The "Cheap" Web

In a sense OP wants E-ink screens and org/vimwiki/markdown. Styling is where the problem it claims to criticize begins; if you want a truly accessible hypertext book, then you don't need styling, the user should be in charge of this, and the default theme would be whatever easily readable font, black on white, headings 36pt bold, paragraphs 14pt regular, every elements as blocks. Epub, in usage, is actually a cool format and I almost prefer this type of "browsing" experience than what websites proposes.

Or maybe OP wants higher level CSS and HTML, and this is a problem because this leads to the invention of a new language with yet another complex ontology that claims to be "simpler than its lower level counterpart" but is not.

   Of course you could create some markdown with basic styling options, but I am pretty sure learning basic html and css is not that more complex than learning that specific markdown and how to operate it to finally get a website served on a specific server.

Web and styling became complex exactly when occurred the encounter between what was styling on the web in the 2000s, the emergence of the nowadays variety of devices that are able to browse it and the actual way people interacted with smartphones.

Today our browsers are almost OSes and it's almost like the complexity of what you can share with them is superior than the complexity of you could create on a 1990 pc natively. I mean, you can run godot engine on your browser : you can develop prototypes of projects with your friends just for fun, create your own private platform for communicating with people you like.

I notice a lot of hate toward JS, but honestly when I first stumbled upon it I felt like a dream came true; it gives you the power to create experiences and share them almost effortlessly, and the fact that bigTechs decided to make boring websites on crazy over-engineered frameworks doesn't change anything to this.

it gives you the power to create experiences and share them almost effortlessly

That's exactly where the hate comes from: I don't want you, the web designer, to have the power to create experiences, because you (collectively) use that power to force experiences on me that I don't want to have.

Don't visit the website

I hope we can have a more reasoned debate than this. This topic is really important and frankly there's so much to it I hope that hearing others I can learn more.

When I read some of the thread above what I am hearing is: "I don't want people to be able to run programs on my computer without asking explicitly". Which is exactly what the modern web and modern web browsers enable.

If you learned computing in the 90's you were very much told not to install things you do not trust. Everyone remembers downloading some game from the internet and suddenly their OS was trashed via malware or non subtle virus. The modern web is essentially a giant way to circumvent that. Sure only your browser can get shitted up but its the same browser you do your finance in and has access to your GPU.

Certainly it enables easy sharing of programs but so does something like java. Most OS's still recognize a .jar file as an executable and ask you about it before you can run. They never do that before javascript starts processing in your browser.

There are several other issues here such as: - Many essential processes such as banking, medical records and education all rely upon the modern web paradigm at this point. One cannot simply access them on an e-reader or using a browser with javascript disabled. Are you saying you think that someone should be left uneducated or lack medicine if they don't like running an unsecured browser? Should those with privacy concerns just incur the cost of having one "safe" device and one they do their "interacting with the web" one? - The modern web sandbox includes very high access to sensitive things. Many people's file systems are from a cloud platform all of which malware in the browser can access for example. - The modern web can be highly performant but generally its not and requires ever increasing hardware cost simply to do things like read a book.

Ironically, that decision to run Javascript without a prompt was probably a huge step forward for the security of most users. The 'do you trust this' model of security doesn't work well in practice once you can download programs from the internet - there's too much stuff you need to or want to trust to get on with things, and even if it's not actively malicious, it may be vulnerable.

Because Javascript can run in the browser without the assumption that you completely trust it, browser developers have put a load of work into restricting what it can do, even within the browser. Of course, sometimes there are holes in the sandbox - nothing is perfect - but I think it's vastly better than giving any program you decide to run complete access to your computer.

(Better for the majority, that is. If you're truly paranoid and have enough time, explicitly deciding what to trust can be better. But I think that's <1% of people - certainly not including me.)

This argument makes sense for the security angle. At least as it pertains to getting malware on your local machine.

But what about all of the "not security" but bad things that happen because we allow people to run code we have no choice over on our computer. The attention tracking features marked as tools to understand user intents are exfiltrating information from you perhaps when you don't expect.

Are you happy for example for someone to be logging where on a screen you are sitting and reading within a book or video. Do you not find it problematic for example that you could purchase a subscription to medium, but medium finds out you pause your computer to read descriptions of guns? Would you mind if they then sold this knowledge to Glock who then showed these ads on your work computer?

I get what you are saying for security. But "knowing" if and when something is happening is important. I may be worse or better at evaluating applications to run on my machine than the chrome team. But at least I know when I am entering into a risky situation.

That makes sense. But I think that's a money problem, not a technology problem - if the money was sloshing around desktop applications, I think we'd get a lot of locked down desktop applications tracking us. We kind of have that on mobile, indeed. Sometimes there are nice community built alternatives, but not that often.

Are you saying you think that someone > should be left uneducated or lack medicine if they don't like running an unsecured browser?

... As opposed to not liking to run an installer for an unsecured program?

App stores have the same issue. You just juggle the trust from some third party to another third party

If you want to do anything more complex than transfering text and images you'll need to trust a lot of things

If you want to do anything more complex than transfering text and images you'll need to trust a lot of things

Yes, exactly; that's why I wish the web were still based on the transfer of text and images.

Do I need to do anything more complex than exchange text and pictures to access my medical bills or records?

"Don't visit the website" is as simple as it needs to be.

This is HN. 90% of the time, the nitpicking here isn't about computer security, bandwidth efficiency, free speech, or whatever the high-minded principle of the day might be. People nitpick because they don't like something about the website, full stop.

It it wasn't the design, it would be the ideological bent of the writer. If not that, it would be annoyance as to why they put up an email signup form or use an analytics script, as if those aren't present on any of the other websites that are featured on the front page every day.

I'd wager that a good chunk wouldn't be happy unless they can extract your article into their homemmade RSS reader.

In real life you have even less freedom. There doesn't exist an equivalent of "reader mode" out in the streets. You can't mod reality or add a few scripts.

Don't like the topology of a parking lot? Tough luck, either navigate it or don't park there.

Hate the hospital's maze-like corridors? Sure they suck, but you are still getting operated there.

Your department's bulletin board is an unorganized mess? I'm sure you'll not drop out because of that.

Jar files have full access to your file system and no fine-grained checks like location and notifications last I checked.

I visit it and use reader mode, it works pretty well most of the time. My browser works for me, not them. Shame about the bloated filesizes though.

I didn't express myself as "the web designer" nor I implied that I wished to share experiences with you. What I meant is that I can share experiences to my friends via the browser, which act as a kind of a simple prototyping platform.

I want the power to create experiences for some people I know, in a niche setting, without intentionally inviting anyone else.

Yeah, the problem is everyone and their mother are "creating experiences" instead of giving me the info I'm actually looking for...

Honestly, I don't understand what you're referring to: are you intentionally visiting websites "you hate" ? I mean, there are a lot of good websites out of here that aren't proposing an experience and that allows you to read text properly, no ?

When I search for something I visit whatever I can find, no? Until I find what I'm looking for.

I may be biased but whenever I search something I go to whatever website that has an authority on X. When I don't know I who has authority on X I search for "recommendations about X", or I go to the website in charge of X. The only problem I encounter generally and recurrently is not js being an obstacle to find information but rather badly written anti-content.

Replace X with Y since X is a thing.

I go to whatever website that has an authority on X

But what do you do when the website gives you an 'experience' instead of info about X?

And no, <the company formerly known as Twitter> has no right to take X for themselves. Keep using it for unknown/unspecified items.

The problem you're describing exists in your own expectations.

Oh, well, that sounds fine! Sorry to mistake your meaning. I wish you could have gotten what you wanted in a way which did not have to change the fundamental nature of the web.

