your fancy site because then don't have an internal model of how a computer works.
You know that when pressing a button a hidden engine runs in the backend (or something runs in the backend). You expect an answer and if the expectation do not match the result, the model in your mind creates an hypothesis about what maybe happened and iterate from there. Maybe you should have clicked something before? Maybe you should mark some form checkbox?
Old people don't have that because they didn't grow up with computers.
What is on the screen is what they see. I clicked next and nothing happens. Well... the site is broken.
You known when you plug your refrigerator and nothing happens and instead of reflecting on the possible blown out resistor that you can bypass with a small wire you understand that your only relationship with the refrigerator is plug and unplug or call for help? That is an old person using your site. They won't fight against it. They'll give up immediately.
I would argue it's not even old people. Most people do not have any understanding of what's going on when you click a button. Website either acts as expected, or it doesn't "work"
If the button doesn't work, the average user is going to say "this most be broken" and then use a competitor (or contact your support). That's why it's really important to error-proof one's design (eg https://en.wikipedia.org/wiki/Poka-yoke).
So instead of the button failing because you didn't check a box, pop up with a message telling them "Please click $box before continuing". Or if you want to be fancy, feed them whatever form you're giving them piecemeal, so that they can't continue until they finish this small part (e.g., have them input a name, then the next page only has a spot for an address, then the next page only has a spot for card information, then the next only has a spot to select shipping). Simple bite sized chunks anyone (well, anyone you would ethically want to sell to) can understand.
Yep, I changed town a few years ago and my new social circles are mostly like that.
Either it works right away without any further questions, or they'll not do it.
Sadly, also if they can't do it on their phone, they will not do it. It's actually very hard to get people motivated to do anything that has to do with sitting down at an actual computer anymore. Which is a bit hard if you're in a very technical political advocacy group, kinda makes me the guy to do everything remotely complex... XD
Yeah, it's really, really not an age issue. If there's an age distinction, it's the range of people who were brought up on computers before the UI was really polished, but even then it's not consistent, and they may not have deep understanding. Kids brought up on iPads, and who aren't forced to learn by their interests or educators, have no clue.
When did the industry put the onus on the user to understand how the computer works? What happened to the old days of Xerox PARC's HCI studies putting the user first? The computer is in service of the user, not the other way around!
If I need to build a mental turing machine to understand your application, it is a bad application. It is rather the engineer's job to build a mental model of the user and their needs, and if you can't do that you should not call yourself a software engineer.
As a firmware engineer, my philosophy is this: if I'm doing my job properly, the user should never even know I exist.
Maybe this isn't applicable to all software devs. If you make web apps, users actually see your UI, they click an icon or type in a URL and hit enter with the intent of using the thing you made. With firmware, that's not how it works.
When you hit the "mute" button on your laptop keyboard, it should just do it. The audio should turn off and the little LED should light up. If that fails, even once, the mirage is broken. The user is forced to think about the fallibility of firmware, which is a word they might not even know, and still struggle to conceptualize if they do. I think it also has a lasting effect on the way someone thinks about the pruduct: Is this going to work today? Why did that happen? Was that a virus? So on.
OTA firmware updates have the same problem. Most users don't know what the hell firmware is. All they know is their computer is showing a loading screen they've never seen before. It's unfamiliar and weird.
Like I said, I don't know if this mindset translates perfectly to other fields, but the priorities that fall out of my philosophy certainly apply. Reliability over everything, and get it right the first time.
> When did the industry put the onus on the user to understand how the computer works?
Just after the turn of the century, that was when there was a big shift in the industry to move all user interface design work to artists, and away from original user interface specialists. The whole profession of user interface designer (one who understood "form follows function") ceased to exist in the early 2000s, as all work was assigned to graphic designers (who only understand aesthetics).
This movement can be seen clearly in the evolution of widget toolkits for desktop applications, which peaked in usability in the late-90s/early-2000s, and have been losing usability while getting prettier with every iteration since.
Automation generally goes along with a transition to more "self-service" approaches that require the user to model internal states and workflows of whatever they're dealing with.
This is even true for things as seemingly non-technological as getting to your flight once you arrive at the airport. People who are used to dealing with a service desk might just show up with their printed ticket without even having looked at it, take it to the counter, and expect instructions on what to do next without having read or considered all the fields present on the ticket.
It's not just about understanding the technology, but sometimes about understanding the business, policies, whatever. When a human agent or customer service worker is handling that stuff for you (typical in the pre-computer age), you barely have to think about that stuff and even if you're told, it can be "in one ear, out the other". Automation very often means pushing a requirement of more understanding onto customers/users.
Usability was thrown to the wolves in favor of more readily available designers from non-UI backgrounds, brand identity (“UI as branding”), and pretty screenshots for slideshows and marketing.
The pendulum is overdue for swinging back the other way, but I don’t know who or what has both the capability and will to give it the push needed to send it on its way back.
> 60 year people can't user your fancy site because then don't have an internal model of how a computer works.
I think this is a bit outdated. I'll be 60 in a month, and have been practicing and writing about machine learning, for money, for a straight 10 years now; and I was a young man (and a full stack developer) during the digital revolution.
If anything, GenX had to work harder to get into these brittle emerging technologies and paradigms. There's no-one of my age group, at least that I know of, who is remotely as tech-illiterate as your comment depicts.
Truth is that it took so long for smartphones to dumb down everyone's tech acumen that those of my generation had already learned to do it the hard way.
> practicing and writing about machine learning ... full stack developer ... digital revolution.
my mum, a boomer now in her 70s, would have no bloody clue what you're talking about. she used to work helping out a guy who was doing punchcard programming back when she was young. she ain't dumb. if i broke it down into normal human english words, she'd probably get a sort of idea (or at least nod along to humour me).
i've lost count of the number of conversations i've had with my dad, late 70s boomer, where he complains that they've changed the UI. "It's all different and i don't understand, why did they have to change it? I don't know where anything is now." he's been moaning about things like this for over a decade now (so since his late 60s).
there are definitely technically not-very-literate 60 year olds and the general point about older folks, whether that's >60 or >70, is very real:
older people exist who don't have a clue about SPAs/PWAs, and chances are they're either asking their offspring for help (my mum does this), trying to phone someone instead (my mum does this) or just walking away from it (my mum does this).
60 year olds have been using computers most of their working life. Word processors and spreadsheets having been ubiquitous for office workers from at least the early 90s.
> 60 year olds have been using computers most of their working life.
Absolutely. I am in full-time work, and expect to be for another decade. I have worked my entire career in IT, doing tech support, training, systems design and implementation, tech journalism, and tech writing (i.e. documentation).
I will be 60 in less than 18 months.
> Word processors and spreadsheets having been ubiquitous for office workers from at least the early 90s.
You did say "at least", but still... longer than that.
I started work in 1988 and they were already ubiquitous in my world. Richer companies had the fairly newfangled IBM compatibles, which were still big and expensive. The cheap Amstrad PCs were just starting to appear.
Older hands had multiuser boxes with SCO Xenix or DR Concurrent CP/M or Concurrent DOS and a bunch of dumb terminals. My company had switched to these from Alpha Micro systems running AMOS -- and again, dumb terminals. One of my clients had a DEC PDP-11.
The real old hands had 8-bit kit: some CP/M, and a few BBC Micros.
The first big migrations I saw were from standalone (or multiuser) PCs to LANs, and from pre-PC systems to PCs and Macs.
That would be "60 year olds who have been office workers most of their working life"
60-year-olds who worked blue-collar for a significant part of their life, this is not so obviously true for.
Also probably not true for 60-year-olds who worked in other non-office jobs, like acting or sports.
There have been a variety of well-paid jobs that didn't use computers that a 60-year-old might have done over their life, meaning that this can't even be broken up along class/income lines.
I'm 55 and likely have been using computers longer than that poster has been alive. Regardless of the fact that I started young, by the time I was in college the PC revolution was in full swing and everyone had and worked with computers.
My mother, born in 1934, had no problem using computers. She didn't internalize how they work, but she learned the workflows she needed. How to launch applications and so on.
The situation described in that comment is just a broken app, it has nothing to do with the age or the understanding of the user.
Dude, you know what he meant. Don't be the internet pedant. No need to be the protector of your class, especially one so inconsequential as "literal 60 year olds".
Instead of "old person" he put a number on it. (Cue the people who need to cut me down because I used a male pronoun for an unknown poster)
Yeah. And reflect on the fact that you know that a 404 error is not a form error. Old people won't. Heck then won't understand that a chrome error page is not your site.
I think he means the submit button is disabled because some field failed validation. It might have turned red or have a message appeared next to it - and it might be scrolled off the screen so you have no idea. Or just the common "I have read and agree to ..." checkbox you didn't tick. Possibly because you didn't read the T&C or because it won't let you tick it until you click the link to the T&C. Nobody else read them either but they've learned through trial and error that when the checkbox doesn't work, you have to click the link to fool the computer into thinking you read it.
"Old people don't have that because they didn't grow up with computers."
You know, it's time to stop this trope.
People who are 60 today were born in 1966, they probably entered the workforce in the mid 80's. They probably are not even retired yet. They know how to use computers, they own a smartphone (or if they don't, it's probably for economic reasons unrelated to their age).
As a founder and product manager, this kind of thinking is unhelpful as we design the future. In many ways it's actually ageist to imply that old people are unable to utilize everyday technology.
I was building public service websites (BBC News website) back in the early 2000's where accessibility was a real and important consideration. Technology progresses, and the bar for accessibility has moved up.
My father is about to turn 80 - he checks his heart with his Apple watch, video calls his grandson from his iPad, and asks ChatGPT questions from his iPhone and MacBook Pro. Maybe he's more unusual for 80yo's but it's time to stop this lazy trope that old people are technically illiterate.
(also, shit, I'm only 15 years away from being 60 myself :/ )
I think the relationship people are seeing is "80 year old can't fill out my form, must be because they can't use a computer." But international surveys like PIAAC [0] indicate that adaptive problem solving is the real problem. 30 percent of adults are at or below Level 1; I've copied and pasted what that level represents below:
> Adults at this level are able to understand simple problems and develop and implement solutions to solve them. Problems contain a limited number of elements and little to no irrelevant information. Solutions at this level are simple and consist of a limited number of steps. Problems are embedded in a context that includes one or two sources of information and presents a single, explicitly defined goal.
This test is administered on a tablet, so I think scores can be interpreted as a sort of combo of computer use and problem solving.
A full 40% of adults from 55-64 are at that level or lower. Wikipedia thinks that a novel online form would be a Level 2 task, provided it involves navigating across more than 1 page. Based on that framework and assuming your dad can use the sort function in his email, he represents a top 20th percentile adult for the 55-64 age group. It's probably even higher considering his age as the trend is towards older groups having a rougher time with the survey.
I also think it's ageist to assume that older people can't use a computer. But assessments like this indicate that a full quarter of the US regardless of age would have trouble with some of the basic tasks we associate with computer use. So designers should consider their intended audience when deciding what's simple or not simple enough.
Yes, but even non-technical users have a different relationship that formed over the young (neural plasticity) days.
"I click the <next> button on this form, nothing happens. Given that I have an internalized notion of forms as a multistep flow, maybe it couldn't advance because something is missing on the current step? Maybe it will give me those messages below or above the form? Maybe some read message somewhere? I'll search for those and try again"
vs
"I click the <next> button on this form, nothing happens. Not sure what would happen next. Maybe the thing will tell me? I can't see no dialogs, no messages, no error screen. Nothing changed. Maybe if I press it again? Nothing? Hum, maybe is disconnected from the network? It's the wifi again? Maybe if I power cycle the wifi router it will reconnect? I click next again now I get a clear error message: no internet connection. I suppose my internet is broken again. Will call my tech-savvy friend."
"Why did the bank change the layout? I want the old one back!" - Don't like it? Change bank then. That's what I did.
I get that changing to another bank is a big unknown, but it's probably still worth it to show your displeasure. Plus her bank are morons when it comes to several other things.
The problem with changing banks to search for a ux you like is that it's not easy to see the ux before you've invested in signing up.
My main bank changed their UX not too long ago, and I liked the old one better, but the banks I've signed up for since are even worse. I signed up with them for other reasons though, so I put up with them because it's worth the pain. It does make my main bank look better though --- mobile style on desktop is annoying, but at least I can easily find everything I need... And it's not really their fault there are seven different options on the transfers page (including my favorite: same day transfer vs next day transfers... at one point same day transfers had a fee but they don't anymore so from a user perspective, it's the same thing but you can have it slower if you want...)
Also, signing up for a new bank these days is an exercise in KYC frustration. And then you can't actually transfer your money and use it, because banks responded to the dumbass check fraud that was being promoted on social media by limiting new accounts.
> The problem with changing banks to search for a ux you like is that it's not easy to see the ux before you've invested in signing up.
I agree. Better to deal with the devil you know rather than potentially one you don't.
Thankfully, I switched to a bank with a UI that was known good going by all the chatter I'd heard, but that's not really something you can guarantee to know. And even if you do know, if the rate is drastically worse at that bank than any of the others', then that's kind of moot.
Yeah. We grown being trained to solve those small puzzles that are websites and apps, so we learned _how they are projected_, not how they work.
I mean, we learn that a enroll is normally a flow. Flows have steps. So if you came to the end of the flow and the finish button is gray, you think.
Hum… I'm used to flows. This is a multistep flow. Flows normally need me to fulfill some small checks and won't let me proceed between steps if something is missing. But some won't. Maybe this is one of those? Some flows have warnings in the end, some have next to the thing missing. I don't see any warn in the last screen, so I'll go back every step and check field by field for errors. That'll probably do.
This is the model you have in your mind, of how a website or an app works.
People that came to computers, apps and websites later in life didn't learned the puzzles.
Next time I have trouble checking in on an airline site I’ll remember that there are so many other sites to interact with that whatever I was trying to do probably doesn’t matter.
I wouldn’t sweat the broken fridge either though, there’s so many other electrical appliances in the house to use.
A lot of the people I build web pages for are poorly educated. The text we use for web is written for people with an eighth-grade education. Print material is fifth-grade.
People in the SV bubble can't imagine that there are tens of millions of people in America who cannot understand how an SPA works.
These people are invisible, even as they ladle out the food in the Google cafeteria, and polish the chrome in the Meta lobby.
I'm going to link my favorite survey [0] to illustrate this point. Its called the PIAAC, and it's a internationally standardized assessment of adult skills in literacy, numeracy, and problem solving.
28% of the US is at Level 1 or below. That level roughly corresponds to reading a simple text and being able to answer a question about it that requires limited inference (i.e. if the question is "what color are pigs," the answer would be in the text as "pigs are pink in color.")
It's still present. JSON/JS parsing still has a delay. And in either case (as the author states) not everyone is using an iPhone over 5G. Heavy React apps are a miserable experience on low end Android phones, even when the connection is fast. I've seen JS/JSON parsing times in the multiple seconds.
Read the article. Typical users had old browsers often with poor reception. One user was using a PlayStation Portable which had very limited WWW capability.
"What, support Safari? Isn't that, like, less than 20%? And its standards support is abysmal! No, not worth my time, they can upgrade to a normal browser like everyone else."
If Rick Rubin could take a tape to his car to listen to his mixes, your product people can try their websites on £20 phones from Tesco. They can ask to sit in on user tests with minority groups. Extending your knowledge like this is trivial, but rarely done.
May i ask why, specifically, Rick Rubin? I don't know who that is, but whenever we finished mastering a new song, we had a series of "systems" we listened to it on. We went out to my dad's work van and listened there. We called up our friend with a street-comp sound system in his car, and listened in there (neighbors must have loved us!), and then a "cheap" boombox with large-ish speakers but cheap.
if it sounded "clean" on all 3, without the bass muffling everything, and the highs not hurting the eardrums, we called it "good" and released.
i don't work in the industry, sorry. We just made music and released mp3s, 1997-2008. Co-creator of Def Jam, alright.
I wonder if David Lynch watches his stuff on a tiny screen just to make sure everyone has a good experience.
hint: no, he thinks small screens are stupid.