Well, most of the websites that proposes an experience are often rather lacking information-wise, no ? Don't you know of any website that reads like simple text ?

How is that different from native apps. I'd take the imposition of a website over the imposition to download a mobile app anytime

That's actually a matter of taste: in many cases I prefer desktop applications instead.

Do you sandbox them? I wouldn't trust random apps running on my PC

It doesn't. But on HN everything is terrible and get off my lawn.

I notice a lot of hate toward JS, but honestly when I first stumbled upon it I felt like a dream came true; it gives you the power to create experiences and share them almost effortlessly

The complaints about JS are almost never about the concept of a programming language that runs in your browser, and the upside you describe seems to apply to any language that runs in your browser.

The only languages browsers should understand are markup languages, scripting and programming: no, none of that. Documents, not programs.

Why?

Not OP, but the number of "documents" on the web that "require" programming/scripting that bring no value to those documents, is vast.

I'm not talking about documents/articles that have useful embedded video, or informative interactive graphs and things.

I'm talking about "documents" that require a multiple javascript libraries to load in order to inject a "<p>Hello, it's a nice day and I had waffles for breakfast" into a DOM.

There is no "graceful degredation" with these things anymore; it's either javascript on, or no content, which is baffling and frustrating. It's not just a "get off my lawn" thing, it has real implications on energy efficiency, bandwidth, and most of all accessibility: https://shkspr.mobi/blog/2021/01/the-unreasonable-effectiven...

Not OP, but the number of "documents" on the web that "require" programming/scripting that bring no value to those documents, is vast.

I mean, sure. But if our rubric is whether a technology being misused constitutes it having no business existing, that puts a shitload of tech on the chopping block.

I'm talking about "documents" that require a multiple javascript libraries to load in order to inject a "<p>Hello, it's a nice day and I had waffles for breakfast" into a DOM.

You're misunderstanding the function. The function is not to distribute documents efficiently. The function is to place documents somewhere they can be distributed in such a way where the business majors can edit them without needing to pay a web developer to do it correctly. Hence the absolute plague of BMS systems, which are sold to the aforementioned manager/consultant vampire class so they can make the website pop and don't need to deal with a 20-something rolling their eyes at them when they ask for that.

That's false, OP talks about that, many people wish the web to be static only and many people wants to be able to read pages "without js" and complain about the fact that they cannot read a page without js.

"... the user should be in charge of this..."

As a user, I choose textmode. I use a text-only browser as an HTML reader, works great for EPUB. No graphical fonts, no Javascript, no CSS, no color. Often I will save to .txt so I can read with less(1).

I have done the same, very probably with the same browser you're using (links). Just curious, do you handle images (or graphs) in EPUB files in some specific way?

Mostly, ebooks are perfectly fine to use without images, but not 100% of the time.

Skip the images.

Below is a quick and dirty script I use to make a single HTML file from an EPUB file. It outputs a _script_ that when run will output the HTML files in the EPUB as a single file to stdout. It is not perfect, e.g., sometimes the order of the HTML files may need to be adjusted, but I rarely bother. I open the HTML file in links and optionally save as .txt. I don't care for UTF-8 or other non-ASCII characters.

EPUB to single HTML file.

    #!/bin/sh
    test ${#1} -ge 1||exec echo usage: $0 1.epub;
    printf "(\n"
    for x in ${1-*.epub};do
    echo printf  \'\\n\\n,,,,,,,,,,,,,,,,,,,,, "$x" ,,,,,,,,,,,,,,,,,,,,,,,\\n\\n\'
    7z l $x|sed -n "s/.* /7z x -so $x /;/nav.xhtml/d;/[tx]ml$/p"|sort -n
    done
    echo ")|tr -cd '[\\\12\40-\176]'"

Yeah, I had a similar script of my own, utilizing the fact that EPUBs are essentially zip files. Thanks for sharing yours. I'm a great fan of your scripts for text-only browsing etc.

There's a difference between content that is meant to be communicated vs. "widgets" that do something on your screen. CSS wants to support widgets, therefore it is complex.

Apps vs. Content. Web is for both but if you can and want to do just content it should be easy and simple.

Has anyone proposed CommonMark over HTTP? Like, just shove markdown in a GET response with

  Content-Type: text/markdown

And let the client decide how to render it. It'd be like Gopher, but modern.

Does Gemini[0] look interesting to you?

[0]: https://en.wikipedia.org/wiki/Gemini_(protocol)

Gemini is like modern Gopher, but it is also very opinionated. The protocol does not support extensions and revision, it refuses HTTP, mandates TLS, etc...

That's not the same as having browsers support the markdown content type.

You might disagree about where to draw the line but you can see why they drew a line in the sand. text/markdown would very quickly grow to be everything HTML is.

There's probably a pithy "every internet protocol expands until it can serve ads" lesson in there somewhere.

I like your pithy remark! I'm calling that "Flir's Law" from now on, after your username.

I just stole it from jmz - "every program attempts to expand until it can read mail" - I think attribution might be overkill. But thanks for the thought.

Markdown generally does include everything HTML is (because you're allowed to write HTML in markdown)

Gemini isn't over HTTP and it's not Markdown.

Could we go with a format that doesn’t have so much unnecessary complexity?

I've never heard markdown referred to as unnecessarily complex before. Genuinely curious, what features / complexity would you strip out?

I’d prefer everything to have the same syntax, like in HTML (or I should better say XML).

When every feature has different syntax, like in Markdown, it makes parsers much larger and leads to an extreme amount of edge cases that every parser has to handle. For example, when you see an asterisk, you have to decide whether it denotes italics, bold or a bulleted list, which might not be trivial.

Generating correct Markdown from a different is also much harder. In HTML, you just need to escape all your less-than signs, ampersands and quotes. In Markdown, you need to check whether your content is going to trigger various edge cases and deal with them accordingly.

You don't need to escape all your ampersands and quotes. An ampersand that doesn't form an html entity will render as is, quotes are perfectly fine unless used in an HTML attribute. And in an html attribute, less thans are fine. The body of a script tag has different parsing rules.

You don’t need to escape them, but it certainly doesn’t hurt. And script tags are generally not a problem when building something from a template, but they are a reason why XML is more “pure” than HTML.

Yeah, I guess that's a really good point. I've never thought much about it before, but you're right that the simplicity of markdown is probably more from the point of view of the human writer/reader, and is really just passing the complexity down to the parser.

Good luck getting the browsers to implement anything so useful. They won't even update their default stylesheets.

I set up my blog with that kind of mentality - I just want to write some markdown and have the browser render it whatever way it needs to - and found a one-liner to pull in some Javascript to do that. But the kind of people who like this stuff tend to hate Javascript, so I get it from both sides.

I don't use other browsers so I can't be sure if this is widespread, but Firefox already has a "reader mode" where it cuts out ads/sidebars etc and applies its own very basic stylesheet. Markdown-over-HTTP could very easily use that preexisting frontend.

I set up my blog with that kind of mentality - I just want to write some markdown and have the browser render it whatever way it needs to

I made my own blogging platform where I write Markdown, and it transforms the markdown into HTML before delivering it to the client. The only JS involved is highlight.js, and that's only loaded on posts where I specifically ask for it - the whole blog works normally with no JS at all. It's still missing some core features like RSS, but I'll get around to that.

I made my own blogging platform where I write Markdown, and it transforms the markdown into HTML before delivering it to the client. The only JS involved is highlight.js, and that's only loaded on posts where I specifically ask for it - the whole blog works normally with no JS at all.

There are plenty of platforms that will render markdown into HTML (e.g. Jekyll); I specifically wanted to actually serve the markdown as-is, so that the "source" for any post is always available.

I've thought about adding this to my own platform. Just checking the Accept header for text/html or text/markdown, then sending them whatever they prefer. I just don't know if anyone would actually use it.

“But but but.. how do we shove ads everywhere and track the targets?”

That's the philosphy behind Markus Docnet

https://github.com/markusdocnet

This used to be something one could sketch up in a couple hours on the web, with registerContentHandler.

Removed in 2018. I don't totally disagree with the reasoning, but when the web kills amazing flexibility points for lack of heavy adoption it makes me sad. There's so many good ideas that should be left in, even though society isn't actually so bloody awesome as it should be. https://groups.google.com/g/mozilla.dev.platform/c/jeTDLz38_...

Also, scoffing at 0.2% of users of a web browser is multiple orders of magnitude away from where the line should be; these people are a fucking joke & shame the fuck on them, losers. Jonathan Kingston, what the heck dude? What the frelling heck?

Also no other browsers implemented it! Adoption needed help! Not to be murdered.

Tl;Dr, adding new content types used to be easy & simple & you could register a site as a content handler but they always kept the feature locked down then killed it without ever letting people go for it at all. Gatekept.

This would also be amazing for screen readers and other accessibility software.

"Most websites should be compatible with screenreaders", and yet, the first line of text in the website is "♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡♥♡"; With the spoken form of: "Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart"

Ha! Thanks for catching that! I think that demonstrates how difficult it is to make even a simple website accessible.

I assumed that since my browser's "reader mode" was all good, then screen readers would know how to parse it too.

I do like the heart aesthetic, so how exactly would I make it more accessible?

role="presentation" will remove it from the a11y tree! Easy fix. Most browsers will notice the characters in the element are not human readable and make some tweaks to make "heartheartheartheartheartheartheartheart" less likely though, even in the current version!

Thank you! I added it.

In case you're maybe interested, I wrote up a little thing about how I think of accessible interfaces and designing for them:

https://graypegg.com/2023/11/25/the-private-definition-of-ac...

Edit: decided to post it here, https://news.ycombinator.com/item?id=38683774

Thank you!

The problem is that accessibility has been treated as an afterthought, so designers, UX and people crafting HTML usually are not in habit of embedding accessibility into their works. It often comes when someone disabled complaints or if there is a regulation they didn't know about and there is "oh sh..." moment where accessibility is tacked onto already done project.

So it's difficult in the sense that people don't think about it, there are not so many tutorials available and so on.

YES. I think it's this unique form of after thought, where people imagine accessibility as a "score" that you can "increase". The web developer lowest-common-denominator cares about it, but only at this weird surface level. More aria tags = more accessibility points. It becomes this weird game where we end up with insanely noisy markup, which creates a horrible described interface for anyone actually using assistive technology.

It's an interface like any other. We'd be weirded out if designers graded their screen specs as "visual" enough. More pixels are required in this area, think of our screen users! They need more pixels!

Depends - do you want screen readers to hear something? If so, add a `aria-label="heart"` or similar. If not, add `role="presentation"` to remove it from the accessibility tree altogether.

You might also want to change the element from a `header` to something else, like a `div` or a `span` or even a `figure` instead. Then put "potato.cheap is home of the... The "Cheap" Web" inside the `header` instead.

How should an interested personal web developer find how their website sounds through a screen reader?

Voice Over is insanely easy to work with, I recommend it to everyone I work with. I would say it takes maybe 20minutes to get a hang of the basic shortcuts and navigation modes.

I've been thinking about making a "VO for Web Devs" cheat sheet at some point, would that be a useful thing for you?

(Of course, I'm making assumptions about being in the apple garden.)

Well, as it happens I'm not, but I imagine it would be useful for people in general, and I'll promise you an upvote on HN if I see it. :)

I imagine you can make a solid guide if you really focus it on the use case of "here's how to evaluate your site for this specific purpose".

Noted! I sadly don't have much experience with the options for windows / linux flavours, but I'm sure something like NVDA [0] could be useful for you to test these things! The idea would be to learn the tool first, (feel at least capable to open the browser from your taskbar and jump around a few tabs for example. This is generally a 20min task.) then set up a few goals to do on your site. Finding a specific page, or making an account. Even with sight, and knowledge of your site, you'll very quickly hear the odd-ball places where the descriptions give duplicated info or even false info. Just don't touch the mouse! Experience the interface, as if this was your portal into the internet.

I think I'll get moving on that cheat sheet then! Thanks eh! :)

[0] https://www.nvaccess.org/download/

Good accessibility tip: Enable any accessibility options you need to, and then turn your monitor off. If you can't navigate your website, there's an issue.

Seems accurate? ;)

My browser (Chrome 120) does correctly intuit that the header is not a landmark, so it at least won't be included in the landmark rotor. (In VO)

Still though, it would be ideal to remove that from a11y tree since it's totally useless in there.

The thing that makes me disagree with proponents of the "smol"/"cheap"/etc Web is their totally misguided insistence on personal webpages as a sui generis medium for self-expression wrt layout and styling.

Until we adopt simple and stable building materials, all websites will continue to look the same.

Yeah, big deal. Who cares if websites look the same? Wait, let me put that another way: we should care whether websites look the same in the sense that more of them should look similar to one another than they do today—namely, they should look how you and I (the readers) wish for them to look for our own comfort and convenience—which means fewer dark starry sky backgrounds, fewer text drop shadows and typefaces reminiscent of comic book speech bubbles, and fewer distracting marquees. Way less kitsch, not more.

HN, Twitter, Reddit, Facebook, SMS, etc. all have a limited range of expression through form. Somehow we get by. Completely dissimilar books look a lot more "the same" than websites do, and that's not a bad thing. You never see book authors complaining about control over the printed page the way that you see self-styled web designers insisting on their manifest destiny over the user's viewport for the purposes of self-expression. Webdevs need to get over themselves, and the wider world needs to stop letting them get by with (or encouraging) this kind of narcissism.

You can have all web sites look the same by setting reader view as default in your browser. It is a great improvement for browsing the web, both on desktop and phone.

If somebody invented a "forum view" to make all forums look the same on our browsers, then I think the circle would be complete. It shouldn't be too difficult either, since most run on similar software.

You should learn about the Gemini protocol! It's an alternative to HTTPS that uses a document markup language similar to Markdown instead of HTML, and uses a deliberately simplified, feature-frozen protocol with the explicit goal that anyone should be able to write their own Gemini browser "in a weekend". No scripting or styling support, just documents with links and images in them.

Here are some informative webpages on the topic, neither of them mine:

https://geminiprotocol.net/docs/faq.gmi

https://github.com/skyjake/lagrange

I think the Gemini stuff is really neat, and I've already experimented with Lagrange and Kristall browsers. However, the information I'm interested in isn't published on those networks, and I think that's true for most of us. Reader view is a seamless way for me to have control of how the web is presented to me. An RSS client + browser with reader mode is the perfect combo for most online stuff that is information (and not function, like an e-shop).

As I understand it, forum functions are not a feature of the Gemini protocol?

Forum functionality is possible; the protocol allows for user input in the form of prompting the user to enter a string and sending that to the server before loading the page. Example: gemini://chat.mozz.us/

Webdevs need to get over themselves, and the wider world needs to stop letting them get by with (or encouraging) this kind of narcissism.

What a presumptuous statement! Authoritarian, even. It's one thing for someone else's website, such as HN, Reddit, Wikipedia, etc., to impose a homogeneous form on posts submitted by guests on the site. It's quite another thing to assert that "the wider world" should "stop letting [independent webdevs]" deviate from a similar standard of homogeneity.