ETA: after like 3 years of mastering and reviewing this way i trusted my ears and my studio monitors enough to know what it would sound like. I also wrote in headphones and mastered on speakers, then remastered in monitor headphones. Anyhow, i think the whole point i was making is "yes, this is a good thing to do, for music, for websites, for software, etc"
This is a matter of perspectives here. Some of my friends are absolutely brilliant lawyers and they trump me in their reasoning abilities (while I am a typ. HN member, high-tech engineer). Yet, they would not know some tech basics. I see lawyers struggle with formatting in Microsoft Word all the time :-), for one.
A brilliant physicist friend likewise once told me he is clueless how real numbers are handled in computers (not talking about floating point encoding specifics but conceptually itself). Yes, I can say that feels dumb, but I cannot deny that he is a brilliant physicist.
You're not a good and modern engineer who knows his craft if you aren't defaulting to react and tailwind.
And don't dare to contradict me, the fact that MIT-bred leetcode ninjas paid half a million per year can't produce a simple (mostly static) website on that stack it's only because of management that wants to ship the next product. /s
The landing page for our latest tool is a static website (exported from nextjs). It has no marketing on it. It just has simple instructions to copy a prompt into ChatGPT / Claude.
I think it is a good idea. JavaScripts and CSS can enhance it but should not be required.
Accessibility functions are also helpful; if designed well then it would be helpful for everyone (not only if you are blind or other disabilities), in many possible circumstances.
I've tried this before and I think this constraint is something that has to be kept top of mind for a designer not just the engineer. Most designs these days assume a single page app and there are interaction patterns that make plain HTML not suitable. But if you incorporate this in from the start and stick with it, there's no reason you can't do this.
My go-to for spinning up a site has been Jekyll + Bootstrap with the occasional bit of React for well over 10 years now.
While it still does the job, I'm a little curious to explore more modern options, if for nothing else to understand the choices a more junior dev would face/make today.
I'm seriously considering giving Atro a go. Is it worth it?
I've started getting traffic on my website only after I re-build it with a locally-brewed MD parsing engine that uses Astro to spit out the final version of the site.
I guess the main argument is how easy it is for an LLM to ingest the content, since I can bet all of the crawlers are llm-enabled one way or another.
I was a little confused by "doubled our users" since that's more about inbound traffic than site experience. I guess it's really shorthand for "halved form abandonment" which is still pretty great.
I think that's even more significant, since it's measuring people who cared enough to click the form in the first place, which is juicier than just page loads.
> Javascript and modern CSS should be used to enhance the experience
When messing around with my blog's Javascript, this mantra is so thoroughly embedded into writing it, that I try to include "enhance" in function names where it makes sense. I might have to do likewise with my CSS.
All the text is invisible for me in Firefox on Linux when the `--font-body` is set to `"Atkinson", sans-serif`. Setting it to `"Atkinson Hyperlegible", sans-serif` fixed it.
I still to this day is baffled by all the idiot geniuses who thought it was OK to have "Please enable javascript" when opening a website without JS enabled. A good chunk in many pages are just static content. You should be able to render HTML. Ashte!
Recently I had to migrate an old SpringBoot app that had a React front-end to a new cluster. Not wanting to mess with super-old dependencies, I opted to rewrite it on a new version of Java/SpringBoot. When it came to the frontend, I paused. I couldn't come up with a single good reason why this app needed React. I rewrote the frontend in straight HTML with a little bit of JavaScript for DOM manipulation. I literally used `var` instead of `let/const` just to drive the point home... (yes, that was overkill). But you know what I didn't need? A BUILD PROCESS! No npm deps. No vite/rsbuild/etc. It was like I had forgotten we could even DO that.
Don't get me wrong, I actually have enjoyed React over these past 10 years. But, including it blindly is just silly.
esm.ah let's you include "complicated" JS that isn't usually found in CDNs.
it doesn't work for everything and imo is worse for (p)react due to the lack of native JSX, but it does allow for bringing in stuff that usually takes an `npm install && npm build`
Yeah in this case, I needed to pop up a <dialog /> and take some form info, persist it via POST and then show the result of a "used" card/token. So there just wasn't a lot of need for libraries. I'm from the VERY old school so I do recall the fresh hell of including many deps via script tags (pre-Bower!)
If you're a "React person", as the article puts it, friendly reminder that you can render components to HTML and serve that to the user.
I have done exactly that on a project that was under similar constraints. The UI models live in .tsx files and the browser gets pure HTML with zero JS by default.
We are still in the early days of software and web development. This is the food service equivalent of a restaurant operator discovering that if they serve dog poo vs edible food they get more return customers.
in the bio ... "has over twenty years of experience building highly accessible and usable web applications"
why not take the html5 standard (see https://html.spec.whatwg.org/ ) and if needed (dont think so for these use cases... "for clients ranging from energy companies to political parties") htmx or alpinejs ...
No, it's actually people. FTA: 'My client was a utility company, and they had a big problem. To apply for their services, customers could either use an old ASP form on the website...'
Ie customers of the utility company were completing the form, not random users from the internet.
The downside mentioned by the author's replacement in the article is the unfortunate explanation for why this is rare in practice.
> "but that’s a lot more work for us."
And it's not that any individual or team is lazy. Most teams have a constant barrage of priorities to balance and are paid by companies valuing efficiency over everything. That said, I think the article makes a great case for adjusting our prioritization. Going a bit slower won't kill anyone, in fact doing so will probably save some.
Maybe this is heretical in today's AI hype climate but...weirdly due to the rise of AI, then AI-slop polluting everything, a lot of old fundamentals are coming back. Clear, well-structured, descriptive content on a well-built page has a better shot of being picked up for SEO/AEO/whatever which are the same best practices from 2005. A lot of these tips and tricks and hacks just aren't going to move the needle as much anymore imo.
SEO and accessibility laws have always been the most effective way to convince someone to build clean, well-structured webpages. Guess what, both are measures of how easy it is for a machine to extract content from your pages. AI is just the latest machine that wants to slurp up your soup of tags.
HTML doesnt have hundreds evangelists writing blogs about how cool it is. So it doesn’t really matter what the benefits are if juniors are going to keep showing up writing client-side react components. And the product managers letting them because of their sparkling (but naive) personality.
I will never stop singing the praises of plain HTML. It's accessible, it's portable, it's simple, but for some reason we need horrid JavaScript nonsense to operate.
It will still work 30 years from now! 300 years! 3000 years. There will be no strange error messages in the log, no browser updates breaking stuff.
Personally, rather than this luxurious approach, I just do one giant form and store all values in local storage. If something is wrong have one message at the top listing which fields failed validation and why. Generate some css to put a red border around the fields.
Local storage might not be a good idea for such sensitive data but if you can get away with the simplicity it's lovely.
Totally agree, gov pages should be widely accessible. Also gov services should NEVER mandate internet access. There should always be a way for tech-illiterate people to ask someone, and fill their forms
Yes, there 'should' be a trail. And there's someone and a process to check that trail. And a process to find the issues, correctly so, and fix, and do all that on a timely basis, and reimburse the person for damages, ...
PS: I have actually suffered through some of this, badly so, though in an unrelated setting.
It comes back to which system is more robust/mistake-proof and more practically recoverable.
>She’s connected to the complementary WiFi and is browsing the GOV.UK pages on Housing Benefit. She’s not slicing fruit; she’s arming herself with knowledge.
>The PSP’s web browser is - charitably - pathetic. It is slow, frequently runs out of memory, and can only open 3 tabs at a time.
Alluring, an annoying property of private software development is that making websites and software in general inaccessible to lower end hardware is actually a positive effect, as it filters out 'undesirable' lower-income prospects.
That, along with pressure to produce fast, without much concern for quality (with notable privileged exceptions of luxury software like Apple or 1B+ user software like Google), as well as a disregard for sourcing "I don't care if you do it yourself, or npm install software from effectively unpaid volunteers", ends up in a state of software lacking craftmanship, software that one is not proud of to work in.
Actually, in English, Chai does not mean tea, it means a specific flavor of tea. If you don't believe me, try ordering some Earl Grey Chai, see what happens.
it's redundant at a place that serves chai, but it isn't redundant at a place that does not serve chai, because you're skipping the "what is chai" question from whoever you're querying.
Not in the Wikipedia page (but check the Italian version): it started as "Mons Belli" (Mount of the Battle) because of a battle fought by the Romans a few years before the Hannibal campaign. Then the original meaning was lost and it gained another "of battle" in the 1800s. Mount of the Battle of the Battle. Hopefully there won't be another one to add.
Except places are now offering Chai Latte Coffee so if you don't specifically order Chai Tea Latte, you could get some thing totally different than expected. I learned this the hard way.
As a teenager I remember going to a website for... a city, I think? And their 'sidebar' was a Java applet that did nothing but provide links for you to click with on-hover effects. The page used frames; the applet was in the left-side frame and the content was in the main frame on the right.
The applet took 30 seconds to load. Once it loaded, it showed five buttons to click to get to different sections of the site. When you clicked on one, instead of changing the content frame, it sent you to an entirely new frameset. This, of course, caused the sidebar to take another 30 seconds to load. Hitting the back button did the same thing.
Meanwhile, I knew someone whose friend made a little applet that he showed me; it was a Java applet that you could provide an image URL for and it would load the image and then, below the image, show a rippling effect as though you were looking at something on the shore of a rippling lake. This applet took less than a second to load and ran incredibly smoothly.
Java was a curse, not because Java was bad but because Java applets were written badly and used badly simply because they were neat.
Every language can say that bad developers write bad code with it while good developers write good code with it.
I would like to say the early interweb was just a learning experience, but today's interweb hasn't learned any of the lessons. It's just changed which language the lesson is being relearned
A lot of these tools, like React, are designed to embrace, extend extinguish the web. Why Microslop and Zuckerberg spend millions of dollars of dark PR claiming anyone who doesn't like React doesn't know what's going on is because it makes the web worse and less useful, which means you spend more time talking to Co-Pilot or bots on Facebook.
I did some work for a company that spent nearly a grand on a Flash animation for their title page of a red bouncing ball that would bounce from right to left along the letters of the word "Yipee" (yeah totally not ripping off Yahoo! were they?) until it landed in the crook of the Y, where it would spread down the middle - the finished logo had the Y made out of blue, yellow, and red stripes.
Every single person I showed it to including my then-70-something mother said "that just looks like menstrual bleeding".
Every single person said that.
They still went with it. Conversion rate? Dunno, never got numbers high enough to test the script.
As a non-web dev, I have a question about this part:
> There was a sad coda; as is the way of contract work, I moved on. I explained what I had built to my replacement, that it always worked even without javascript. He was appalled and said, “but that’s a lot more work for us.”
Why is it more work? The approach described in the article seems honestly reasonably simple: just write the standard <input> components for the form, have a submit button at the bottom. When I was making my own websites many years ago now, that's how it worked, and it wasn't that hard. Maybe it's reflecting my ignorance in this field, but doing fancy front-ends seems much harder to me.
Starting a few years ago, I realized some junior and medior engineers never once considered the possibility of building a website (app, experience, etc.) in anything other than a heavy SPA framework. But they're not stupid people! If you directly asked "Can you build a website without React?" they know the answer is obviously "Yes." However, if you asked them to build a new website, they would unthinkingly start a new React project, mostly out of familiarity and a desire to get the job done.
A few of them would outright not know how to do anything else. No knowledge of how to stand up a boring HTTP server to send pure HTML. No experience building a form that validates or submits without JavaScript. These are not the people who post here on HN. They are not engaged in online discussions of new tools and skills (or old tools and skills!). These are people who learned just enough from a bootcamp, or their uni's single "web apps" course, to get a job. Since then, they have just-in-time learned whatever their employer required, or whatever particular tools someone else on their team chose for a project.
As an old, it took me a while to recognize/realize it, but I understand them now. Depending on their career path, someone will encounter the simplest aspects of HTML, CSS and vanilla JavaScript after they learn the complex, framework-specific aspects of each. It feels (to them) like more esoteric, advanced, or tertiary knowledge.
Tying it back to to the quote "that’s a lot more work for us", that's not necessarily an intentionally false claim. It probably does feel like a lot more work to perform a task using unfamiliar tools, even if they are less-complex tools.
You are far too empathetic to them. They should not hold the jobs they have.
These are the people writing React monstrosities for government benefit websites, and testing them on fast iPhones and fast 4G, without realizing that every page load for actual users will take 30 seconds on their old $200 Android on 3G, and users won’t complete the form.
It’s a culture of not giving a shit, that’s the deeper issue.
I had a contract once to save a government website that had serious performance issues, it was so unusable that people preferred to go in-person and wait 4h in a queue rather than try to fill the forms online.
The frontend was in React because the company that got the contract initially used React for everything. The frontend was a 5MB SPA, but it could've been (mostly static) HTML files with some interactivity for forms like TFA. Everyone working on the project agreed React didn't make sense, but we couldn't do anything about it because someone from the government IT department would have to admit they made a mistake. There was no budget for rewrites in the contract. The few times a developer attempted to remove any "React monstrosity" they got in trouble.
Sometimes developers care, but the people in charge don't, and in government environments every change must go through them first.
I’m curious if - and when - LLMs change this. They’re very good at web apps. And they’re great at rewriting existing stuff. Just give them a well scoped /goal and go get coffee.
Theres lots of open questions about the future of our profession in the age of AI. But, playing with opus and fable, I think the future will be bright for our users. There is no reason any more for teams to put out junk that’s worse than what an LLM can do.
Unfortunately the LLMs are trained on what we've made, and there's going to be a ton more React garbage[1] in the training set than there are carefully-crafted websites like the article describes, so I don't expect a decrease in overengineered, bloated junk. If anything, I predict that the fact that you can shit one out in less time than before will have a different effect: A modest increase in bloat since an LLM won't mind adding a half dozen redundant and competing ways to do the same things in a large codebase, combined with a shorter mean-time-between-full-rewrites.
I think most of us have seen incredibly creaky codebases that are too buggy to be maintained any longer, where we make the hard choice to wipe the slate clean and build a new one.
We might find those rewrites happening every 12-24 months instead of after a decade.
[1] Frontend people, I mean no disrespect -- just that React & friends are (ab)used for nearly every website now, even those which map perfectly onto the "Simple document viewing with occasional submission of incredibly simple form data" model that plain HTML has always been perfect for.
> Sometimes developers care, but the people in charge don't, and in government environments every change must go through them first.
To be fair, the same thing happens in private companies. How many UI changes have people gone through that didn't actually make anything better and just made everybody relearn everything? We would have been better of scrapping many of those and let people continue to use what's already familiar, but that too would have to involve someone admitting failure, which is a hard thing to do for some people.
I've used many a government website in the Navy, and they were almost invariably bad, but it had nothing to do with React per se.
A very slow website I can think of had something like 200 GET requests required to load the landing page, and it used Liferay with Material Design Bootstrap. That was closer to the "style at the time". React is the style of this time, but you can write very slow websites in anything, I'm convinced.
> That was closer to the "style at the time".
So I tied an onion to my belt.
In Canada you can't call yourself an engineer unless you have some kind of association behind it; the title holds meaning including partially accountability. Something that is lacking in the tech world. I'm not saying I want to live in that world but also I worked hard for the knowledge I have starting in the IE days of web dev; it was hard earned experience making things work across the web without loosing performance. The idea that we have developers out there now getting paid higher than me that are clueless on how auth works, how the browser works, why css and browsers maintain backwards comparability for a reason.. well it's sad; but good for them I guess?
The behaviours of developers as well being beholden to their managers rather than the craft; meaning not saying No we will not move forward without proper unit tests, or pushing back when business demands quick corner cutting solutions.
Anyway, decades of bitterness. I wish we had associations to uphold some level of accountability on developers as much as protect developers. I think things would be a lot more expensive and slow if we did that though.
Fundamentally I agree with your take, not just on dev side but just the web/dev/produce' a culture of not giving a shit.
Junior and midlevel devs aren't decision makers for government benefit websites. The culture of not giving a shit is real, but the responsibility goes far beyond these roles.
If we're talking a government site, chances are you don't have the budget to be able to hire much above junior or midlevel devs. And the project manager probably has a small budget [^1] and little experience with what the web design choices really mean (and what the trade off are).
I think you'd be surprised who ends up making those decisions.
Which goes back to the original point (that's valid for any project) - keep your user in mind. If your users will be using recent-ish iOS or Android devices, use as much flair as you'd like. If your users will be using mass-market low-end devices or used devices from 4+ years ago, then maybe dial down the interface.
Knowing your user is important, no matter what level you're at.