If you don't like the appearance of someone else's webpage, maybe just don't go to that webpage?

Ironic take for a comment that exists only as a consequence of what started out as someone complaining about how boring other people's websites look; there's no way out of this—either you agree with the ability to criticize, in which case mine is fair, or you disagree, in which case it's not but then yours isn't either and neither you nor anyone else should bother yourselves or others about the sameness of anyone's websites (instead, focus on not going to the ones that bother you so much in their sameness).

(And choosing to read my comment as an "authoritarian" imposition on "independent webdevs"—not actually stated anywhere? How presumptuous.)

I'm going to put a tiled JPEG of an artist's rendition of the night sky in the background of my blog, to assert my liberties and affirm my right to be silly and have fun

Author here! I largely agree with you.

I love Del Taco. I love the speed and consistency I get from franchises.

At the same time, my best friend recently sent me a picture of a fancy swordfish dish he cooked up, and I'm bummed that we don't get to see each other as often as we used to.

And as my daughter grows older, I'm becoming a bit frustrated at how difficult it is to purchase foods that were not doused in neurotic pesticides. So I'm learning how to grow our food.

Growing food isn't easy, but the farming community is cool, and I've been having a good time of it.

And so sometimes I wish building a website could be more like learning farming, and I think it could be.

Authors like Mark Z. Danielewski do more with the layout in their books than most and I love them for it. I like seeing people try out different ideas and do things that might be considered less-than-ideal to view

HN, Twitter, Reddit, Facebook, SMS, etc. all have a limited range of expression through form. Somehow we get by.

Disagreement exists as to how well this is going for us. (cf., content collapse)

Who cares if websites look the same?

A lot of books look the same too, but that works pretty well for me tbh. I do not find books to be homogeneous as a result.

We're probably past the point in the web where formative experiments with layout are useful for most people, and most situations can be handled with a light CMS. That still leaves room for massive variance and personal control. Even just linktree levels of customization have a lot more spice than facebook. I think current big tech is still reacting to myspace allowing users to write raw CSS, and there's a lot of space for a swing back, especially given the current web's focus on "creators".

Love it — I think something of value was lost when we collectively decided in 2005 that everyone’s page should look exactly the same (i.e. TheFacebook™ profiles)

Sadly that, combined with the changes Google made to deprioritize results from personal sites like these, all but destroyed these neat bespoke pages.

I mean, that's because Facebook Pages are cheap. Businesses don't want to spend money on anything they don't have to, and people aren't much different. The argument for learning to write even basic HTML only to upload it even to something like Amazon S3 (free) and direct a web address to it (not free) or spend money on hosting to not have to learn how to deposit files into an S3 bucket is a LOT of time to spend to accomplish this task. Instead, you can outsource all of that time, effort, and labor to Facebook who already built a website. And yeah, it sucks ass and it's full of undesirable people, but it's free and it accomplishes the same goal.

Google, meanwhile, sees that everyone is setting up Facebook profiles instead of websites because it's easier, and now businesses websites are sometimes quite out of date compared to their Facebook pages so they begin prioritizing results from network sites over regular pages. And now people can follow their favorite businesses on Facebook and be at least somewhat assured they'll always know if, for example, they have to close up one day unexpectedly and won't be doing business again until tomorrow.

This is not meant to be an argument against the cheap web, mind: I'm just saying that people trend towards what is the lowest-cost option (by money, time, effort, or any combination) and use it because it works and gets the job done. I still keep a website, because I value my independence and have no interest in having my thoughts dumbed down to suit the preferences of corporate America. But the vast majority of people do not care about that (until they get smacked with a ban-hammer anyway) and will just stick to the easy option.

Also, I would state: there is no reason we couldn't have a social media site that functions as a public utility instead of a for-profit corporation. We could just... have a Facebook that doesn't need to piss everyone off to make money. Social media without profit motive could be a massive boon to our society.

The other conclusion I've come to is that web designers way overcomplicate stuff. So I don't want to do a Facebook page. What are my relatively low-cost options?

Well, I can have a really basic HTML site but that's a little late-90s.

OK, a Wordpress site then. Now I'm dealing with introductory rates that shoot up after a year unless I just use a standard VPS and now I'm having to do a lot of stuff myself. And, as far as I can tell, most of the literally thousands of templates/themes out there don't make it straightforward to just have a nice homepage with a simple blog post that doesn't require a high-resolution photograph to go with it.

I'm probably settling on just sticking with Blogger using a new theme. I'm not really happy with that solution--but I'm not a web designer. I'm willing to spend some money but I haven't really found an answer I love.

The other conclusion I've come to is that web designers way overcomplicate stuff.

I mean, there's a floor right? The floor being that you need a server that can respond to HTTP requests, yours or otherwise; you need said server accessible from an address to which you can map a DNS record to; and you need HTML files that can be served in response to that request. Few of these are free, but many can be had for very cheap.

So I don't want to do a Facebook page. What are my relatively low-cost options?

Amazon S3 with Cloudflare in front of it and a domain name. Probably about $15 per year unless you manage enough traffic to get billed by Amazon.

Well, I can have a really basic HTML site but that's a little late-90s.

I think that's a little reductive. HTML sites can be just about anything and they can be quite fancy if you're willing to put the time in to learn something like Jekyll. As a side hustle I maintain a couple of basic HTML websites for local businesses, and just bill them for the annuals + the time I spend tweaking things for them, which thanks to my workflows, isn't much.

OK, a Wordpress site then. .... I'm willing to spend some money but I haven't really found an answer I love.

I mean, that's just the platform-ification of the Internet described. There's a floor of technical requirements and software experience required to publish on the Internet, and the vast majority of people like yourself do not possess it and aren't interested in acquiring it. That's fine, but it's akin to saying "I want to print a newsletter but don't want to learn how printers work." You've selected yourself out of your own goal.

HTML, CSS and a bit of JavaScript can make dynamic and beautiful websites. But it's not without a learning curve, and if your response to that is "well I want to do this without learning HTML/CSS/JS" well, that's the central conceit of Facebook pages, Shopify websites, and Wordpress.

What I've been doing is a homepage on S3 with a few static HTML pages and a link to Blogger for the blog. And a domain name I've had for ages pointed at it.

But, if I'm going to be on Blogger anyway for blogging, my best bet is probably to find a reasonable way to also make it my home page destination. Or maybe not. We'll see. I'm just not sure I'm especially interested in spending a lot of time on tech that isn't my ultimate objective.

So I don't want to do a Facebook page. What are my relatively low-cost options?

There's some level of complexity in all of those choices compared to just having a FB page.

I liken it to taking the bus everyday over buying an e-bike. Buying a bike is complicated, expensive, easy to make mistakes. But at the end of the day, you'll have a vehicle to take you from A to B.

You don't have to buy one, as long as the bus keeps running. But one day the bus route or schedule might change, and it no longer takes you where you wanted to go. This is when having a bike, warts and all, would've been be useful.

States like to control, internet state of mind is 100% liberal, and both live in a capitalist world. Internet won its freedom, but freedom did not overcome the greed from capitalism.

Good news is you can still go to mastodon, after all.

Internet won its freedom

I'm gonna need a real big citation on that. If by freedom you mean what a lot of people mean, which is that it opens potential revenue streams and markets and the oppression has it's origins in capital holders, then sure. "Freedom" in it's modern vernacular tends to mean this, hence "bringing freedom to Iraq" actually meaning "making sure Iraq's oil and other assets are open to exploitation by western corporations." By this definition the Internet has never been more "free."

By the liberalism/capitalism definition: anyone can make its website, its app, its blog etc.

Which is the opposite of "we need an official social network" intervention

In retrospect a lot of the customization you could do on MySpace was madness by todays standard.