[1] Unless we're talking about some kind of large system that's being redesigned by a consulting company on a cost-plus contract. Who knows how those decisions are made.
Even if this were the case, and I wouldn't be surprised, it's still misplaced blame.
> Knowing your user is important, no matter what level you're at.
I agree, but it's absolutely ridiculous to expect a junior dev to make excellent decisions on this. Software development is a massive industry with no prescribed methods. It's not like these folks are going through a residency before getting the job. Even if they went to uni for CS those programs don't teach these skills.
I am always baffled by people who blame developers. Like some mid dev or junior would calling shots what stack should be used for project.
You'd be surprised, then. Some managers don't know squat. I rolled onto a project once and found that an entire application was being delivered as a 300MB ActiveX control, to run in a browser because that was cool and "cutting-edge" at the time.
Looking at the code, I found it was using UI elements for data storage and other such nonsense. A colleague and I had to tell the manager that the entire thing had to be rewritten. I'm not sure he actually went pale, but that's how I remember it.
It is EXACTLY the type of people that are hired to make decisions, because of either nepotism or impressing with portfolio filled with overcomplicated, 3.js frontpages.
When you give the project to a bunch of junior devs the stack is necessarily decided by one or more of them since there's nobody else to decide it...
The tech stack is almost always decided by someone in leadership that has no developer experience. Or by the consultant company that will chose the most complicated and difficult to maintain stack because then they can invoice more and will win all future contracts. The trick is to hire someone that is not corrupted by money, someone like the author of this post, who cares more for the users then how much he gets paid.
Preach! Amen! Hallelujah!
New cheap android phones are just as slow as old cheap android phones. The bottom of the market has been stuck in performance limbo for years, and modern web dev frameworks are ill designed to meet them where they are at.
My cheap phone has 6 cores and 8GB I think. My first cheap phone had 1 core and 256MB. Both ran Android. I think you're mistaken.
old != oldest
They are referring to this:
https://us-east-1-shared-usea1-02.graphassets.com/cluuijofv0...
This is a leadership failure. People who want to do a good job always get fired for taking too long.
I've found what works really well on 3G an MPA with streaming HTML with brotli compression rendering the whole page on every change.
I use to have an old pentium 2 computer for testing websites. Sometimes you cant make things fast enough for the old box. A fun trick is/was to have <script>elm.textContent="loading images"</script> between each "heavy" section, all targeting the same elm. If the computer, network or server is truly extremely slow you will get a nice message at the top describing what they are waiting for. On a normal slow computer you won't see the messages unless something went wrong.
Oooh! Like a status bar!
shit, I'm too old to remember those...
It's more of a culture of "but everybody else does it".
I like how HTMX does SPAs. It straddles the divide nicely between simple and capable.
I just had one of these people, a contractor working for a state government, argue vocally with me in a meeting stating that "500 JavaScript requests is not a problem" for a single page. Un-cached, of course, despite there being a CDN in front of the site.
You can't win against cargo-cult coders because they just assume you're from a different, competing cult.
They have no concept of engineering or science, they have never encountered it.
Heh this is one nice thing about doing engineering work in Australia. Our round-trip time to US data centers is often about 200ms. There’s no hiding from sloppy choices in the performance panel.
I had an argument a few weeks ago because our page took 4 serial requests before content appeared. I argued - with solid data - that it should be 1. If we could manage that, cold load time would ~ halve.
I see no reason not to be empathetic. The frustration is fair, but it's aimed at the wrong layer. These people were guided into this spot by bootcamps and curricula that start at React and never go down the stack.
My experience was the reverse. I learned HTML and CSS first, then Rails in college to serve templated pages. I understood the client/server boundary fine as a concept, what I couldn't see was where it actually sat in a web context. I sort of knew JavaScript ran in the browser, but then I'd see ERB templates stamping values directly into script tags, so the server was writing the JavaScript that ran on the client, and my mental model fell apart. Where does my code actually execute? Why does this variable exist here but not there? Why does the page have data the network tab never fetched? Nobody ever sat me down and explained the request/response lifecycle as its own thing. I had to assemble it from fragments over years. This was around 2017 for context.
How you learn something shapes how you keep learning. If your mental model is misaligned, everything downstream is friction. The thing that finally made it click for me was reading the actual HTTP RFCs, which is apparently a weird thing to do, because HTTP itself is absent from nearly every guide and curriculum. Tutorials teach you the framework, maybe the language, and just assume the protocol underneath. These days I make newbies read the MDN docs like a book and skim the HTTP wiki page, learn the history of the protocol. It's short! It's not even a book! That gives you a firm foundation. But if your foundation starts at React, drilling down is like digging past bedrock. People don't know where to start, and Googling only shows them wrong answers because they don't yet know how to ask the question.
Are you sympathetic to a doctor who specialized in surgery and now always recommends surgery, even for a common cold? Or would you say they are in the wrong job, if they are anywhere but surgery?
Well that's horribly reductive. I certainly do not expect everyone in a given field to know absolutely everything there is to know in that field.
Crazy enough, I also hold doctors and surgeons to higher standards than web developers.
If those web developers fail a critical government service, or online pharmacy or financial service, it can mess up peoples life pretty badly as well.
Ridiculous example that does nothing to argue the original, fair point. Obviously health interventions demand more finely tuned solutions than information technology
FWIW, maintaining at least a moderate degree of empathy even in systemically frustrating situations is good for the empathizer and thus in one’s interest
Yup. It’s exactly like the dismal state of CSS frameworks we're mired in.
All these new kids walk in and learn the CSS framework du jour first, then find themselves stranded when things move on. If they had just learned CSS the first time, they'd be set for life.
Nobody should learn React before learning HTML and vanilla javascript.
HN last week: Learn SQL Once, Use It for 30 Years https://news.ycombinator.com/item?id=48347483
The practical question most face is which is more likely to land them a good role. Does standards level CSS knowledge meet that bar?
(I’m team CSS standards although my knowledge is a bit tilted towards late 2010s, currently revisiting)
Kinda sorta analogous to the cloud engineers who can standup complex monstrosities in AWS-land, but don’t know the first thing about how to troubleshoot say a connectivity or simple problem where they have to ssh to an ec2 box and do the needful
well... just because you know how to ssh into the prod DNS host and manually update the prod zone files in vim to remove orphan A records + duplicate CNAME records, in order to fix an ip address exhaustion issue that is blocking new VM's from spinning up for your customers... it doesnt mean that you should, lol.
that was 8 years ago in my first gig. now i kinda wonder... having those skills made it easier to put off implementing a robust long term solution. it was playing with fire, sure, but i was a rookie
or the materials engineers who are great at making mems tools, but couldnt for the life of them design an aircraft prop
I’ve long thought frontend web developers are the ones most threatened by LLM-assisted programming for a bunch of reasons and now I can add “many don’t understand web fundamentals” to the list.
Don't you think they have a legit skill issue here and should they be better off upskilling themselves?
This is a direct effect of being a low barrier industry to enter. Most of the ppl among us are mostly here because of a good paycheck. And it SUCKS!
>Don't you think they have a legit skill issue here and should they be better off upskilling themselves?
Absolutely agree! Just because I understand how they got there doesn't mean I think it's a good state of affairs ;)
My post was already quite long, and I didn't want to append a treatise on what one should do when encountering those engineers. It depends on many details. Avoid hiring them, if that's a power you have. If you are stuck working with them, depending on your authority, encourage them to learn or force them to learn. If you're coming in to clean up after them... well, hopefully your comp is worth the annoyance.
We are all simultaneously in the position of encountering "the world as it is", understanding it, and doing what we can to improve it.
Yep. It’s also an attitude problem. A lot of devs are able to up-skill just fine, but some are downright demeaning towards anything they don’t understand, or towards anything that doesn’t come from a FAANG.
“HTML only? Nobody is doing it!”
In the olden days, people wouldn't take office jobs or factory job necessarily because they thought: "Yes! That's my passion! That's exactly what I've always wanted to do." Passion isn't your first and foremost thought when you have a family to feed.
A few decades ago, IT jobs were for the most part done only by people who were in it for the kick they got out of working with computers. They already hacked at their dad's computer in their early teens (or sometimes even younger), and just could just never let go. It was for people who loved it because it was a niche.
But today, IT is no longer that. It's the backbone of much of our society. And so the field no longer attracts just the die hard fans, the nerds. It attracts ordinary "career people", who just need to have a job to feed the family. Who turn the machine off after 8 hours. Who don't go on coding all through the evening on their hobby project. Who don't try out new tech just for the heck of it.
I think it's hard to understand if you belong to the first group, the nerds, that anyone working in the field isn't like you. Because they all used to be! But those days are gone. We live in the times of enshitification for a reason. If you have the hacker spirit, you don't enshitify because you simply can't. You know what is the right way to do it. Sometimes that's a React app but sometimes it's just an HTML page.
You're not just in it for the money. You care. Not necessarily for the end user, although that would be nice. You care for the tech. And when - like in this article - both come together, sweet things can happen.
Absolutely
I'm not a web developer. I built a few websites in high school, but these days I write safety-critical real-time code for robots.
A few years ago I was back in grad school and I took a class with undergrad senior CS students. We had to write a fairly simple web service, and I was blown away by how complicated they were making it. Based on the requirements we easily could have written 90% of it in plain HTML, but everyone else insisted it should be 100% react. Part of that is honors students wanting to do everything the most complex way possible to impress teacher, and part of it is them simply not knowing that other options exist.
A few of them would outright not know how to do anything else.
It's like how a lot of people these days reach for an electric drill/driver for even the most simple projects like tightening a screw. It never occurs to them to use a screwdriver, or even a butter knife.
Yeah, my dad is like this. But even just for one screw I am at least 10x faster with the electric option.
But I also achieve faster HTMl with vanilla and never touched react, so .. I will continue to stick with using the right tool for the job.
Try to get a Java developer to do something without Spring.
There is a reason we use React today - if you want dynamic and complex site, HTML is PITA to work with. Either you reload everything on any change (not good), or end up with a maze of partial reloads in event handlers (not good either). It was done, it was bad, React (and similar frameworks) was the solution. If you don't need complexity of React, fine, use HTML. But if you do, who can guarantee that 5 min later someone will not start requesting site not to reload here and there, but update this bit here. So I don't blame people for just using React. And in this case the issue wasn't that react was used, but that it was used poorly.
As someone who has built both react based frontends and html based ones (with htmx), there is a law of diminishing returns at play.
To start off, writing a basic crud website with forms is much easier with htmx.
But when you start building more complex components, and integrate with other systems (OAuth for e.g.) there are tons of libraries and SDKs for the react ecosystem, but not many for pure html components.
At this point, it's much easier to use off the shelf components than it is to manually write html to handle all the bizarre UI edge cases.
Two years ago, I started a new company, and decided at the outset to avoid using any heavy JavaScript SPA framework. We stuck to simple server-rendered html and only use progressive-enhancement style JavaScript.
Our app was fast, and simple, but it also came at a cost: we were limited in our ability to take rich UI elements off the shelf with an npm package. We had to do a lot more work to provide a rich user experience. Everything took longer, and the user experience was worse as a result. We cared, but sometimes you don't have time to carry through.
The company failed, and I don't think react would have saved it. But I can tell you first hand that righteous adherence to "simplicity" didn't help either. It's always a trade-off.
I also prefer simple web tech, but I'm really glad you brought these points up! Ecosystems matter more than a lot of purist devs think.
> Why is it more work?
As someone that reads a lot of code written by others, I'm confident that "learning a new way to do something" is perceived by many as the hardest thing in the world.
It probably has to do with what technology people are used to. There has been a couple of generations of web developers who have only known javascript and its ecosystem for building webapps, and so anything other than pure javascript solution looks foreign.
A lot of developers have made or just perceive very strong silos between frontend and backend today. Any coordination that needs to happen between frontend and backend is potentially a communication challenge.
It seems like a lot more work because you have to keep the backend and frontend in closer sync. The backend has to be aware of and able to store every sub-form in the full process (which sounds like a "wizard form" with a multiple sub-forms to get through the full "form" process), not just accept a "finished" or "complete" submission. If a sub-form needs a change the backend, the backend's storage, and the frontend all need to change. The backend and frontend have to agree on validation logic for each small piece of the form. The backend and the frontend need to both validate every small piece of the form, and maybe can't share that validation logic (depending on what language the backend is written in), especially if one of the goals is to do as much of the frontend validation as possible with Browser native validation tools (`<input required` and `<input type="email"` and so forth) so that you get the most benefits of progressive enhancement.
The original ways of making websites were "full stack" and from a full stack perspective it shouldn't seem that hard to have a coordinated frontend and backend, especially when a progressive enhancement approach likely means a smaller more agile frontend, but current siloed world where frontend and backend are different teams with different goals and alignment makes that seem like way too much work.
This was the jQuery way. It was called Graceful Degradation.
The entire approach went out of style with the advent of single page apps, React, Angular, VueJS, etc.
That's generous. I always heard people espouse that ideal, but I rarely saw them actually do it. And I never saw it at work.
There were always certain UX requirements that required JS, and that meant the company wasn't interested in testing to make sure it worked without JS. None of their customers were going to use it that way.
Angular, React, etc helped force it further, but they didn't cause it.
I know I always did. CakePHP and Rails made it really easy to determine if a request came from AJAX or direct and you could slightly tweak the response to match the medium.
Agree that most people didn't, but I was always an advocate.
the shockwave and flash way was simpler.
download the whole app and run it in browser. you can even run it off line!
jQuery and graceful degradation are different things. The vast majority of sites in the jQuery era that used jQuery did not gracefully degrade.
I think the problem is hearing just one side.
Someone is saying that they delivered a very reasonable solution that's simpler than most would come up. Person taking over was not happy.
Do we know if the code being handed over was high quality? Were they reacting to the fact that it was "not React"? Maybe they have a template they enforce in the company about how apps are built?
We don't know.
As an application becomes more stateful it gets harder to keep that state aligned across the frontend and backend, especially if they're in different languages.
The short and sad answer is that most people that work in web development do not understand how the web works or how web apps work.
It is harder to do more with less. There is a reason React and other is used so much. Makes it easier to make interactive websites. It’s like asking why backend engineers think it’s harder to code an api using C instead of using Django.
I used to think that was true but I now think it’s only true for very interaction-heavy apps: if you have hundreds of interactions on a page over many minutes, using an SPA is amortized across a lot of time, but if it’s something you could do with e.g. a simple Django app you’ll not only be done faster but will spend an order of magnitude less time on maintenance and accessibility work.
I used to do web dev. (Got out of it due to js frameworks and their bloat.)
I'd pee myself in happiness to take over a project like this.
I was confused by that too, but then I thought maybe the new replacement was commenting that using javascript in the app would provide more work for contractors, which he perceives as a good thing. So not using javascript provides less work. I could be wrong though.
They don’t know how to compose and style elements without someone providing them with a prebuilt library of $framework components.
As a web dev a lot of this is simply ongoing maintenance of a largely unknown quantity. Most web devs know React and use it extensively; Astro is something they'll have to learn on the job or hire for specifically.
It's akin to writing a backend in Haskell. Chances are you could write something performant that leverages FP in a way that serves as a magic bullet for your domain. But now everyone after you needs to learn Haskell and how to model all future problems in a way that conforms with it - or rewrite things again.
Not a web-dev myself and I was wondering if, apart from unfamiliarity with astro or HTML being treated as unknown technology, it also has to do with having to handle fallback cases, eg the 3 point validation (web component, browser default, server), esp when one is used to have react (libraries) just handling it all without any more considerations.
> Astro is something they'll have to learn on the job or hire for specifically.
Before LLMs I would have agreed.
LLM + framework you don't understand goes in ... unmaintainable garbage comes out.
Before LLMs, learning on the job looked like reading documentation. Now it’s a guided tour with verification. When I produce things in this way, I’m not just blindly accepting it. The goal is that by the end of it I have learned more about the codebase and architecture, not less. I feel that’s important.
Many people don't understand this, even big tech engineers. They see LLMs as a bottleneck. It's more that they don't understand how to use it to multiply their skills, just basics and code gen.
I use multiple Claudes at a time, daily. It's precisely because of that experience that I wrote:
Claude follows code patterns and structure. If you setup that structure and those patterns properly, it will produce great code. If not, it will follow ... whatever it feels like, with each commit.If you just have it built something with a framework you don't understand, it will do so just fine! But over time every "vibe coded" change you make will drift it further and further, until you are left with a mess of vibe-coded spaghetti.
Using it to understand a framework is fine.
I agree that's fine in that at least it's doesn't cause unmaintainable garbage. And might even get you up to speed quicker that reading the docs old school.
But the GP point, that you're better off finding people that already, truly understand and are familiar with the tech (ie. Astro), imo still stands.
We cant all be wise enough to use php.
I read a fun comment the other day from a frustrated windows user who failed to configure linux in previous attempts but now with LLMs it was very easy for him.
By this point people don't appear to have any real clue how to write HTML anymore. Writing semantic HTML isn't significantly harder than say writing Markdown. You copy some HTMl skeleton and you literally just stack your elements into the body. I managed to do that as a 13 year old on MySpace without any deep instruction. Sure you have to close elements as well so the syntax is slightly harder than markdown, but that allows you to differenciate between for example <article>, <section> and <aside>.
I am convinced the one single thing that made HTML unusable over the time was that people wanted or needed a way to re-use parts of the page across multiple pages, like headers, navigational elements and footers.
This meant people used frames, PHP, templating engines or any other new technology mainly for the purpose of creating shared elements, simply because HTML failed (and to this day: fails) to offer a way to include one HTML file in another without having it suck (like frames definitely did, since the browser treated each subpart of the page like its own entity including caching).
> (and to this day: fails)
The `<template>` tag has gotten very close. Right now you still need JS (optionally wrapped as a Web Component) to load a `<template>` from an outside HTML file (at which case, yeah, it's so easy to just use a JS-based HTML renderer instead of a template today), but discussions are ongoing about closing that loop for simpler "JS-free Web Components".
I don't know when that will be accepted into the web platform, but it still feels more like a matter of time that it may happen eventually.
I've found at least some of my static page generation has moved to just dumbly appending `<template>` tags to the bottom of a page rather than use some other template language, so it feels like we are closer every day to finally having "HTML-native" simple part reuse.
<div><div><div><div><div><div><span></div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div><div>></div></div></div></div></div></div></div></div></div>
writing straight html and css sucked. building reddit css themes sucked
But I believe the point is: if you were writing HTML by hand, you would never have written the above.
Large websites resorted to PHP and server side includes to get headers and footers. Smaller websites resorted to frames and copy/paste. It wasn't perfect, but it also wasn't horrible or unusable either.
Ikea relies heavily on Edge Side Includes, assembling static pages from parts at the CDN level and then having small islands of JS interactivity
I think it was... SHTML? that allowed for server includes. My recollection from... 25ish years ago was that it was generally quite well supported and worked quite well (and was dead simple to implement). Not sure why, if that was the issue, the fix didn't quite catch on (but it's totally possible I'm mis-remembering the state of browser support).
Server-side includes worked fine but weren't enabled by default in any of the mainstream web servers. I think the lack of default-enabled status hampered their adoption. Joe User couldn't just FTP a bunch of ".shtml" files up to their shared web space and expect it to work right.
I certainly used the heck out of them in the late 90s, though.
It would have been very cool if HTML had been created with the ability to do client-side includes without having to resort to using a Turing-complete VM in the client to do it.
[delayed]
yeah editing all the footers and navigation parts in html is too annoying to me so i've resorted for my websites to just a back button to a page that has links to everything else lmao
I'm risking being wrong here, but I think the difficulty is getting conforming behavior across all device and browser combos.
Modern web/frontend devs do NOT understand whole page navigating form POST calls from a pre JS era.
I suspect they are both more familiar with client-side rendering, and also thinking of things being able to share components, reuse existing libraries, and so on. So re-implementing everything with vanilla HTML and forms feels like reinventing the wheel to a team used to an SPA component library, it's not that it's intrinsically harder, it's that they don't have the existing building blocks they'd normally reach for.
Modern frameworks such as Astro allow for a similar development experience (and can optionally use JS, React and other client-side libraries) while still being able to generate a static site if desired.
I think server-side rendering and static site generation are less familiar to many web devs who came up in the React/Vue/Svelte era. The patterns and mental model are just different. In an ideal world we combine both: fast static HTML that works everywhere, with progressive enhancement for interactivity. Astro does this well; Next.js supports it too, though the SSR parts have a learning curve.
Because before there was AI Slop, there was React.
I think Facebook with their money and Vercel with their VC funding tried hard to push the React and then the Next.js everywhere. So it arrived in time for AIs to all train on it. And now it’s the one true way :)
But do we really need all that stuff? Build steps, bundling, tree shaking, all for what? And is it really simpler… hmm
There are some advantages, but the main one is probably that it stops everyone from using NoScript and breaking tracking.
Simpler doesn't mean easier. Consider a chef who at their previous job started using a wood-burning stove. This is an objectively simpler tool than a gas or electric stove, yet it would be very difficult (even impossible, depending on local architecture and regulations) for a new kitchen to add one.
I'm interested to hear what architecture and regulations prevent the use of something that is foundational to web develpment and backwards compatible by design? Which also, by the way, comes with the advantage of not incinerating other parts of the restaurant (accessibility, user experience...), forcing expensive countermeasures or total rebuilds of the things destroyed every time you turn it on.
90% of the SPAs I use could be Django/Rails/Flask apps with no noticeable difference, other than that they'd be many times more responsive on slower devices.
"Old" SSR apps are mature, not obsolete. It's ridiculous when they're not considered even when they'd be the right tool for the job. I wouldn't try to clone Google Docs in Django, but something like Linear could 100% be modeled as a CRUD app with new page loads on click without a substantial loss of functionality from the user's POV. Not to mention built-in, automatic support for pervasive "deep linking", aka "linking", because you access a URL's contents by GETting and rendering that URL is just how it works.
I agree but you and I aren't the audience and I think experts in general should be a little more holistic when critiquing other people's choices. For any given problem there always exists a perspective from which your solution is over-engineered. People who (like us, I presume) understand processes, files, the command line, compilers, a computer language or two, a bit about computation theory (e.g. Turing, big-O, Knuth..) can get to a very broad swath of places along many different (often shorter!) paths. This is not where most people are starting from.
Speaking of Knuth, imagine being asked to "write a program to add two numbers" and using something like Python instead of assembler, or because really that's complex too, machine code. Do you think that the amount of housekeeping and computer activity is justified for adding two numbers? Objectively, it is not, its just that steady-state dominates the transient over time.
I don't understand how having to pay 20 different vendors so hackers can run commands on your server barely impeded is somehow simpler.
The message you just wrote involved how many complex systems, from your keyboard switches and firmware to your BIOS and OS interrupts, to your browser, the internet and middle boxes, just to say one sentence to someone. It would be much simpler (and more secure!) if you just told me with your mouth, but you didn't do that.
Of course, and if you use all these services you can be a pedantic ass who never has to actually ship a product.
Computers are very good at repeating a known "recipe". They can add numbers billions of time per second. Yes, billions with a bee.
The hard part is coming up with a recipe that solves your problem and that the machine can run without breaking things when it runs around with a few billion steps per second. You have to think ahead for it and handle edge cases in the recipe.
That is the really hard part.
This takes me back close to 15 years ago: using backend session management in Grails and the html forms that were enhanced with “some” JS, using responsive CSS. The difference at that time was browser tech not being as advanced as now, we had to care for different browsers and deal with IE7 and even IE6, it was difficult and we needed extensive QA (Browserstack would appear later). There is a reason why we had JS library evolution. Dude there was no npm, not even bower. Then we had Backbone.js - loved it, then AngularJS - amazing, then Angular version which had huge breaking changes then React, Polymer etc. Native browsers can do a lot these days, it is easy to enhance the functionality as well. But it was not always the same, the decisions to use React made sense for a variety of reasons at the time, maybe it was the case here as well.
I haven't heard much about in a while, but the HTML Triptych proposal [0] is still something I hope to eventually land in browsers. HTML forms speaking to REST endpoints are a good pattern. (meaning user-aiding validation is handled via the input attributes, real validation is handled on the far side of the request, and the flow is GET /form => POST /thing => GET /thing/1) It would be a great pattern with the triptych features implemented!
[0] https://triptychproject.org/
Most of my apps are now simply HTMX + Go + SQLite.
I've found it's enough for most projects.
One of my sites is image heavy and serves 10 TB of traffic per month. For this, I use the following setup:
1. S3 (I wanted reliable data storage) 2. In front of it, I have Cloudflare (with Tiered Cache enabled, which makes POPs prefer pulling from Cloudflare rather than the origin). I've set rules to cache everything on both the browser and Cloudflare for 1 year, ignore origin cache policies, ignore query strings, etc., and I simply use immutable objects that require revisioning. 3. BunnyCDN in front
Cloudflare will not let you run an image heavy site on its own, so I use this approach to massively cut the bills. Their policy says you cannot use it primarily for images; it must be used for HTML, CSS, JavaScript, and other site content.
And if you run only S3, the bills will be huge.
But yes, lately I’ve been building mobile apps. PWAs are limited; the OS can evict IndexedDB storage, so I cannot offer people reliable data storage in the app without sign up or involving a backend.
What can I do? So I was forced to switch to Flutter on Android, but I ran into another pain point: app updates sometimes spend a lot of time "under review," which is frustrating. For the same app, I maintain a web app that is very quick to update by comparison.
I wonder why there isn't a mobile OS that simply lets you build apps with JavaScript, HTML, and CSS and gives you reliable storage without all this effort.
I like how quickly you can update PWA app.
> I wonder why there isn't a mobile OS that simply lets you build apps with JavaScript, HTML, and CSS and gives you reliable storage without all this effort.
There is! You just have to time travel all the way back to 2009 when webOS was launched by Palm. Time travel is the easy part, you then also need to somehow prevent Palms demise and webOS fading into obscurity as a smartphone OS.
If 2009 is too far back you can try your luck in 2012 with Firefox OS.
Joking aside, people and companies have given it a go. But a combination of bad timing and various other events never made that reality happen in our timeline.
Maybe I'm missing something but aren't PWAs pretty dead-simple on both iOS and Android? Maybe it's the "reliable storage" part that's the gap?
The amount of effort that goes into keeping Termux barely functional, has a lot to do with Android and the platform making it harder and harder, to access a dev environment on a phone.
Running `npm install` on Android isn't so easy.
(Caveat: The new Android Terminal that only works on a handful of models.)
I thought I read that one or both of them removed or heavily restricted PWA support to funnel more apps to their 30%-taking app store.
No that's not correct, they both support PWAs with a large feature set. I built and have been maintaining a client's PWA (an internal tool for their employees) for over 10 years now.
Not my site, but someone on HN (can't remember who) built it: https://pwascore.com
You can still buy KaiOS phones, I have one.
Sure, but they aren't exactly widespread or even close to mainstream.
Go is so awesome for server apps. I should have discovered it much sooner. It somehow sits in the exact optimal point having no bullshit overhead like C, yet also getting out of your way so you can focus on the business logic like Java (not Rust).
It's not great for every task - in particular the lack of abstraction-building capabilities - but it's great for business-logic-heavy server apps. It feels like it's specialized for that and not trying to be a jack of all trades.
>One of my sites is image heavy and serves 10 TB of traffic per month
I can't imagine this kind of traffic without acting as a CDN, advertising broker, pornographer, or part of a massive ecommerce site. I have to wonder, what are you doing that generates 10TB of traffic per month?
Are you me? I've been on a tear building stuff with HTMX + Go + SQLite. It's like the trifecta of boring technology that jives with me. Stuff gets deployed to a colo server using a generic bash script.
I created a couple libraries to abstract the SQL and HTMX/web/OAuth bits; my apps are now very similar and easy to copy features between.
https://github.com/cattlecloud/webtools
https://github.com/cattlecloud/litesql
> I wonder why there isn't a mobile OS that simply lets you build apps with JavaScript, HTML, and CSS and gives you reliable storage without all this effort.
Because there isn't a 30% walled garden you can create with that.
I've learned about several frameworks that are system language + web frontend. The general approach is, ship a small binary compiled for the specific platform, and use the platform's native browser (and whatever html frontend tech you like) for frontend. The whole binary can be 1mb or less.
There's a couple of Rust libraries like this; right now I'm building an app in Tauri+Svelte.
It looks like Wails might be a similar framework in Golang.
Can't vouch for it's effectiveness, but Hotwire Native might meet the needs for mobile paired with plain HTML. Despite the name, it doesn't seem that using Hotwire for the web UI is necessary to use Hotwire Native. I could be wrong about that though.
https://native.hotwired.dev/
Hotwire doesn't work anywhere you can't guarantee a high-quality low-latency connection. Every UI interaction in Hotwire includes that round trip.
That's pretty much true of HTMX too (which is what GP mentions using). I mention Hotwire Native (different from plain Hotwire) because it makes it easy to wrap a web app as a mobile app. Then you can replace performance critical parts with native views, but keep everything else working through the web app portion. This is easier to maintain, albeit with the downside that everything requires network round trips. Depending on you use case, that may be an acceptable tradeoff.
> Cloudflare will not let you run an image heavy site on its own, so I use this approach to massively cut the bills. Their policy says you cannot use it primarily for images; it must be used for HTML, CSS, JavaScript, and other site content.
Pages has a 20k-100k limit on static files, but if they just guide you to R2 to offload it, which is still Cloudflare.
Did you mean the CDN? In which case, I'm not seeing that in the terms. [0] Though, I would have expected they'd have a similar thing. R2 resources don't generally count towards your cache limits.
[0] https://www.cloudflare.com/service-specific-terms-applicatio...
10TB is nothing these days. All Hetzner virtual servers in Europe have 20TB/month traffic included (excess less than $2/TB) and all their dedicated servers have unlimited fair use (which is probably about 200TB/month averaged over many months).
On the mobile point, there is https://github.com/instawork/hyperview
Your Go server can have endpoints that render XML instead of HTML and basically get the same server-driven experience of your HTMX site. Fully skips the need for the app review process since you're not updating the actual client app code to make UI changes.
> Most of my apps are now simply HTMX + Go + SQLite.
Very cool!
> But yes, lately I’ve been building mobile apps. ... What can I do?
I am currently building HTMXNative.
Together with Objective-Smalltalk, which has linguistic support for REST built-in.
The idea is that you create your model in a natural way and then thinly wrap it to deliver wherever.
have you considered Capacitor? it works pretty well for cross platform web development
Since you're using HTMX, I have to ask: do you have any tips or idioms for composing complex forms and UI without things getting out of hand? I love the approach, but I'm having a bad time figuring out where the ideal balance is between too few or too many HTMX-replaced areas in a page. Thanks.
My #1 advice is not creating separate server endpoints for every HTMX fragment, unless you are 1000% sure that endpoint will be used in multiple different pages.
Working on a "simple html page" that is actually 5 different independent "subpages" (routes, views, templates) in the backend is awful. The UX was improved, but the DX was sacrificed.
I recommend having a single view function for each page/SPA and do sub-routing within that function to handle page fragments. In other words, use a GET/path/Header parameter that indicates which fragment is currently needed, defaulting to the full document as normal. Just make sure you are considering request logging and client-side caching in your solution.
This makes it very easy to add/remove async content from the page, since you are just editing the one view function/template and you can easily reason about the entire page as one logical unit.
It also means you don't need to duplicate security logic or other middlewares for the page, since it can be implemented once at the start of your multi-faceted view function.
Generally you don't even need to do the sub-routing in the handler. You can just render the entire page and have `hx-select` attributes pluck out the part that you want.
That is a good solution for reusing content across pages, but most of my HTMX usage is for fetching data that would otherwise delay first page load significantly, or for seamless interactivity. Very different use cases.
This is huge, thank you.
Where do you store the SQLite database files? What is your strategy for partitioning your data into SQLite files? One per user or…?
Developing webapps only for iPhone was Jobs original strategy. Only changed when developers complained.
> Most of my apps are now simply HTMX + Go + SQLite.
Would like to hear about your Go stack for building htmx apps.
This post is good, and it's a great example of taking a problem and solving it with the appropriate tech with the right amount of depth. It really helps to have full domain knowledge of your customers as well.
However, I do not like how it is framed as "simple html is better than react" - because you could just as well have told the same story as a react developer.