Being able to set your own background, color, font, potentially animated, with auto-playing music of your choice was like letting the average person be a web developer.

I don't understand this sentiment. It implies that an arbitrary web developer, who has never met someone, knows better how the want to portray that person, than that person themselves.

It also feels very anti-hacker to me. A core ethos across open source software is that each user should be in charge of how their software behaves, should they choose to be. That includes the lowly "average person" who is not a web developer.

Sounds like this article is advocating for everyones page to look the same. If every page is built using simple technology, the user can style it themselves, not the developer.

Has there been a serious attempt to define an html/css subset that achieves these kinds of goals? Something that a mere mortal could implement and would cover the vast majority of web designs?

I understand the urge to throw out the old and replace it with a new system, but that would be a huge blow to accessiblilty and adoption.

I have been asking for a while if it could be a good idea to make something like asm.js but for webpages:

Something to put in a meta tag or something early in the page that lets the browser know this webpage will only use a known-to-be-fast-and-predictable subset of html and css and only use js from a standardized library that provides things like autocomplete and other actually high value interactions.

i'm interested in the answer to this question as well and did a bunch of searching a while back to no avail. maybe the world is waiting for us to start?

I have absolutely no idea about html rendering so my only contributions so far has been to ask/suggest to people who could know.

But it is an idea that I keep getting reminded about.

It could work adoption wise, since if it can be done technically it can be adapted independently of everyone else on backend and frontend:

- backwards compatible, if a browser doesn't take advantage of it it renders completely ordinary and in supporting browsers if a website doesn't use this trick it just renders as any other website

- If just one browser and one website does implement this then there will be a full setup. If the idea worked and for example Firefox supported this and Wikipedia or someone else took advantage of it suddenly everyone who used Wikipedia would se faster load and less battery usage.

The tag is `<!DOCTYPE html>` and it should be the first line in the file/stream. Doing anything else will enable "quirks mode"[1]. Beyond that, the best thing you can do is avoid features that haven't been standardised yet or that were only standardised recently.

[1]: https://developer.mozilla.org/en-US/docs/Web/HTML/Quirks_Mod...

You don't need JS for basic autocomplete in 2023, you can use "datalist" in HTML: https://developer.mozilla.org/en-US/docs/Web/HTML/Element/da...

A lot of basic high value interactions are more directly encoded in HTML today than a lot of developers expect to need JS for. Some other tags to pay attention to: summary/details, progress, meter, input type="color|date|time|datetime|range".

It's an interesting relearning project, sometimes, how much the high level interactivity bits of HTML have changed since, for instance, the jQuery era.

You don't need JS for basic autocomplete in 2023, you can use "datalist" in HTML:

This does not seem to load data dynamically, it seems to be a way to show data from a predefined list?

Most "basic" autocomplete isn't all that dynamic and you can prepopulate on the server side reasonably well.

Sure, you still need JS to fetch a dynamic changing list, but you can also just update datalist options elements with JS rather than implement a full separate UX for autocomplete today. You are limited in the ability to CSS style datalist options so a lot of developers are still going to feel pressure in 2023 from "pixel perfect" UX designers to continue to reimplement that wheel, but the version of the JS that just updates a datalist after a fetch is likely much simpler than building a full "autocomplete control".

https://amp.dev/ is that. No one wants to use it because the very first line of the mandatory js file is about advertising metrics.

There are reasons why I didn't mention amp :-)

amp isn't meant to solve our problem but Googles problems.

maybe a good starting point would be the subset which is supported by dillo or elinks today?

This is why we should be using decentralized XMPP servers to chat since they actually run on potato hardware unlike other chat options.

Matrix Homeservers like Synapse can run on anything now, and there is conduit which should be lighter too.

Matrix is not lightweight or simple.

Matrix shifts much of the burden to the server so that clients can be simpler (except for e2ee, which has to be done clientside).

Which if true is still a burden that prevents simple self-hosting. You can (& I did briefly) install an XMPP server on a 10-year-old smart phone with postmarketOS on your network & have a reasonable experience.

Last I checked, the way the protocol works still causes a huge amount of load for minimal usage. Because of the way events need to propagate and the types of events that propagate

IRC never died, you know :P

It’s not decentralized so somebody’s server has to take the hit even if the protocol is about as barebones as it gets. You’re also missing some basic features like usernames & rooms having collisions—heck usernames & rooms can’t even contain Unicode which hurts accessibility.

Google Chat on my 2008 Blackberry (XMPP client) was the fastest IM experience I've ever had. I finally understood the meaning of "Crackberry" when I had that setup. The combination of the instantaneous speed and keyboard was something I haven't experienced since.

Very interesting, thanks for posting!

I am not sure if the way to go is a cheap web, small web, slow web, indie web, or maybe some combination of these attributes, but something about the current web feels off, mass produced and superficial.

The old sites created by people who did cool stuff just for the sake of doing it are so hard to find now. I miss them.

They're hard to find because they're not really linked anywhere. We used to have Web directories arranged by subject headings that tried to be mostly comprehensive and put some serious curation effort into that goal, such as DMOZ.org - but there's no modern equivalent to that. People like to complain about how the whole SEO issue has made search results useless as of late, but that if anything is downstream of the overall lack of manual curation.

That's so true, it's almost like a return to the beginning (Yahoo Web Directory). We need to find a way to curate content that scales but in a way that eliminates the incentive for bad actors to try SEO-like approaches to game the process.

That feels impossible without a feasible alternative to the ad-based revenue model.

Pinboard and similar bookmarking sites could play that role.

Blogrolls, folksonomies (tags), etc. just pretty much faded away once the web really went mainstream. How many people ever really used lists on Twitter?

Directories died because

a) Websites change hands frequently and enshittify rapidly.

b) Nobody was getting paid to check in on websites and make sure they were still on-topic

c) Most 'good' websites don't update every day, and people hate RSS and email signups. So activity and engagement on those sites are completely overshadowed by the bottomless pit of content that is social media.

I always thought search engines could go back to using titles, description, and meta tags, but limit the number of characters/meta tags accepted.

They're also hard to find now because the web has become so much more than what it was back then. So many more people and companies are on the web now, so _good_ smaller personal pages make up a much smaller percentage of what's out there. I think this might be a larger cause of this issue than less curation.

That comment about a "simple markup language" is interesting - anyone know what OP is referring to? They mentioned they're attempting to build one.

I don't really get this point... HTML is not _that_ complicated. I mean is:

really "more complex" than something like:

# this

## this

### this

You still need to know that '#' means header / title, and that more of them leads to a lower header level... and this is just for the most basic concept of web pages: headers. Okay, even if it's not markdown, no matter what system you cook up, the syntactic complexity to support the huge number of elements that exist in HTML has to live _somewhere_ right? (Yes, you can argue maybe not ALL of the 100+ HTML tags aren't needed, but probably most of them are!)

While there is something to be said for mega purists who don't even like JavaScript, we're supposed to not even like HTML anymore either? I think at some point you have to accept that building things for the web involves some minimum level of complexity...

Author here!

If you haven't done it before, try teaching a non-programmer how to make and deploy a website. It's easy to forget where the sharp edges are because we cut ourselves on them so frequently.

I've observed that HTML/CSS/JS sets a minimum level of complexity that is a non-starter for most people.

I personally believe that most webpages could be written, arranged, and styled in formats easier than markdown (without a scripting language). But I could be wrong! Maybe nobody's done it yet because it's simply not possible haha

The vast majority of people who wrote their own websites back in the 90s were non-programmers. And the complexity of HTML/CSS/JS is oversold in a world where nine year olds are already building games on Roblox.