(Nb. I could go on forever about the complexities and intricacies of storing things session based on a server vs browser based and etc - and lots of other things that were skimmed over in this article, but that would be too long)
All of those things that are simple in html are also simple in react.
It's literally the same code - there's nothing preventing you from using browser based html validation in react - all the same code that gets complicated in react (overly complicated validation logic) also ends up being complicated in astro - they have their own thing around schema validation etc and integrating it within an astro site means you have to integrate their client router etc etc.. so it's very easy to go overly-complicated there as well.
The comparison is also with an off-shore team doing development for you with probably incomplete knowledge and the way projects are structured they have an incentive to create the solution as fast as possible, in as little time as possible, with the biggest amount of complexity as possible.
The last point is devious - it's not necessarily that the contractor does this by design, but the incentive structure makes it so something that's overly complicated actually benefits them, so they don't have a direct incentive to go with something simple.
Anyways, a simple solution, directly addressing the problem at hand is always better - no matter what stack you pick.
(I'd like to say that I don't have anything against Astro's form validation, I was just trying to highlight how there's more to it than "native html browser validation")
Seems like LLMs embrace that last point as well.
The counterargument: In Defence of the Single Page Application:
https://williamkennedy.ninja/javascript/2022/05/03/in-defenc...
Funny enough. I’m opening this on mobile internet connection and it stuck at loading spinner. I don’t know if the problem is with my internet (probably not) or support for mobile so I can’t even read the content.
I think that's the joke :) all other articles load fine instantly, just this one that has a spinner
Heh it got me then :)
Whoosh! Right over my head. I feel a little silly, but also, got me!
Got me too :-)
My guess it’s a joke.
it's a sarcasm loader
Sufficiently advanced parody is indistinguishable from reality.
Ages ago when writing.com was first modernizing its site, they started by hiding story content to display a spinner, waiting a second or two, then re-displaying the content. It was on the page the whole time, they just made it look like it was loading in the background.
https://medium.com/luminasticity/on-the-triumph-of-satire-fa...
"Satire isn’t dead.
Satire won.
This is what it looks like from inside, looking out. "
Discussed at the time:
In Defence of the Single Page Application - https://news.ycombinator.com/item?id=31275178 - May 2022 (32 comments)
All I got is a "loading" animation. Gave up after 10 seconds. So, not a counterargument, but a confirmation of the article's thesis.
It's a joke/sarcasm
I wish I hadn't read the replies. I love being made the fool (though not to worry, I'll have plenty of other opportunities, likely today!)
I got wooshed.
Nice work!
I have to say that I hate SPAs. It is often a far worse user experience than the vanilla multi-page websites.
Like most defenses of Single Page Applications it managed to make me angry, at least at first.
lol this got me
Excellent article but I am always torn when I read inspirational articles like this - it makes perfect sense to me and I love the idea of simple, non-nonsense sites that work well, load quickly and don't rely on the latest browsers to function.
Then I start to wonder if that's just because I'm not smart enough to understand React or whatever the fancy technology of the day is.
Feels like I have a hard understanding threshold that cannot be breached - give me a simple editor like Sublime and ask me to make a web page - even with JavaScript - and it's my happy place. Give me VSCode or Zed, Claude/Copilot/ChatGPT plugins everywhere, React tutorials and my brain goes to mush.
If it makes you feel any better, the people using the fancy frameworks and whatnot usually aren't smart enough to understand them either.
you might enjoy this post on complexity and why it is bad: https://tengstrand.github.io/blog/2019-09-14-the-origin-of-c...
keeping things simple is not a bad thing, and often requires you to be smart enough not to overcomplicate them.
I love the web. I hate what the React cretins have done to it.
Embrace Extend Extinguish is real, and the people going along with it deserve to be replaced by a LLM that lies and spits out garbage code just like they do but faster.
React has helped folks like vercel/convex/cloudflare to build fantastic dashboards. There's just as many examples of well done React as there are the opposite.
I generally prefer solidJS nowadays, but the react ecosystem has enabled lots of amazing user experiences (and developer experiences too if you don't fall into the trap of overcomplexity).
Sadly, this is the story of nearly every React project I've reluctantly inherited. In my experience, it's because React is not opinionated like its peers Vue and Svelte. So, a bunch of devs will use something for state management and and another team will use something else completely. Eventually both teams leave after making a mess. But, if you look at the graphs and numbers that MBAs chase, they will all look like everyone was productive until the very last minute. The ultimate casualty is unfortunately the user. Even Facebook hasn't figured out React across their properties. Just use Instagram / Facebook on the web. Bunch of spinners to load a static list of items in a drop down menu. Not even joking, click on the bhamburger menu on Instagram web. It makes a dozen requests, shows you a loading skeleton and takes 5 seconds before you can see a finite list of menu items. Ironically Facebook was super popular in the 2014s because they didnt have much React based BS going on. Everything was just good old hyperlinks.
If the creators of React haven't figured it out, what makes you think you can?
> Ironically Facebook was super popular in the 2014s because they didnt have much React based BS going on.
facebook was super popular in 2014s because they were just starting the transition from mostly being about people to mostly being algorithmic rage bait, and they tapped a new market of fox news/conservative talk radio listeners
This isn't "We replaced a React app with an HTML form and performance improved." It's "We replaced a bad web page with a good web page and performance improved."
Attributing this to the technology driving the browser experience is silly. You can make a brilliant user experience with React. You can make a terrible website with plain HTML.
The improvement comes from the change design, not tech.
> The improvement comes from the change design, not tech.
You could argue that the constraints of using HTML-first (as they call it) helped them stay away from the bad patterns they were using before.
But you’re right: The user change came from fixing the design, not the technology used.
This is a lot like those bad resume bullet points where someone tries to claim an increase in business was due to their code change. “Increased visitor count 100% by rewriting website to be HTML-first”. Then when you ask them about that point they concede that the entire site was redesigned to fix some design problems or add a feature and that’s what drove the visitor increase.
Plain old HTML with vanilla JS is not exactly a pit of success but React is much closer to a pit of despair. The former tends to be clunky but effective, while the latter requires a PhD in complexity avoidance to stand a chance.
JavaScript: The Good Parts by Douglas Crockford is a comically short book. React: The Good Parts would be shorter still.
Of course, is just what with React is 100x harder, and when you fail the fans will blame you instead of the technology.
The standard answer to that is that some technologies make one harder than the other. That's kind of true from first principles, but it requires making the case that e.g. React is actually harder to make good than a plain HTML page.
Fun thing, TFA describes a kind of multi-page wizard style form that I haven't seen a lot anymore in the last decade or so. But when I did see it, it's always some dogshit enterprise system. Some Oracle product for expensing expenses last time.
The problem with those things always seems to be that they are slow in the middle of doing your task. Every button is seconds of waiting. Doubly annoying if you have to go back a step or two. The badly coded SPAs seem to be slow at the start. It takes a while to load, but once it's loaded its performance is usually okay.
I don’t like multi-page forms when I’m not able to see at a glance all the info needed to fill out the form. Though I guess if my progress is durably saved that makes it not so bad. The worst is when I fill out a bunch of stuff and then realize it’s asking for info I don’t have, and after I get the info I have to fill everything out again from scratch!
Is it slow though? Like in practice? This demo [1] using Datastar (a streaming HTML framework) every action including scrolling roundtrips to the server. Even the checkboxes changing colour is a roundtrip.
https://checkboxes.andersmurphy.com
The advantage of SPAs, like the checkboxes page, is that they can do the round-trip less visibly. The user can still continue the next thing. So even if it is slow, it's less of a deal than loading and rendering a page anew.
> Is it slow though? Like in practice?
The multi-page wizards? The ones I've seen were. Enterprise crap systems.
I'm personally of the mind that any .gov or utility website should work on old browsers, or at least have a lite fallback to ensure everyone can at least read how to do something the pen and paper way
I built apps like these on GOV.UK over 10 years ago for the Ministry of Justice. We built our own form wizard library that let us validate long forms in steps and break them out into multiple pages because Ruby on Rails didn't support doing that out of the box. It was a very important principle back then that everyone should be able to make use of these digital services regardless of whatever users were using to access them.
I've always liked basic HTML pages where one can upload a document without having to restart the entire application. That's a great practice you have there with general forms. With each session ID, it can cross reference a page in a multi page application with that session ID, so that the user can maybe type it in if necessary, but it should be able to determine that with enough information, like IP address, upload date, browser, OS and so on. But the most accurate session would be within the browser so that the cookies for a single application aren't mixed up with another applicant, like a relative, who might be using the Playstation Portable.
I am actually impressed by the gov.uk websites. Its minimal and just about right for its purpose.
May I ask what tech is used for mobile apps? I am guessing not a full mobile app but it uses webview.
I'm curious how many web issues can be solved by having the people responsible for the relevant sites only be allowed to use them on a Windows 11 machine with only 4 GB of RAM using Firefox with the network throttled to 3G speeds.
Assuming the processor isn't horrible, I can still browse plenty of sites with those specs without much issue, and on the sites that do require more, it's very rarely because the sites actually needs it (i.e. I'm not running Windows XP in a VM in the browser or something). It could just be normal HTML and CSS and normal forms, sprinkled with some light JS to help out a bit. But the amount of sites made with that level of care and attention are sadly rare, since the people responsible rarely feel the pain or have the empathy to fix the problem.
+1 to the idea. Developers should have a powerful machine to develop, but a minimal machine to run and test.
> Windows 11 machine with only 4 GB of RAM using Firefox with the network throttled to 3G speeds
And their mobile device should be the cheapest Android phone from a monthly cell provider you've never heard of. This is what the real world is for a lot of people, and a huge number of developers simply don't know or care.
Don't tell me were going to rediscover progressive enhancement all over again after more than a decade. Back when we used to actually care about the end user whether you were programming frontend or backend.
Too much VC money and big tech influence in the JS ecosystem made the web worse in some ways.
I was going to comment on the Terence Eden excerpt quoted by the author about the woman researching housing benefits on an old PSP browser, when I noticed that you (the OP) are Terence himself. It's strikingly powerful, and a reminder of the duty we have in building our infrastructure.
> Of course, your javascript-based analytics package doesn’t see the users you are bouncing because of javascript failures.
It is frightening to think of how many people are alienated from critical systems every day because of this bias reinforcing the idea that they do not exist.
just use firefox with an adblocker like adnauseam and a fairly decent chunk of the internet stops working, including chase.com and several other massive corp sites.
I can't imagine trying to use links/lynx or a browser with less market share than FF that isn't based on chromium.
Chase works with ad blockers, though.
It doesn't work if you disable JavaScript...but it wasn't always this way!
They had a mobile version of their online banking service at https://m.chase.com that was EXTREMELY FAST and did 85% of what you need to do in an online banking portal (check balances, transfer funds). They scrapped it when they moved to their current bloated monstrosity of the portal that they have today.
It was a big reason why I moved to a credit union (who outsources their online banking services to Alkami, which maintains a very tight portal and supports 2FA AND passkeys!).
i am unclear on if it's because of the adblocker (specifically i use ad nauseam which does block some JS. some.) or because of firefox. I can load it on edge every time it fails on firefox. last week, chase.com worked fine on firefox. the previous 15 months where i needed to log in, it did not.
Someone at chase isn't checking their work on firefox.
FWIW I use Firefox with uBlock Origin and Enhanced Tracking Protection, and use Chase's website almost weekly. No issues that I've noticed on MacOS or Windows
When I have to log into Chase on my computer, I do via Firefox, and it works fine. Maybe you flipped something on in about:config that's breaking it?
i love that i'm holding it wrong. thanks, HN.
And that is what pixel tracking is for. :)
There is one hard wall that stops very old clients from connecting: Not supporting a new enough version of TLS. TLS 1.2 is from 2008, and TLS 1.3 is from 2018. Web browsers older than 2008 can't connect to modern websites since TLS 1.0 and 1.1 were deprecated from web servers in 2021.
I still think it's worth it to provide connections over plain HTTP for this reason. It probably doesn't apply to many people, and you shouldn't allow anything really important to happen over plain HTTP (logins, payment), but normal viewing should still be possible.
Sadly, the internet as a whole seems to disagree. Even the most useful resource on the web one could use over plain HTTP, Wikipedia, only allows connections over HTTPS. I guess it kind of made sense as part of the campaign to push the internet as a whole over to HTTPS, but anyone who's connecting to any website over normal HTTP these days is doing so because they literally can't use HTTPS.
I agree; you should allow non-TLS connections as well as TLS. (At least, access that does not require authentication should not require TLS, but should still allow it if that is what the client wants.)
If you are concerned about accidental login or API keys without TLS, then you can consider supporting mutual TLS, which improves security (and flexibility) in other ways as well. (You do not necessarily, have to require mutual TLS, in case someone prefers to use a username/password login, or 2FA or something else like that instead.) (In the case of login forms, you can have the links to the login forms to always use HTTPS, in order to avoid the problem.)
Having HTTPS as mandatory it more mistake-proof.
I'm not convinced from the article that HTML-first was the thing that fixed the problem. What fixed the problem was 1) the person building it knew what they were doing and 2) it had design constraints from the get-go to be user-friendly. You can do that with React. It's arguable whether it's easier or better, but you can get there regardless of the approach you use.
1. Presumably the team that made the previous app also thought they knew what they were doing. Presumably they were not hired on the assumption that they couldn't make a good app.
2. The design constraints had always existed, the previous app just failed to meet them.
1a. They may have thought they knew what they were doing, but their work product shows otherwise.
1b. They may have thought they knew what they were doing and spoke confidently enough to convince whoever was doing the acquisition, likely non-technical, of the same, but the bad hire and the bad hire’s work product shows that neither was the case.
2. Ideas merely exist. To be constraints, they must be enforced.
Related recent threads about react performance:
JavaScript-heavy approaches are not compatible with long-term performance goals 178 points by luu 3 months ago | 237 comments https://news.ycombinator.com/item?id=47029339 (should be called "Why React is not compatible with long-term performance goals")
Does anybody like React? (jsx.lol) 242 points by brazukadev 15 days ago https://news.ycombinator.com/item?id=48274077
I remember when YouTube implemented a new html player and suddenly, unexpectedly, they had all the is traffic coming from rural Africa. Size matters.
People who built a crappy website using React are just as likely to build a crappy website using Astro, HTML-first approach or any other technology
I disagree. An HTML website which uses links, forms, buttons and inputs will by default:
* Have working back/forward buttons * Have working progress indicator as provided by the browser * Show errors to the user - even if they are ugly * Be accessible to keyboard navigation
With SPAs these are all things the developer has to get right.
So often when using a SPA I'll click a button, you get a spinner and then nothing will happen. Is it still in progress? Don't know. Eventually I'll open developer console and trace the network requests to find the JSON HTTP request that returned "ERR_BAD_EMAIL" and fix what I've entered. With a normal form submission at least the user will see the error message and can press back and then fix it.
"What one man built, another will always find a way to break"
True. Crappy developers will build crappy websites irrespective of the tech.
The article is clearly aimed at non crappy developers or developers who want to do better for their users.
And it provides an anecdotal experience where an HTML first option developed by a good developer was far superior to what a JS necessary option would have been, given the user base of this application.
At least it'll be a fast and crappy website.
It's definitely possible to make slow server-rendered website. Most of the slow client-side apps are slow because they're waiting on slow network requests.
(I still very much support fast, simple HTML websites. The good ones are a fantastic user experience)
but the host (the company) will need to pay the price in the form of server equipment. Not the user as is the case with client side rendering. If server side rendering becomes slow it will affect all users regardless of their hardware or connection, prompting earlier response from management and devteams.
The cost difference between client-side and server-side rendering is pretty non-existent these days.
yeah, I love when shit loads immediately, so I'm not wasting seconds of my life just to see shit.
I agree. Server-rendered React can also send down 100% HTML apps.
But I 100% see where the author's coming from, considering the massive fragmentation of react codebases/patterns and decision paralysis of React development in general. I really doubt most React apps, even the more accessible ones, are testing their multi-page form wizards with JS completely turned off.
HTML-first does seem highly underutilized in the commercial web, and I learnt a lot from reading this (as a solidJS/react dev).
You can do a bad job with any tool but you cannot do a good job with any tool.
This is patently untrue, give a craftsman terrible tools, and they'll still produce a decent end result. That said, defaults matter, and astro is going to be significantly more friendly out-of-the-box to low-end clients
> give a craftsman terrible tools, and they'll still produce a decent end result
This is an absurd statement. Just because something is a proverb, doesn't mean it's automatically true for all cases.
> Just because something is a proverb, doesn't mean it's automatically true for all cases.
A professional woodcarver armed with only a metal spoon will still make a nicer woodcarving than I can given a full wood shop. Similarly, if you only give me notepad.exe, I can still make a pretty nice website.
Does using the right tools make our lives easier? For sure. Using mildly-wrong tools (react, in this case) isn't going to hold us back very much though.
How good are you at hammering nails with a screwdriver? Tools are important. We have a more apropos saying in our industry: use the right tool for the job. Don't let your brain get fried by the 'poor workman' adage, it's talking about the difference between an expensive tool and a cheap tool of the same kind, not two tools of different kinds entirely.
i dont think thats a proverb though? the actual idiom is more like
> A poor workman blames his tools
which is towards the opposite meaning
I find it's way easier to build crappy React apps than an HTML-first approach.
An "old school" Ruby on Rails/Symfony/Django app, with templates, usual get/post forms etc, frames you and pushes you in using the standards and relying on browser default behaviors.
In JS-heavy apps, it's as easy to code normal `button` elements as it is to code clickable `div` elements. But with the divs you just forget to handle keyboard nav, proper element roles, etc. It's easy to create fake links, not relying on `a` tags, using an internal JS router that doesn't expose URLs, doesn't handle middle click mouse, for no particular reasons.
In less JS-heavy contexts, the easiest way to do is to use proper HTML so you are less inclined to mess up.
Even on codebases that use a decent framework like Next.js that handles those for you on paper, it's often we see people not very aware of the benefits of using proper semantics and standard behaviors, and you easily end up with web apps with poor UX in the end.
true, but where can I find the smallest functional react website where react is needed...?
Not really, no. Astro requires you to opt a component in to client-side rendering, React (with its server components etc) require you to opt out. Defaults matter in scenarios like this and I'd bet the average developer of crappy websites would have a much faster site with Astro than React for that reason alone.
Small correction: it's other frameworks that require you to opt-out, most notably Next.js. Although I've been seeing so many people confusing Next.js with React lately...
Astro itself works just fine with React, and it can still be HTML-only.
But you can also render React on the server yourself using renderToString, if you don't want a framework.
with this logic, why discuss any technology?
We have the "social media" companies to thank for this innovation. They introduced these stupid patterns primarily to keep users locked into their walled gardens and the rest of the industry followed them to the bad place like lemmings. Software is a fashion industry.
They replaced an ASP form on the website with something better.
To apply for their services, customers could either use an old ASP form on the website, or follow a manual process
"customers could either use an old ASP form on the website"
Did I miss it, or did the article neglect to explain why this was a problem?
You ever tried to maintain or enhance an ASP form that was built 20 years ago and enhanced by random devs with no documentation? I was working with ASP forms 15 years ago that were hopelessly outdated, included tons of technical debt and were not in the least bit user friendly. You can accomplish so much more with things that ship in every browser now. It’s not a very modern technology and you miss out on things that are now very standardized and much, much easier and quicker to develop and maintain.
Also, Microsoft.
That this article even has to be written proves how stupid web development as a whole has become.
I loathe this area of software development, loads of unnecessary frameworks and spinning gears that add complexity while making less and less functional applications
I am not familiar with this astro framework they used. But having built some sites using Pure HTML/JS back in the day, React, Angular, Vue, Rails ERB, Rails Hotwire, and HTMX. I think HTML first websites are absolutely the way to go. Rails Hotwire with View Components makes rails sites super fast, faster to develop and easy to re-use components. HTMX more generally, but Ive used it with Spring boot and Thymeleaf. I really don't want to go back to SPAs. Development time is less and the website performance is better, and I haven't really seen any regressions in capability. With HTMX and some url parameters, I can make a pure HTML site that seems like a Single Page Application but without the excessive loading times.
I first tested Astro on my site and never went back. Now every new project defaults to Astro and I have to have a reason not to use it. So far no reasons. It's simple, fast and it kinda fits my desire to keep things minimal. For example, yes, page content matters, but all but one page on my site is under 10kb, most hovering in the 3-4kb range (100% of the downloadable content)
Yep Hotwire and Hotwire Native are amazingly useful tech.
Good post, but:
> A venerable web application pattern that has had a small modern renaissance thanks to Remix
Remix is not that popular. I don't think attributing this to remix is accurate. Next.js quite possibly.
The full context of that quote makes it clear that it's meant more as a wry joke:
> A venerable web application pattern that has had a small modern renaissance thanks to Remix, form submissions and redirects took a while to explain to my colleagues, on account of everyone being used to heavily client-side web applications.
(Although it's not really a joke, it's pretty amazing how many professional web developers these days don't know how to use forms without JavaScript.)
The opposite is why I'd never be a good web developer. I grew up messing around with PHP and if I spent the time to learn the modern stack, I'd constantly be thinking it's stupid.
I can relate to that.
I recently had to intervene during the latest office holy war to explain that you don't need JS for file uploads.
It was eye opening.
I think the author is suggesting that Remix was the inspiration for the renaissance, not that it's necessarily the most popular method for doing so.
I'd be curious to see the stats on how often Next.js users lean into the server component model that makes the frontend fast. My anecdotal experience is that it's an afterthought for many. By comparison, Astro (as mentioned by the author) makes you think about this stuff upfront via opt-in rather than opt-out. It's a wonderful framework.
Opt-in = action is required to opt in = off by default.
I think Remix brought back interest in Form Actions and other meta frameworks took inspiration from that.
Remix has been nonetheless influential in the space, in the same way preact and signals have been.
HTML first, minimal JS, not using any frameworks unless absolutely justified is already part of my prompts to LLMs when I build stuff for myself. I’m not a frontend developer, and have not kept up to date with the ton of frameworks. That HTMl first approach already makes the LLM outputs faster/fresher for me. I wish more people went this route for public facing services.
> this was a regulated monopoly, and if their customer satisfaction dropped below 96% (if I remember correctly) it could result in millions of pounds in fines.
OK, I'm still at the beginning and irrelevant to the article, but as a USA-ian, I am so jealous about that. Unheard of here.
That hit me, too, specifically thinking about my current gas/electricity provider. I have not heard one single piece of positive feedback from the public, and there's only ever problems. I feel like that's a pretty universal experience here. Even outside the scope of websites, it holds so very true.
Personal anecdote: Recently they were updating everyone to "smart meters" on the gas lines. They needed me to be home so they could enter my apartment and bleed the gas out of the line by turning on the stove prior to replacing the meter. I played phone tag with them for 6 months, setting up countless appointments, and nobody ever showed up, the meter remains un-upgraded. At the same time, I have received weekly phone calls and monthly physical letters stating that if I don't upgrade the meter, my gas will be shut off. I just moved, so the new tenant will have to deal with it now.
I moved to the UK in 2016.
The public sector, simple, no frills, accessible, no flashy graphics, websites were a massive eye-opener.
They just worked. They had a job. They did it. I wasn't going to buy more from them because of it, and they didn't care. It was great.
I've heard that recently they've dismantled the centralised team that wrote all the rules, enforced it, and started moving to decentralised hosting, but so far the whole still seems to hold to together really well. I think, I hope, they have embedded the expectation that the local council, the tax office, your visa status, etc, should just be utilitarian in nature, and work for everyone.
I worry how long it will last...
Hmm I cannot load the font on Firefox 151.0.3 Arch Linux. All I see is only a title and empty paragraphs. So I end up reading the article in the source code mode and felt pretty on-brand.
Back to HN comments it looks like this wasn't actually intentional?
> Arch Linux
You've only got yourself to blame, there.
The quote from Terence Eden almost made me cry. Actually the whole article did.
This is such a great story. I am glad more people are sharing stories like this. I hope my article the other day inspired more to develop lightweight websites:
https://inavoyage.blogspot.com/2026/06/im-building-parallel-... https://inavoyage.blogspot.com/2026/06/how-about-new-java-ba...
What do people like for form validation?
In this article he recommends the “validation-enhancer” library:
https://www.npmjs.com/package/validation-enhancer
I’ve also seen one called “formisch” that the author of valibot is working on:
https://github.com/open-circle/formisch
They’re both pretty new. Has anyone tried them?
I actually like the built-in stuff and the Constraint Validation API. I wrote about using it with htmx: https://dev.to/yawaramin/handling-form-errors-in-htmx-3ncg
My biggest tip to reduce complexity of data validation if you are using React is to stop using React controlled components and switch to React uncontrolled components. They are an underused part of React. You usually don’t need React to handle every single keypress and every single character being entered by the user. In fact before React popularized it, it was unusual for form components to update on each key press; traditional desktop apps tend to validate when a field loses focus only, not on each key press. This has at least three benefits: (a) good for performance, (b) reduces unnecessary error UI when the user is in the middle of entering data, and (c) simplifies your own code by not having to deal with prefixes of valid input that’s not itself valid.
Also allows user scripts to interact with the forms, eg I can run a bookmarklet to fill out certain forms. With React controlled components all these changes are wiped out and reset with the state that React has in its app memory.
> When we launched, the number of people completing the form doubled. The analytics people didn’t even know where these users were coming from. Of course, your javascript-based analytics package doesn’t see the users you are bouncing because of javascript failures.
Yeah, reminds me of the b52 story re holes in wings on the planes that made it back from missions, leading them down the wrong path of strengthening wings. They weren't looking at the planes that never came back with holes in the fuel silages.
Certainly not the b52, which arrived after the end of WWII…
How many people are using browsers which don't support Javascript in 2026, and doing so out of necessity rather than out of choice? I can't imagine this number to be >1%.
How many such devices can still support modern TLS certificates anyway? By this logic, shouldn't we also use plain HTTP instead of TLS?
It may only be 1%, but that small fraction of users are also probably the people who sure as hell don't need even one more tiny thing going wrong in their life.
If you're using a decade old phone to sign up for a utility, you've got bigger problems in your life and no self-respecting person should be adding to them.
> By this logic, shouldn't we also use plain HTTP instead of TLS?
Better would be to use plain HTTP in addition to TLS, rather than instead of TLS. TLS does have benefits, but if it is optional then it can also be used on computers without TLS (as well as potentially other situations where you do not want TLS or where it is not useful).
I like JavaScript-light websites just as much as anyone (my own website works on the same principle). However, I do wonder how much of the increased traffic has to do with AI agents that now have an easier time working with the more standard web forms. My own contact form had a bunch of bots quoting Scorpion lyrics before I added a rate limiter.
None of the increased traffic was from AI agents. FTA: 'The results? When we launched, the number of people completing the form doubled.'
All completions were real people. It's a government website.
> form submissions and redirects took a while to explain to my colleagues
This is so sad
> My client was a utility company
> Adding a lot of pressure, this was a regulated monopoly
> Some requirements I derived: > ... > We had to meet WCAG accessibility (the team settled on AA rather than AAA)
The author doesn't doxx their employer by giving any dates, but if we take the story at face value and assume it took place in last few years, it is pretty shocking.
How does it take a single hero to be fighting for AA compliance as an afterthought for a project with this scope in the 2020s?
I've worked on much more niche projects that treated this with the respect it deserves as a quasi-legal requirement.
As a public policy specialist (and a tinkerer when it comes to tech), I read the article with pure joy.
This is the "wonder" that happens when you build for the client instead of for yourself. It takes being truly intentional: sometimes the client (especially that kind of client) does not know what he needs or want; sometimes you as a dev simply build something with good intentions but just out of habit.
Taking the time to ask true, relevant questions, and build backwards, is one of the ways you achieve this. It takes time, energy, intentionality (once again)... but it's worth it.
Everyone pays lip service to 'user experience' and 'putting the customer first', but, only with experiences such as acing customer service forms can one really claim to know what 'putting the customer first' means.
Sometimes it is just a form that you need, with the web page loading a new one after 'submit' has been clicked. Yes we can do fancy things to ajax-submit the data, but did the customer want that?
Getting to the form in its 'final form' takes work, particularly if asking the customer for more information, such as proof of purchase or important documents. Do you just start the conversation with a 'contact us' box to have customer service ask for the bits they need later, or do you ask the customer for all information up front, reducing the need for back and fore?
You need to actively test what your customers will do, with metrics such as time filling in the form and how big the ticket queue is.
There can also be internal problems that prevent getting the form right. For example, if there is some manager in charge of customer service that is an empire builder, with a vast team. If your form decimates the team because everyone is efficient and able to go home at the end of the day with an empty queue, then you have undermined the empire builder, so he will want you gone!
There are some huge benefits to getting customer service right. You can brazenly have phone numbers, email addresses and even physical addresses, all published on the website. If the forms work then the phone never rings, the inbox is empty. And nobody can accuse the company of hiding behind a wall of corporate website small print.
The word of mouth aspect is also not to be overlooked. You can harvest reviews from happy customers that should have been unhappy under the old system. If you fix their problem in an hour, or get a replacement product to them the next day, then they will write you a rave review, with that being great for the customer because they explain better than you can how dedicated your customer service team is.
I use the word customer lightly here, there is the term 'service user' that is used in the public/third sectors, but that doesn't sound good in front of the 'service user', probably because they have an actual name.
Getting to the form 'in the final form' means quite a few small changes that can be easily reverted and monitored. It could be just making an input box only show capital letters, or show a numeric rather than standard keyboard.
Ideally, a submitted form does something when completed to place the ball in either the customer service court or that of the customer. If the customer need to provide some information before anyone need look at the ticket, the form needs to send out that email, then park the ticket awaiting whatever the customer does next. There should be no need for someone on the team to do that step.
I know AI does everything awesomely under all circumstances, but the 3-6 month journey needed to deeply understand the customer, the product and the team is something that needs a human, simply because you are dealing with humans and their emotions.
What has proven to be a huge bonus is CSS grid styling. Inputs and labels can be written without the div and span cruft, with everything lining up nicely with a few align 'center' CSS things.
What a fun time to be doing forms that actually work!
Old people. They exist.
Not even that old. 60 year people can't user your fancy site because then don't have an internal model of how a computer works.
You know that when pressing a button a hidden engine runs in the backend (or something runs in the backend). You expect an answer and if the expectation do not match the result, the model in your mind creates an hypothesis about what maybe happened and iterate from there. Maybe you should have clicked something before? Maybe you should mark some form checkbox?
Old people don't have that because they didn't grow up with computers.
What is on the screen is what they see. I clicked next and nothing happens. Well... the site is broken.
You known when you plug your refrigerator and nothing happens and instead of reflecting on the possible blown out resistor that you can bypass with a small wire you understand that your only relationship with the refrigerator is plug and unplug or call for help? That is an old person using your site. They won't fight against it. They'll give up immediately.
I would argue it's not even old people. Most people do not have any understanding of what's going on when you click a button. Website either acts as expected, or it doesn't "work"
If the button doesn't work, the average user is going to say "this most be broken" and then use a competitor (or contact your support). That's why it's really important to error-proof one's design (eg https://en.wikipedia.org/wiki/Poka-yoke).
So instead of the button failing because you didn't check a box, pop up with a message telling them "Please click $box before continuing". Or if you want to be fancy, feed them whatever form you're giving them piecemeal, so that they can't continue until they finish this small part (e.g., have them input a name, then the next page only has a spot for an address, then the next page only has a spot for card information, then the next only has a spot to select shipping). Simple bite sized chunks anyone (well, anyone you would ethically want to sell to) can understand.
Yep, I changed town a few years ago and my new social circles are mostly like that.
Either it works right away without any further questions, or they'll not do it.
Sadly, also if they can't do it on their phone, they will not do it. It's actually very hard to get people motivated to do anything that has to do with sitting down at an actual computer anymore. Which is a bit hard if you're in a very technical political advocacy group, kinda makes me the guy to do everything remotely complex... XD
Yeah, it's really, really not an age issue. If there's an age distinction, it's the range of people who were brought up on computers before the UI was really polished, but even then it's not consistent, and they may not have deep understanding. Kids brought up on iPads, and who aren't forced to learn by their interests or educators, have no clue.
> Old people don't have that because they
Aren't insane.
When did the industry put the onus on the user to understand how the computer works? What happened to the old days of Xerox PARC's HCI studies putting the user first? The computer is in service of the user, not the other way around!