Agreed, pure semantic HTML5 with minimal (if any) CSS is pretty much equivalent to what Gemini and markdown are going for. HTML5 is also a proper standard unlike markdown, which exists in a variety of incompatible versions.

OP here!

I've got a few different competing ideas at various stages of development.

One idea builds on the programming language I'm building. I am growing more doubtful that it's going to provide the beginner-friendly experience I yearn for. It might be as easy as Elm one day, but that's not good enough for grandma.

[1] https://scrapscript.org

Another idea is like "markdown, but with a grid model". That one also does not feel very good right now.

I've also been toying around with something called "BYOD" (bring your own data), where you stitch things together in a simple layout format. This ends up feeling like powerpoint via plaintext, and surprisingly pleasant.

Love that - I also have a few competing ideas around something that sounds very similar. My goal is to create a textual interface that eschews traditional programming conventions in favor of a grammar and syntax that feels familiar to UI designers. I'm calling the overall effort "matry" - right now it's a language but that could change in the future: https://matry.design/

If you are spending less than $1 per hour on your entertainment (podcasts, videos, articles, games, books, etc.), consider finding ways to support creators and the infrastructure that supports them.

So I should send them extra money because I keep using it? I bought the discography of a band (0.2665 € / hour according to last.fm so far, obviously falling), should I now set up a monthly donation just because I keep listening? It seems like advocating to only use subscriptions and never owning anything.

This is such a weird take.

There are other ways besides subscriptions and ownership: Patronage is a big and popular one, through platforms like Steady or Patreon, or even just direct bank transfers. Another big one is commissions, though that's usually more of a thing with artists than with musicians or videomakers.

Yes, there are different ways to subscribe. And again, I find it extremely weird to be supposed to keep paying indefinitely because you keep using something you bought.

It's not to keep using something you bought.

It's because you value the artist, and want to support them continuing to make new things that you'll like in the future.

Rapidly changing economic structures can strip away sufficient economic reward via the ordinary way of 'purchasing' things, such that continuing to make them is no longer viable. E.g., Spotify paying artists $3.3 per one thousand streams, but almost no one buying CDs.

They are proposing ways — in this environment — to direct more of the funds directly to the creator, so that they can keep creating. But if continued creation is not of value to you, you are free to skip it.

If you get a lot of value from their music, you may want to support them via merch and concerts!

I've found that "$1 per hour" is simply a good starting point for figuring out which things in my life are worth supporting. Obviously, it may be different for other people :)

Going to concerts is sadly not a "want to go" thing, but "is there anything that’s close by", and I say that as someone living in a metal country like Germany.

According to this blog, I owe Trent Reznor hundreds of dollars for a second hand Nine Inch Nails CD I got for $5

the whole cheap/small web thing has me thinking of those people who buy used clothes but think nothing of paying $15 for a coffee at hipsterCafe. something virtue signally about it.

maybe it's the "solarpunk philosophy" and the "synthesize serendipity"-ies or quoting steve jobs, one of the biggest a*holes in the history of tech (imo).

although i agree strongly with everything here https://potato.cheap/#cheap

buy used clothes but think nothing of paying $15 for a coffee at hipsterCafe

The funny thing is Starbucks is by far the most expensive and crap quality coffee you can get most anywhere. "hipsterIndependentCafe" is consistently cheaper and better if you're a coffee snob.

Also people like me prefer to buy used as much as possible not to save money but to reduce waste. The romantics amongst us also prefer to have our goods well-storied

Offtopic - but that is wildly untrue.(for the US, at least)

Starbucks makes consistently good espresso. The amount of times I had - "oh, this hipster cafe is number one in the area" to just go dump the coffee and dip into Starbucks - is very high.

But going back to the topic - the consistency of Starbucks and the inconsistency of coffee shops (even in places like NYC) is reminiscent of why we gravitated to repeatable designs that we see today.

The visually appealing websites, that are not hard to read is "Starbucks". Be it visually "extravagant"(TheVerge) or simple(Yahoo Finance) they are fairly standard... these follow time proven visual standards set out by printed magazine, books and newspapers.

This potato.cheap website is a pain to read. Hard to focus on the message and the text is way too clumped. So are many of the sites it lists as examples.

HN is barebones Starbucks, not a pointlessly overbearing hipster cafe.

the whole cheap/small web thing has me thinking of those people who buy used clothes but think nothing of paying $15 for a coffee at hipsterCafe. something virtue signally about it.

This says nothing about the article or topic, but says a lot about your mental baggage.

people who shop at goodwill but think nothing of paying $15 for a coffee at hipsterCafe

I don't go in for fancy coffee with any regularity, but I do care a lot more about what goes in my body than what covers it, so I don't really see a contradiction here. Honestly, the older I get, things that may help my health (and I think there's a lot of "woo" there, but I think we can all agree that a decent restaurant is on average healthier than a McDonald's) are where I'm willing to spend money.

the whole cheap/small web thing has me thinking of those people who buy used clothes but think nothing of paying $15 for a coffee at hipsterCafe. something virtue signally about it.

Very often hipsterCafe commoditizes a demand for quality in food by offering something slightly better than average at five-fold the price, but I can see that hipster-vegan-bum would find themselves in a perfectly coherent way of life buying clothes from thrift shops and spending a lot for food, sourcing for ethical cooperatives or whatever. It may be a reflection of their privilege, it may be a virtue signal-ist modus operandi, but aren't contradictory habits.

maybe it's the "solarpunk philosophy" and the "synthesize serendipity"-ies or quoting steve jobs, one of the biggest a*holes in the history of tech (imo).

Don't you think that Steve Jobs scammed a lot of artsy people selling them a dream of olistic tech and sustainability, and that it's perfectly sane and just for them to clasp at that dream and trying to realize it, rejecting the evils of the false prophet while not throwing the baby with the bathwater?

I agree that we need a new - less crufty - standard to replace HTML/CSS/JS but there's another point which I feel has been sorely missed in previous efforts.

I want a FUN standard. I want to make cool looking sites that are relatively scalable based on user hardware.

Alternatives like Gopher and Gemini never really scratched that itch due to how sterile they are.

The problem you have is that by defining a "FUN" standard, you've probably created a basin of attraction [1] for what is basically the web today. It doesn't take much before you've basically let the ocean in.

There are clearly many points between Gopher/Gemini and the modern web... but I'm not sure any of them are stable. Between the difficulty of keeping out features in a principled manner in what will inevitably become a group effort and how easy it is to accidentally spec something that turns out to be a lot more complicated than you thought it was, you're pretty much working in a space where the Horrors of the Web are lurking just outside your door, and you'd be surprised which missteps will let them in.

[1]: https://en.wikipedia.org/wiki/Attractor

This is very deep. I like this mathematical explanation of why we get either Gemini, which nobody uses, or the modern web.

When people say math doesn't matter and is useless, I see it as rather akin to standing in Adam Savage's workshop and wondering why he has collected so much useless stuff and proudly declaring their inability to understand or utilize the contents.

As a bonus, here's another IMHO highly underrated gem: https://en.wikipedia.org/wiki/Coevolution Not only are there characteristic patterns intrinsic to evolution (in all its forms, far beyond just biology), there are characteristic patterns to interacting evolving systems. And that's, like, basically everything in the world. Super useful tool.

The problem will always be that 'fun' rapidly develops complexity to support 'cool looking'.

OH yeah, thank you for saying that. Gemini was a short 1 month obsession of mine. I just hit this brick wall of "... oh... this is it?".

I do think HTML is a fun standard though. There's so much "big business productization" chained to it. Web companies announce updates to frameworks as if it's an apple keynote now. I think that gives a false impression of the "bones" of the web though. HTML/CSS is still fun! :) Nothing beats the visual-focus and short lag time between code->website in browser.

Is that the Comic Sans I've heard so much about?

"Playpen Sans"

The other day, I was talking to my daughter about where she might like to study in future. We happen to went to a big University for some showcase event and I asked her if she liked that one. She replied, “No! Their logo is in Comic Sans.”

It is. It's all real and it's spectacular...ly ugly.

I do wish we could go back to how things were in the early 00s where view source on any page was extremely telling and would show you cleanly formatted code.

Another huge reason for the status quo is a lot of companies actively try to obfuscate their frontend code for a variety of reasons, and a lot unwittingly do it as part of minification to squeeze out a little bit of extra efficiency both in terms of payload size and parsing time for clients (every token counts!).

That said, I would love a world where the de-jure frontend syntax was less ambiguous to the point where minification is essentially a lossless operation other than the actual names of things.

At the browser level we have to make the decision: do we want clients to be able to figure exactly what is running in their browser? If the answer is no, the current situation is great, if the answer is yes, the current situation is pretty bleak, and will actually get bleaker with the advent of WASM-based payloads where we now need to disassemble on top of everything else.

to the point where minification is essentially a lossless operation other than the actual names of things.

Is that not the case now? I don't do a lot with javascript myself, but I always assumed the minification process wasn't changing the code itself.

It will shorten variable names, replace true with !0 and many other things. Newer technology might even unroll loops to improve performance but that's a bit different.

view source on any page was extremely telling

Yeah! This was basically how I taught myself to write effective HTML. Great fun!

I'm guessing Cheap does not count for the blog post with no included image weighting over 1MB....... 315KB for HTML and looks like your "favicon.ico" is literally a copy of the webpage...

The images are embedded into the CSS.

I mean the blog post itself has no images.

Cheap to maintain: Most webpages should work indefinitely without falling over.

Simple HTML will do this, so long as a browser that supports that version of HTML and CSS is still around. A static Go binary serving dynamic HTML will do this. I struggle to imagine Python, Ruby, etc being able to accomplish this though. That's to say, I'm not sure that's a good rule or could use some reworking to make it more achievable.

Once upon a time side bars and tool bars were made up of iframes that could talk to each other. Those sites wouldn't work today for "security reasons"

If they have the same origin, then they should still work. But there's always postMessage[1].

[1]: https://developer.mozilla.org/en-US/docs/Web/API/Window/post...

The link to Marginalia in the "Explore" sidebar should be "marginalia.nu", unless this is some alternative Marginalia.

Thanks! Fixed

The font (because of the drop shadows I guess) looks jarring in dark mode.

I'm a bit conflicted, there's some really good points mixed in with some odd-ball hurdles that I don't think anyone benefits from jumping over.

Cheap to maintain: Most webpages should work indefinitely without falling over.

YES, holy crap this is what I find most annoying about some of the things I've made recently. Moving between computers, bringing over old projects, and of course dependancies break down and rot. Why can't I start this Gatsby 3 app? Who knows, the install fails. Aggravating.

Cheap to leave: Opting-out of the web should be painless.

The age of total anonymity on the web is gone I think. Even if you try to be totally anonymous, it's pretty advantageous to have successful online media tied to your name. That makes it pretty hard to opt-out. It's a good ideal to have though!

Cheap to access: Most websites should be compatible with screenreaders, etc.

Machine readable information is a shallow measure of accessibility. Just because everything is labeled doesn't mean your website is usable. It's a whole interface you need to consider. I would reword as "Most websites should be usable with only a screenreader" or "assistive technology".

Cheap to participate: Interacting with the web should be possible on a Wii.

... what? An old, outdated, version of Opera?

Cheap to explore: Exploring the web should be pleasant on 1W of power.

Why are we measuring power draw? Yeah your website can be intense enough to draw more power, but this just seems like an odd metric.

Cheap to contribute: Making/hosting websites should be easier than scrapbooking.

Yeah, squarespace, wix, wordpress, etc... facebook?

> Cheap to explore: Exploring the web should be pleasant on 1W of power.

Why are we measuring power draw? Yeah your website can be intense enough to draw more power, but this just seems like an odd metric.

Regardless how you'd measure it, a pretty useful metric imho. A nice catch-all encompassing:

Complexity of a website + efficiency of common software implementations + efficiency of common hardware platforms.

Any content that requires a 'beefy' setup on client side to view (or consumes a lot of power doing so), automatically cuts a large chunk of potential audience.

That is a fair point, maybe I'm thinking to concretely about it. Was imagining getting a kill-a-watt power meter, and making sure my OS + browser + a website is drawing less than 1W from the wall.

What you're saying as a more all-encompassing "efficiency in general" metric makes sense.

I love this. I've seen a lot of similar "use HTML as HTML" type things recently (probably a lot coming from the HATEOAS crowd).

I'm not a web developer, and I really don't understand how we've ended up in this state where:

- web starts as a means to share documents, HTML is built around that

- eventually people want to build general applications (a la google docs) so they build tools that let you build non-document things by pretending they're documents

- everyone thinks those tools are great and starts using them

- 90% of the web is still documents, but now its all built in frameworks where you build a non-document by pretending its a document, even though in actuality, we're normally just building a document in the first place.

The whole thing is a batshit crazy mess and I don't understand how as a global engineering culture we view it as anything other than completely insane.

The platform isn't that much of a mess imo. I think if you look at individual web application code bases they're kind of a mess because the penalty for doing it wrong isn't that high compared to real physical engineering just throwing the business thing in without planning actually does have a higher ROI. If anything the 30 year old legacy code bases are still much worse than the typical React application and services tangle you get today.

HATEOAS = Hypermedia as the Engine of Application State (in case this saves a search).

The JavaScript tools for building web applications are nice when used correctly. I can't imagine the product I work on would be easier to build with HTMX, but perhaps I'm ignorant. I wonder though, are there more enjoyable, non-web UI engines out there? Given that web app UI was glued together on top of a document sharing platform, it makes me think native UI dev should be more coherent.

This is the standard tirade^W manifesto about how web users forced corporations to ruin the Web.

Oh no. Why did you patron Walmart when you could have gone to your local mom&pops. That but for the Web. Petite bourgeoisie sentiment.

What the hell does "petite bourgeoisie sentiment" even mean?

A sentiment coming from the petite bourgeoisie or those of that viewpoint.

Reading through I got this:

- Cheap = a pain to read

Good and modern visual design exists explicitly to deliver information to a person in the most accessible way. If you show a Japanese website to someone who grew up in Ireland, they will be overwhelmed. Not catering to different cultures, like the website suggests - only makes it worse.

- Cheap = a lot of spam

Having low cost of equipment and almost unlimited anonymity, makes fighting against spamming or "enshittification" exponentially hard. Just ask anyone who bothered to run their own mail server or open forum board. It's just not feasible for us to be able to have a nice experience without a level of trust.

"Good" modern visual design barely exists at all according to that definition. Most modern web design exists to jam marketing in your eyeballs while drip-feeding information at the minimum rate to keep you from closing the tab, which is basically the opposite of "delivering information in the most accessible way".

Also: The YouTube homepage uses twice the memory of the NicoNico homepage to show fewer thumbnails, and has basically been unchanged for years, while YouTube can barely resist keeping their layouts the same for a single month. Japanese web design has quite a few issues with being stuck in the 2000s but I'll take consistently bad over rolling the A/B testing dice every week to see what page I get.

I haven't provided any definitions on what is a good design at all, so you'll have to tell me what it is.

Your issue with the content, doesn't negate the designs. Japanese websites are also filled with a lot of junk content.