If I need to build a mental turing machine to understand your application, it is a bad application. It is rather the engineer's job to build a mental model of the user and their needs, and if you can't do that you should not call yourself a software engineer.
As a firmware engineer, my philosophy is this: if I'm doing my job properly, the user should never even know I exist.
Maybe this isn't applicable to all software devs. If you make web apps, users actually see your UI, they click an icon or type in a URL and hit enter with the intent of using the thing you made. With firmware, that's not how it works.
When you hit the "mute" button on your laptop keyboard, it should just do it. The audio should turn off and the little LED should light up. If that fails, even once, the mirage is broken. The user is forced to think about the fallibility of firmware, which is a word they might not even know, and still struggle to conceptualize if they do. I think it also has a lasting effect on the way someone thinks about the pruduct: Is this going to work today? Why did that happen? Was that a virus? So on.
OTA firmware updates have the same problem. Most users don't know what the hell firmware is. All they know is their computer is showing a loading screen they've never seen before. It's unfamiliar and weird.
Like I said, I don't know if this mindset translates perfectly to other fields, but the priorities that fall out of my philosophy certainly apply. Reliability over everything, and get it right the first time.
> my philosophy is this: if I'm doing my job properly, the user should never even know I exist.
This is my philosophy too working in infrastructure. It’s my job to care way too much so downstream users don’t have to.
> When did the industry put the onus on the user to understand how the computer works?
Just after the turn of the century, that was when there was a big shift in the industry to move all user interface design work to artists, and away from original user interface specialists. The whole profession of user interface designer (one who understood "form follows function") ceased to exist in the early 2000s, as all work was assigned to graphic designers (who only understand aesthetics).
This movement can be seen clearly in the evolution of widget toolkits for desktop applications, which peaked in usability in the late-90s/early-2000s, and have been losing usability while getting prettier with every iteration since.
Automation generally goes along with a transition to more "self-service" approaches that require the user to model internal states and workflows of whatever they're dealing with.
This is even true for things as seemingly non-technological as getting to your flight once you arrive at the airport. People who are used to dealing with a service desk might just show up with their printed ticket without even having looked at it, take it to the counter, and expect instructions on what to do next without having read or considered all the fields present on the ticket.
It's not just about understanding the technology, but sometimes about understanding the business, policies, whatever. When a human agent or customer service worker is handling that stuff for you (typical in the pre-computer age), you barely have to think about that stuff and even if you're told, it can be "in one ear, out the other". Automation very often means pushing a requirement of more understanding onto customers/users.
Usability was thrown to the wolves in favor of more readily available designers from non-UI backgrounds, brand identity (“UI as branding”), and pretty screenshots for slideshows and marketing.
The pendulum is overdue for swinging back the other way, but I don’t know who or what has both the capability and will to give it the push needed to send it on its way back.
> 60 year people can't user your fancy site because then don't have an internal model of how a computer works.
I think this is a bit outdated. I'll be 60 in a month, and have been practicing and writing about machine learning, for money, for a straight 10 years now; and I was a young man (and a full stack developer) during the digital revolution.
If anything, GenX had to work harder to get into these brittle emerging technologies and paradigms. There's no-one of my age group, at least that I know of, who is remotely as tech-illiterate as your comment depicts.
Truth is that it took so long for smartphones to dumb down everyone's tech acumen that those of my generation had already learned to do it the hard way.
Some people know how to fix a fridge.
> practicing and writing about machine learning ... full stack developer ... digital revolution.
my mum, a boomer now in her 70s, would have no bloody clue what you're talking about. she used to work helping out a guy who was doing punchcard programming back when she was young. she ain't dumb. if i broke it down into normal human english words, she'd probably get a sort of idea (or at least nod along to humour me).
i've lost count of the number of conversations i've had with my dad, late 70s boomer, where he complains that they've changed the UI. "It's all different and i don't understand, why did they have to change it? I don't know where anything is now." he's been moaning about things like this for over a decade now (so since his late 60s).
there are definitely technically not-very-literate 60 year olds and the general point about older folks, whether that's >60 or >70, is very real:
older people exist who don't have a clue about SPAs/PWAs, and chances are they're either asking their offspring for help (my mum does this), trying to phone someone instead (my mum does this) or just walking away from it (my mum does this).
To be fair, the UI randomly changing in every update is a pain point for people of any age.
GP clearly isn’t talking about 60 year olds who were full stack devs and get on Hacker News.
60 year olds have been using computers most of their working life. Word processors and spreadsheets having been ubiquitous for office workers from at least the early 90s.
> 60 year olds have been using computers most of their working life.
Absolutely. I am in full-time work, and expect to be for another decade. I have worked my entire career in IT, doing tech support, training, systems design and implementation, tech journalism, and tech writing (i.e. documentation).
I will be 60 in less than 18 months.
> Word processors and spreadsheets having been ubiquitous for office workers from at least the early 90s.
You did say "at least", but still... longer than that.
I started work in 1988 and they were already ubiquitous in my world. Richer companies had the fairly newfangled IBM compatibles, which were still big and expensive. The cheap Amstrad PCs were just starting to appear.
Older hands had multiuser boxes with SCO Xenix or DR Concurrent CP/M or Concurrent DOS and a bunch of dumb terminals. My company had switched to these from Alpha Micro systems running AMOS -- and again, dumb terminals. One of my clients had a DEC PDP-11.
The real old hands had 8-bit kit: some CP/M, and a few BBC Micros.
The first big migrations I saw were from standalone (or multiuser) PCs to LANs, and from pre-PC systems to PCs and Macs.
That would be "60 year olds who have been office workers most of their working life"
60-year-olds who worked blue-collar for a significant part of their life, this is not so obviously true for.
Also probably not true for 60-year-olds who worked in other non-office jobs, like acting or sports.
There have been a variety of well-paid jobs that didn't use computers that a 60-year-old might have done over their life, meaning that this can't even be broken up along class/income lines.
Most people under 60 aren’t full stack devs either.
Well, neither am I - I am talking about my own peer group, non-tech types just fine with computers.
I'm 55 and likely have been using computers longer than that poster has been alive. Regardless of the fact that I started young, by the time I was in college the PC revolution was in full swing and everyone had and worked with computers.
My mother, born in 1934, had no problem using computers. She didn't internalize how they work, but she learned the workflows she needed. How to launch applications and so on.
The situation described in that comment is just a broken app, it has nothing to do with the age or the understanding of the user.
Dude, you know what he meant. Don't be the internet pedant. No need to be the protector of your class, especially one so inconsequential as "literal 60 year olds".
Instead of "old person" he put a number on it. (Cue the people who need to cut me down because I used a male pronoun for an unknown poster)
You are just describing a broken site?
I've seen static sites with these same problems, 404 was invented decades before React...
Yeah. And reflect on the fact that you know that a 404 error is not a form error. Old people won't. Heck then won't understand that a chrome error page is not your site.
I think he means the submit button is disabled because some field failed validation. It might have turned red or have a message appeared next to it - and it might be scrolled off the screen so you have no idea. Or just the common "I have read and agree to ..." checkbox you didn't tick. Possibly because you didn't read the T&C or because it won't let you tick it until you click the link to the T&C. Nobody else read them either but they've learned through trial and error that when the checkbox doesn't work, you have to click the link to fool the computer into thinking you read it.
There's a great book that explains how abstract computer systems have become from the machinery: https://dabacon.org/caelifera/2017/02/05/book-overcomplicate...
"Old people don't have that because they didn't grow up with computers."
You know, it's time to stop this trope.
People who are 60 today were born in 1966, they probably entered the workforce in the mid 80's. They probably are not even retired yet. They know how to use computers, they own a smartphone (or if they don't, it's probably for economic reasons unrelated to their age).
As a founder and product manager, this kind of thinking is unhelpful as we design the future. In many ways it's actually ageist to imply that old people are unable to utilize everyday technology.
I was building public service websites (BBC News website) back in the early 2000's where accessibility was a real and important consideration. Technology progresses, and the bar for accessibility has moved up.
My father is about to turn 80 - he checks his heart with his Apple watch, video calls his grandson from his iPad, and asks ChatGPT questions from his iPhone and MacBook Pro. Maybe he's more unusual for 80yo's but it's time to stop this lazy trope that old people are technically illiterate.
(also, shit, I'm only 15 years away from being 60 myself :/ )
I think the relationship people are seeing is "80 year old can't fill out my form, must be because they can't use a computer." But international surveys like PIAAC [0] indicate that adaptive problem solving is the real problem. 30 percent of adults are at or below Level 1; I've copied and pasted what that level represents below:
> Adults at this level are able to understand simple problems and develop and implement solutions to solve them. Problems contain a limited number of elements and little to no irrelevant information. Solutions at this level are simple and consist of a limited number of steps. Problems are embedded in a context that includes one or two sources of information and presents a single, explicitly defined goal.
This test is administered on a tablet, so I think scores can be interpreted as a sort of combo of computer use and problem solving.
A full 40% of adults from 55-64 are at that level or lower. Wikipedia thinks that a novel online form would be a Level 2 task, provided it involves navigating across more than 1 page. Based on that framework and assuming your dad can use the sort function in his email, he represents a top 20th percentile adult for the 55-64 age group. It's probably even higher considering his age as the trend is towards older groups having a rougher time with the survey.
I also think it's ageist to assume that older people can't use a computer. But assessments like this indicate that a full quarter of the US regardless of age would have trouble with some of the basic tasks we associate with computer use. So designers should consider their intended audience when deciding what's simple or not simple enough.
[0]: https://nces.ed.gov/surveys/piaac/2023/national_results.asp
This describes all non-technical users, not just "old people."
Yes, but even non-technical users have a different relationship that formed over the young (neural plasticity) days.
"I click the <next> button on this form, nothing happens. Given that I have an internalized notion of forms as a multistep flow, maybe it couldn't advance because something is missing on the current step? Maybe it will give me those messages below or above the form? Maybe some read message somewhere? I'll search for those and try again"
vs
"I click the <next> button on this form, nothing happens. Not sure what would happen next. Maybe the thing will tell me? I can't see no dialogs, no messages, no error screen. Nothing changed. Maybe if I press it again? Nothing? Hum, maybe is disconnected from the network? It's the wifi again? Maybe if I power cycle the wifi router it will reconnect? I click next again now I get a clear error message: no internet connection. I suppose my internet is broken again. Will call my tech-savvy friend."
And now also a lot of young people. They grew up with iPhones and many think that "Wi-Fi" = "internet".
Keep it simple and light. HTML+CSS first, JS to expand functionality. Don't re-invent the wheel.
Giving up is a wise choice: there are so many other sites to interact with. On the other side they have only one refrigerator.
I wish I could convince my grandmother of this.
"Why did the bank change the layout? I want the old one back!" - Don't like it? Change bank then. That's what I did.
I get that changing to another bank is a big unknown, but it's probably still worth it to show your displeasure. Plus her bank are morons when it comes to several other things.
The problem with changing banks to search for a ux you like is that it's not easy to see the ux before you've invested in signing up.
My main bank changed their UX not too long ago, and I liked the old one better, but the banks I've signed up for since are even worse. I signed up with them for other reasons though, so I put up with them because it's worth the pain. It does make my main bank look better though --- mobile style on desktop is annoying, but at least I can easily find everything I need... And it's not really their fault there are seven different options on the transfers page (including my favorite: same day transfer vs next day transfers... at one point same day transfers had a fee but they don't anymore so from a user perspective, it's the same thing but you can have it slower if you want...)
Also, signing up for a new bank these days is an exercise in KYC frustration. And then you can't actually transfer your money and use it, because banks responded to the dumbass check fraud that was being promoted on social media by limiting new accounts.
> The problem with changing banks to search for a ux you like is that it's not easy to see the ux before you've invested in signing up.
I agree. Better to deal with the devil you know rather than potentially one you don't.
Thankfully, I switched to a bank with a UI that was known good going by all the chatter I'd heard, but that's not really something you can guarantee to know. And even if you do know, if the rate is drastically worse at that bank than any of the others', then that's kind of moot.
Yeah. We grown being trained to solve those small puzzles that are websites and apps, so we learned _how they are projected_, not how they work.
I mean, we learn that a enroll is normally a flow. Flows have steps. So if you came to the end of the flow and the finish button is gray, you think.
Hum… I'm used to flows. This is a multistep flow. Flows normally need me to fulfill some small checks and won't let me proceed between steps if something is missing. But some won't. Maybe this is one of those? Some flows have warnings in the end, some have next to the thing missing. I don't see any warn in the last screen, so I'll go back every step and check field by field for errors. That'll probably do.
This is the model you have in your mind, of how a website or an app works.
People that came to computers, apps and websites later in life didn't learned the puzzles.
Next time I have trouble checking in on an airline site I’ll remember that there are so many other sites to interact with that whatever I was trying to do probably doesn’t matter.
I wouldn’t sweat the broken fridge either though, there’s so many other electrical appliances in the house to use.
Is entirely context dependent. I can agree in some scenarios but when it’s a utility or gov site that I can’t really avoid it’s less straightforward.
> Old people don't have that because they didn't grow up with computers.
You realize that someone who was 18 when the Mac was first released would be 60 now?
Old people. They exist.
It's not about age. It's about ability.
A lot of the people I build web pages for are poorly educated. The text we use for web is written for people with an eighth-grade education. Print material is fifth-grade.
People in the SV bubble can't imagine that there are tens of millions of people in America who cannot understand how an SPA works.
These people are invisible, even as they ladle out the food in the Google cafeteria, and polish the chrome in the Meta lobby.
I'm going to link my favorite survey [0] to illustrate this point. Its called the PIAAC, and it's a internationally standardized assessment of adult skills in literacy, numeracy, and problem solving.
28% of the US is at Level 1 or below. That level roughly corresponds to reading a simple text and being able to answer a question about it that requires limited inference (i.e. if the question is "what color are pigs," the answer would be in the text as "pigs are pink in color.")
[0]: https://nces.ed.gov/surveys/piaac/2023/national_results.asp
Empathy and respect for users is what product managers should be doing.
Shipping tens of megabytes per web page is impolite, if not outright disrespectful to users.
> Shipping tens of megabytes per web page is impolite, if not outright disrespectful to users.
You're being generous with what I would consider negligence.
"If our users can't afford the bits, we don't need them!"
They don't know what a megabyte is
They feel the slowness of the page load
Not on their iPhones operating over 5G or the corporate WiFi.
It's still present. JSON/JS parsing still has a delay. And in either case (as the author states) not everyone is using an iPhone over 5G. Heavy React apps are a miserable experience on low end Android phones, even when the connection is fast. I've seen JS/JSON parsing times in the multiple seconds.
There's 5 bars 5G and there's one bar 5G anyway... Citing connection types really is completely beside the point.
My old iPhone handles well react apps, but frequently freeze/crash on heavy advertised pages and pages with huge images/auto loading videos.
You don't think there's any palpable difference as long as the connection is any good?
I think there’s a palpable difference but many young developers have no concept of why.
Salesforce and SAP are not fast, even on that. But ubiquitous for building corporate platforms for their customers.
The vast majority of global users are not using iPhones.
Yes, that was kind of the point of my comment. Apparently not well made.
Read the article. Typical users had old browsers often with poor reception. One user was using a PlayStation Portable which had very limited WWW capability.
The person you are replying to is saying the PMs are using new phones on WiFi, not that the customers are.
Thank you.
"What, support Safari? Isn't that, like, less than 20%? And its standards support is abysmal! No, not worth my time, they can upgrade to a normal browser like everyone else."
But if they dont, where is the disrespect? They dont know what a megabyte is, they dont feel a slow page load. Where is the disrespect?
React is too heavy weight for a lot of things. But it's ridiculous to call it disrespectful.
If Rick Rubin could take a tape to his car to listen to his mixes, your product people can try their websites on £20 phones from Tesco. They can ask to sit in on user tests with minority groups. Extending your knowledge like this is trivial, but rarely done.
May i ask why, specifically, Rick Rubin? I don't know who that is, but whenever we finished mastering a new song, we had a series of "systems" we listened to it on. We went out to my dad's work van and listened there. We called up our friend with a street-comp sound system in his car, and listened in there (neighbors must have loved us!), and then a "cheap" boombox with large-ish speakers but cheap.
if it sounded "clean" on all 3, without the bass muffling everything, and the highs not hurting the eardrums, we called it "good" and released.