As for the Japanese websites, they aren't bad. They are bad for you, maybe... they're very much good for people accustomed to Japanese culture.

This is great! I can't really endorse most of the web "innovations" that have happened since the Dot Bomb. There was a time before FAANG when people with low technical knowhow could build something on the web and start earning enough residual income to pay their rent. Like the eBay store on The 40-Year-Old Virgin. Or the Mutiny BBS before that in Halt and Catch Fire. I did it with my business partner via shareware in the early 2000s. Even though the writing was on the wall even then that this was all going corporate.

I think that things started going wrong when the anti-intellectual/anti-education movement gained prominence after 9/11 in the push towards privatization and outsourcing. Before that, we could rely on publicly-funded academia to deliver techniques designed from first principles to decrease our workload.

But since then, we seem to get leftovers like endless Javascript build pipelines handed down to us from the private sector. It's almost like best practices today are designed to mire a startup in endless red tape. While multinational corporations just throw money at it to create a billion dollar pair of scissors that we can't afford.

The way to undo all of that is to do the opposite. Realize that our freedom and prosperity start with our tools and techniques. Find the underemployed exhausted people and give them any resources at all to design better stuff and then leave them alone. Start funding bottom-up and middle-out policies instead of waiting around for trickle-down economics to toss us more scraps. Don't let any one person become a cult of personality preaching how things should be. And stop worshipping capital and focus on getting actual resources (the most important being time) to the middle class through automation and recycling to avoid war. This stuff is so obvious and evident in history that the main challenge is to unlearn one's own programming to be able to perceive alternatives.

Edit: we're talking about HTML not macroeconomic policy. But after spending my entire career having to do things the "easy" way because there's never any time or budget to do things the simple way, I view the complexities of the modern web and the barriers standing in the way of our self-actualization as one and the same.

Edit 2: think scholarships, grants and UBI - not loans, investments and contests.

There's a shift though, just the last couple of years. Today all browsers support native modals with the dialog element, accordions with details/summary, everything but Firefox supports popovers (dropdowns, tooltips, menus, etc) in plain HTML.

I'd honestly love to be a beginner again learning modern HTML and CSS today, it's not bad at all.

I get the spirit of the article, but disagree with some of the specifics.

HTML remains hostile to beginners.

What? HTML is dirt simple. There are some fiddly bits, but if you are creating simple documents (like this one) you don't need them — just use the subset of HTML that you are comfortable with.

HTML is generally unkind to people of limited hearing, vision, etc.

Only if you mess it up. Clean and simple markup is usually pretty accessible.

But really, this page is 1.4 MB of code for less than 20 KB of text (300 ms to load!). You want the 'cheap' and simple web, then set an example.

I agree with your disagreements. I am really bad with any coding, I have tried to learn Python and CSS several times now and can't remember a single thing. At the age of 12 I was easily able to learn HTML to build functional websites. Despite not having written a website in HTML for over 15 years I am still confident that I would be able to build a website using HTML code only right now.

How is that hostile to beginners when it was the only language that I could ever grasp? Are there coding language out there easier than HTML?

I've had this idea for something in between Gemini and the modern web that's essentially supposed to capture how the web was in the late 90s/early 00s. Basically just HTML/CSS with GET and POST, no scripting at all.

I don't have the expertise to write this, unfortunately.

That's literally just "writing basic HTML and not using Javascript."

You can still totally just do that, no one is stopping you.

But even if you went the whole Gemini route and published a custom protocol to be explicitly incompatable with HTTP and a spec supporting only early HTML, you still wouldn't be able to capture the early web, because it isn't the 1990s anymore.

What is a slippy mindset? what does slippy mean in this context? is it an acronym or something? I couldn't find anything about it.

There's a link from the page to here[0] and then out to here[1], which says...

The tadi web is built with a slippy mindset. We try to be as ‘slippy’ as possible.

Being slippy means you’re not stuck.

When you’re slippy, it’s easy to change plan, or rebuild something from scratch. It means you’re not locked in to using a certain tool.

If something breaks, you can choose to fix it, or let it die. It’s ok, because it won’t take long to rebuild it from scratch.

Every time you grow back, you’ll be a bit different. You might be stronger. Or you might be better equipped for your changing needs.

[0]: https://www.todepond.com/wikiblogarden/tadi-web/ [1]: https://www.tadiweb.com/

Absolutely lovely read!! Honestly feels incredible to see other people feel the things I do when most people don't, don't care and don't want to care.

Might be a nitpick, but the point is ruined by the god awful aesthetics of the page

Weirdly I loved the aesthetics of the page. I know they are probably pretty terrible from a conventional design standpoint, but they're so full of personality it was a joy to look at!

What a breath of fresh air is browsing simple websites, where content in mainly text and images are sparingly used. All "modern" websites are so tiring. Now if website pops up anything on screen on first visit, be it newsletter signup, login request or just huge image covering whole viewport, I just switch to reader view.

If I can't close it immediately and resume I right-click->Block Element. If doing that leaves the page unusable, I usually just click off. I'm with you, the web today is exhausting and feels hostile to the user in most cases.

Did anybody else's eyes get freaked out by the blog?

As software rots, multinationals may become the only players capable of making websites.

They are already the decision makers of who open source software is for.

The thing people fail most at realizing is that you can't have a lot of rich people with a lot of power without having a lot of poor people with no power. The interests of the former will always effectively undermine the latter.

Similarly, you can't have software that serves multinationals and regular people at the same time because the interests of the former will always effectively undermine the latter.

Unfortunately, the HTML source for Apple.com is not so "beautiful on the inside", but Apple engineers shouldn't be faulted for ugly HTML. Their only option was to wrap a sleek skin around shoddy materials.

Can ChatGPT et al make good beautiful HTML/CSS pages? Or take an existing one and actually beautify it?

The article pulls together a lot of lovely material and links and ideas, but it starts with such a toxic & conservative outlook.

Sure, apple.com's source might not be as beautiful as their products but potato.cheap immediately turns around and says it's not apples fault, and trashes the web and says it needs sto be replaces by something else.

It just seems to be like the industrial world isn't going to improve without being lead there, but that there's so much leeway for the better awesome web to evince itself. The manifest possibilities are what so many people see as a bad and awful thing, but I think taste & discernment & internal structure are incredibly complicated & any kind of broad scale human hyper-engine of note is going to take more than a decade of effort for us to really suss out where to go & what looks & feels right. There isn't going to be the perfect system created out of nothing that makes everything great; it's power flexibility and possibility that we refine &harness onward to get to exceptional, and the web has ongoingly been an amazing host to that dialog in ways nothing else has been or from what I can see could be.

I'm so tired of those who most strongly clearly advocate for better online systems being so retrograde. I miss the hope & buoyancy of those excited for stuff! Web Components and webrtc and QUIC and moq, our new awesome css flexbox and grid and layers, css variables. I'm still as thirsty and in love & still a believer that we can keep honing new & great. To shock industry into not sucking we just need to lead towards better, towards protocols & interoperation & elegance personally. The virtue will come out, and will be aped, as we develop, if we dare be better.

The web is already fast and performant on all devices from the last five years, and does not seem to be getting slower. Moore's law and optimizations caught up, performance is good enough for me, I won't complain as long as nobody messes it up.

I do absolutely agree we need more accessibility and more sites that aren't endless scrolling trash. I don't think we need old fashioned simple HTML to do it, although that's certainly one of the tools that can be used.

I think we mostly need more things like neocities that make it actually easy and maintenance free to set up a site. To me the important part of the old web was the content and the people, not the tech.

cs professors are happy with this. this is what their "personal site" envisions