Working in the music industry and not being aware who Rick Rubin is… is a bit weird.
i don't work in the industry, sorry. We just made music and released mp3s, 1997-2008. Co-creator of Def Jam, alright.
I wonder if David Lynch watches his stuff on a tiny screen just to make sure everyone has a good experience.
hint: no, he thinks small screens are stupid.
ETA: after like 3 years of mastering and reviewing this way i trusted my ears and my studio monitors enough to know what it would sound like. I also wrote in headphones and mastered on speakers, then remastered in monitor headphones. Anyhow, i think the whole point i was making is "yes, this is a good thing to do, for music, for websites, for software, etc"
Yeah but for those not using £20 phones from Tesco, where is the disrespect?
The consequences of MBs of JavaScript can be perceived by anyone in terms of performance and mobile data consumption.
+1.
I have been asked by someone in late 40s why uploading a video takes a lot longer than uploading a photo.
They are not dumb people. They just do not know.
The onus is on the engineers to design for them.
"Does it take longer to upload 10,000 photos than to upload 1 photo?"
If a 40-year-old can not answer that question, then they are in fact - dumb.
This is a matter of perspectives here. Some of my friends are absolutely brilliant lawyers and they trump me in their reasoning abilities (while I am a typ. HN member, high-tech engineer). Yet, they would not know some tech basics. I see lawyers struggle with formatting in Microsoft Word all the time :-), for one.
A brilliant physicist friend likewise once told me he is clueless how real numbers are handled in computers (not talking about floating point encoding specifics but conceptually itself). Yes, I can say that feels dumb, but I cannot deny that he is a brilliant physicist.
You're not a good and modern engineer who knows his craft if you aren't defaulting to react and tailwind.
And don't dare to contradict me, the fact that MIT-bred leetcode ninjas paid half a million per year can't produce a simple (mostly static) website on that stack it's only because of management that wants to ship the next product. /s
The landing page for our latest tool is a static website (exported from nextjs). It has no marketing on it. It just has simple instructions to copy a prompt into ChatGPT / Claude.
That's it. I will report back if it works.
I think it is a good idea. JavaScripts and CSS can enhance it but should not be required.
Accessibility functions are also helpful; if designed well then it would be helpful for everyone (not only if you are blind or other disabilities), in many possible circumstances.
Someone should send this doc to papajohns.com, which has overcomplicated their website to the point it doesn't even work in Firefox anymore.
Can anyone confirm if Web Components work in the old psp web browser?
Yeah the uk gov website is indeed petty damn good. For once all the money dumped into studies and what not produced an outcome.
I've tried this before and I think this constraint is something that has to be kept top of mind for a designer not just the engineer. Most designs these days assume a single page app and there are interaction patterns that make plain HTML not suitable. But if you incorporate this in from the start and stick with it, there's no reason you can't do this.
There are truly an amazing amount of red flags in this otherwise well meaning and short post.
Designing for failure modes (bad network, old devices, no JS) often leads to better systems even in the happy path. This is a good case study of that.
My go-to for spinning up a site has been Jekyll + Bootstrap with the occasional bit of React for well over 10 years now.
While it still does the job, I'm a little curious to explore more modern options, if for nothing else to understand the choices a more junior dev would face/make today.
I'm seriously considering giving Atro a go. Is it worth it?
Yes
Guess I should have stated the questions as "what makes Astro worth while to try out?" :)
I've started getting traffic on my website only after I re-build it with a locally-brewed MD parsing engine that uses Astro to spit out the final version of the site.
I guess the main argument is how easy it is for an LLM to ingest the content, since I can bet all of the crawlers are llm-enabled one way or another.
> […] it always worked even without javascript. He was appalled and said, “but that’s a lot more work for us.”
Is it more work?
I was a little confused by "doubled our users" since that's more about inbound traffic than site experience. I guess it's really shorthand for "halved form abandonment" which is still pretty great.
Users visited the site and couldn't even begin the form, nor get seen as a visitor, due to javascript metrics and rendering failing.
I think that's even more significant, since it's measuring people who cared enough to click the form in the first place, which is juicier than just page loads.
Doubling conversions, in digital marketing-speak. Marketing firms probably give out fat bonuses for this kind of result.
may be he meant, doubled our users who actually submitted the form
> Javascript and modern CSS should be used to enhance the experience
When messing around with my blog's Javascript, this mantra is so thoroughly embedded into writing it, that I try to include "enhance" in function names where it makes sense. I might have to do likewise with my CSS.
Most of my apps are now simply HTMX + Go + SQLite. I've found it's enough for most projects.
I want to work with people like this. Build the right solution to the problem with the simplest technology you can.
Same. But I never got so lucky as to find a company where this is valued.
Nice clickbait title.
All the text is invisible for me in Firefox on Linux when the `--font-body` is set to `"Atkinson", sans-serif`. Setting it to `"Atkinson Hyperlegible", sans-serif` fixed it.
I read the article in the source code mode and thought it was intentional until I came back to the comments
P.S. your solution seems to have disabled the custom font instead of fixing it
You're right. Seems like it fails when the font is correctly selected. Another machine running Firefox on Linux didn't have a problem.
I'm not a web dev and my needs are pretty simple but I've had a lot of success with jekyll. It's a open source static-site generator.
I still to this day is baffled by all the idiot geniuses who thought it was OK to have "Please enable javascript" when opening a website without JS enabled. A good chunk in many pages are just static content. You should be able to render HTML. Ashte!
Interesting they went with Astro.Makes sense for a form-heavy site. No JS until you need it,and it handles page transitions cleanly.
Recently I had to migrate an old SpringBoot app that had a React front-end to a new cluster. Not wanting to mess with super-old dependencies, I opted to rewrite it on a new version of Java/SpringBoot. When it came to the frontend, I paused. I couldn't come up with a single good reason why this app needed React. I rewrote the frontend in straight HTML with a little bit of JavaScript for DOM manipulation. I literally used `var` instead of `let/const` just to drive the point home... (yes, that was overkill). But you know what I didn't need? A BUILD PROCESS! No npm deps. No vite/rsbuild/etc. It was like I had forgotten we could even DO that.
Don't get me wrong, I actually have enjoyed React over these past 10 years. But, including it blindly is just silly.
esm.ah let's you include "complicated" JS that isn't usually found in CDNs.
it doesn't work for everything and imo is worse for (p)react due to the lack of native JSX, but it does allow for bringing in stuff that usually takes an `npm install && npm build`
Yeah in this case, I needed to pop up a <dialog /> and take some form info, persist it via POST and then show the result of a "used" card/token. So there just wasn't a lot of need for libraries. I'm from the VERY old school so I do recall the fresh hell of including many deps via script tags (pre-Bower!)
OMG its JavaScript, I can't even.
Great work. I love it. We need more developers like you.
I wish more people take this approach, specially public services.
UK government services guidance is to use progressive enhancement
https://www.gov.uk/service-manual/technology/using-progressi...
> form submissions and redirects took a while to explain to my colleagues, on account of everyone being used to heavily client-side web applications.
Ah yes, when you have to explain to your esteemed web developer colleagues how the web works xD
Everyone reinventing what Rails has been saying all along...
If you're a "React person", as the article puts it, friendly reminder that you can render components to HTML and serve that to the user.
I have done exactly that on a project that was under similar constraints. The UI models live in .tsx files and the browser gets pure HTML with zero JS by default.
We are still in the early days of software and web development. This is the food service equivalent of a restaurant operator discovering that if they serve dog poo vs edible food they get more return customers.
in the bio ... "has over twenty years of experience building highly accessible and usable web applications"
why not take the html5 standard (see https://html.spec.whatwg.org/ ) and if needed (dont think so for these use cases... "for clients ranging from energy companies to political parties") htmx or alpinejs ...
> When we launched, the number of people completing the form doubled.
I don't want to be that guy, but the title is misleading. The number of users completing the form doubled.
No, it's actually people. FTA: 'My client was a utility company, and they had a big problem. To apply for their services, customers could either use an old ASP form on the website...'
Ie customers of the utility company were completing the form, not random users from the internet.
FTA? Free Trade Agreement?
From The Article
I understand people need to make arguments for things they like but provide more please.
What were some of the downsides? Illuminating the tradeoffs would elevate this post from good to great.
The downside mentioned by the author's replacement in the article is the unfortunate explanation for why this is rare in practice.
> "but that’s a lot more work for us."
And it's not that any individual or team is lazy. Most teams have a constant barrage of priorities to balance and are paid by companies valuing efficiency over everything. That said, I think the article makes a great case for adjusting our prioritization. Going a bit slower won't kill anyone, in fact doing so will probably save some.
Maybe this is heretical in today's AI hype climate but...weirdly due to the rise of AI, then AI-slop polluting everything, a lot of old fundamentals are coming back. Clear, well-structured, descriptive content on a well-built page has a better shot of being picked up for SEO/AEO/whatever which are the same best practices from 2005. A lot of these tips and tricks and hacks just aren't going to move the needle as much anymore imo.
SEO and accessibility laws have always been the most effective way to convince someone to build clean, well-structured webpages. Guess what, both are measures of how easy it is for a machine to extract content from your pages. AI is just the latest machine that wants to slurp up your soup of tags.
Beautiful story!
HTML doesnt have hundreds evangelists writing blogs about how cool it is. So it doesn’t really matter what the benefits are if juniors are going to keep showing up writing client-side react components. And the product managers letting them because of their sparkling (but naive) personality.
I will never stop singing the praises of plain HTML. It's accessible, it's portable, it's simple, but for some reason we need horrid JavaScript nonsense to operate.
It shows just how far gone webshit is when the obvious must be stated time and time again
"I took a very bold decision and built a new version of the site using Astro"
It will still work 30 years from now! 300 years! 3000 years. There will be no strange error messages in the log, no browser updates breaking stuff.
Personally, rather than this luxurious approach, I just do one giant form and store all values in local storage. If something is wrong have one message at the top listing which fields failed validation and why. Generate some css to put a red border around the fields.
Local storage might not be a good idea for such sensitive data but if you can get away with the simplicity it's lovely.
"And then the whole bus clapped..."
It's true. Also, if you despise bloat, you may like my JS-free LLM site: https://ch.at/
Totally agree, gov pages should be widely accessible. Also gov services should NEVER mandate internet access. There should always be a way for tech-illiterate people to ask someone, and fill their forms
>> There should always be a way for tech-illiterate people to ask someone, and fill their forms
Isn't that more open to fraud and abuse though by the "someone" whether that is a gov employee or otherwise.
If it's a government employee (as it should), it shouldn't be open to fraud and abuse, no. There should be a trail of what happened
Yes, there 'should' be a trail. And there's someone and a process to check that trail. And a process to find the issues, correctly so, and fix, and do all that on a timely basis, and reimburse the person for damages, ...
PS: I have actually suffered through some of this, badly so, though in an unrelated setting.
It comes back to which system is more robust/mistake-proof and more practically recoverable.
Sure, but that's how Governments worked for decades before widespread access to laptops / mobile phones...
FTA: 'To apply for their services, customers could either use an old ASP form on the website, or follow a manual process.'
Use Next.js Bro just know React and complaining it.
It'll be replaced by a new react app within a few hires lol
Use Next.js!!! Bro just know React and complaining it.
>She’s connected to the complementary WiFi and is browsing the GOV.UK pages on Housing Benefit. She’s not slicing fruit; she’s arming herself with knowledge.
>The PSP’s web browser is - charitably - pathetic. It is slow, frequently runs out of memory, and can only open 3 tabs at a time.
Alluring, an annoying property of private software development is that making websites and software in general inaccessible to lower end hardware is actually a positive effect, as it filters out 'undesirable' lower-income prospects.
That, along with pressure to produce fast, without much concern for quality (with notable privileged exceptions of luxury software like Apple or 1B+ user software like Google), as well as a disregard for sourcing "I don't care if you do it yourself, or npm install software from effectively unpaid volunteers", ends up in a state of software lacking craftmanship, software that one is not proud of to work in.
So funny to me that the kids nowadays are rediscovering HTML lol.
Having been building websites since the mid 90's, I laugh at terms like "HTML-first website"
It's like chai tea.
- Hill Hill Hill (Hill):
https://www.youtube.com/watch?v=NUyXiiIGDTo
The Sahara desert. It's not only repetitive but it repeats itself too as well.
These gems are brought to you by the department of redundancy department.
Naan bread
With au jus.
ATM Machine
The La Brea Tar Pits => The the tar tar pits
My favorite from Southern California.
OK, I have to admit, that one I didn't know.
It's only a matter of time until someone posts "Torpenhow Hill" -- which does not exist.
https://en.wikipedia.org/wiki/Torpenhow_Hill
explain?
"Chai" means "tea", so "Chai Tea" is "Tea Tea".
"ATM" means "Automatic Teller Machine", so "ATM Machine" is "Automatic Teller Machine Machine".
Both are mentioned in the animated movie "Spider-Man: Across the Spider-Verse".
Actually, in English, Chai does not mean tea, it means a specific flavor of tea. If you don't believe me, try ordering some Earl Grey Chai, see what happens.
https://www.google.com/maps/search/england+chai+london/
If your server is Indian, they'll likely react positively, and get you what you want.
Sure, but "chai tea" is still redundant. I have never used that term and ordered chai in many places without confusion.
it's redundant at a place that serves chai, but it isn't redundant at a place that does not serve chai, because you're skipping the "what is chai" question from whoever you're querying.
PIN number
Town names too: https://en.wikipedia.org/wiki/Montebello_della_Battaglia
Not in the Wikipedia page (but check the Italian version): it started as "Mons Belli" (Mount of the Battle) because of a battle fought by the Romans a few years before the Hannibal campaign. Then the original meaning was lost and it gained another "of battle" in the 1800s. Mount of the Battle of the Battle. Hopefully there won't be another one to add.
Lake Tahoe (Lake Big Lake). River Avon (River River). https://en.wikipedia.org/wiki/List_of_tautological_place_nam...
Except places are now offering Chai Latte Coffee so if you don't specifically order Chai Tea Latte, you could get some thing totally different than expected. I learned this the hard way.
> Chai, a word for tea in numerous languages
https://en.wikipedia.org/wiki/Chai
Reminds me of “pre-jit”
ie "aot"?
I’m sorry I can’t hear you over the Flash animation splash pages I was forced to sit through before being able to look up hours of operation.
As a teenager I remember going to a website for... a city, I think? And their 'sidebar' was a Java applet that did nothing but provide links for you to click with on-hover effects. The page used frames; the applet was in the left-side frame and the content was in the main frame on the right.
The applet took 30 seconds to load. Once it loaded, it showed five buttons to click to get to different sections of the site. When you clicked on one, instead of changing the content frame, it sent you to an entirely new frameset. This, of course, caused the sidebar to take another 30 seconds to load. Hitting the back button did the same thing.
Meanwhile, I knew someone whose friend made a little applet that he showed me; it was a Java applet that you could provide an image URL for and it would load the image and then, below the image, show a rippling effect as though you were looking at something on the shore of a rippling lake. This applet took less than a second to load and ran incredibly smoothly.
Java was a curse, not because Java was bad but because Java applets were written badly and used badly simply because they were neat.
Every language can say that bad developers write bad code with it while good developers write good code with it.
I would like to say the early interweb was just a learning experience, but today's interweb hasn't learned any of the lessons. It's just changed which language the lesson is being relearned
A lot of these tools, like React, are designed to embrace, extend extinguish the web. Why Microslop and Zuckerberg spend millions of dollars of dark PR claiming anyone who doesn't like React doesn't know what's going on is because it makes the web worse and less useful, which means you spend more time talking to Co-Pilot or bots on Facebook.
I did some work for a company that spent nearly a grand on a Flash animation for their title page of a red bouncing ball that would bounce from right to left along the letters of the word "Yipee" (yeah totally not ripping off Yahoo! were they?) until it landed in the crook of the Y, where it would spread down the middle - the finished logo had the Y made out of blue, yellow, and red stripes.
Every single person I showed it to including my then-70-something mother said "that just looks like menstrual bleeding".
Every single person said that.
They still went with it. Conversion rate? Dunno, never got numbers high enough to test the script.
I thought this article was missing a (1999) in the title.
Same, and it has certainly made me realize that I am now officially entering my "old man yelling at cloud" phase of my life, and I'm "only" 38!