News for hackers. Stuff that matters.
Re:The Page with No Code by asim on Saturday January 21, 2023 @09:04AM (#34466456)
Can I just say. I wish web pages were actually programmable. Every time I put together a website now I think damn, look at all this content stuck inside static html. I wish it was in a database and loaded from there, even if that database was a simple text file or whatever else. The web needs some evolution that isn't react and next.js. Going back to pure text is killer.
Re:The Page with No Code by est on Saturday January 21, 2023 @09:16AM (#34466556)
XSLT?
Re:The Page with No Code by still_grokking on Saturday January 21, 2023 @01:02PM (#34468508)
Frankly HTML5 isn't XML any more. Very bad move…
Re:The Page with No Code by andai on Saturday January 21, 2023 @01:16PM (#34468658)
Does anyone know why XHTML failed? I remember deriving some satisfaction from making my website XHTML compliant back in the day.

Wouldn't it make it much easier to parse? Or does the motivation to "make stuff still try to work even when it's clearly broken" override that? (The same motivation gave us everything that's wrong with JavaScript, so I'm not sure I agree with it...)

Re:The Page with No Code by chc on Saturday January 21, 2023 @02:01PM (#34469145)
XHTML was not backwards compatible and was harder to write for essentially no benefit to most people. "It's XML" was basically the entire list of selling points, and that was just not a compelling enough proposition — in fact, for many people, that was negative value.
Re:The Page with No Code by still_grokking on Saturday January 21, 2023 @02:57PM (#34469733)
> XHTML was not backwards compatible

Sure?

I think it was HTML4 that was not forwards compatible.

XML is a subset of SGML, and HTML4 was SGML. XHTML was / is XML. So HTML4 parsers (as being SGML parsers) shouldn't have issues with XHTML.

It's the other way around: XML parsers don't like HTML. So the attempt failed to change the default parsers to XML; as old webpages had than issues, and the web crowd didn't like to fix that. People wanted to continue to use the horrible SGML mess—only because nobody wanted to touch their HTML4.

So Google created their own web "standards council" and ignored the W3C henceforth. The rest is history.

Re:The Page with No Code by still_grokking on Saturday January 21, 2023 @02:47PM (#34469624)
You can still write HTML in XML syntax. It's not wrong afaik.

But HTML can be also written in some absolute quirky way that isn't even SGML. Just a horrible mess!

Re:The Page with No Code by cableshaft on Saturday January 21, 2023 @01:22PM (#34468726)
Have you tried coding XSLT before? I had to for one job, it was a cool idea, but extremely difficult to do anything, and very restrictive.

As much as I wanted it to work out (and tried making a personal website using it for a little while), I would never go back to it today.

Re:The Page with No Code by avaika on Saturday January 21, 2023 @10:43AM (#34467150)
Sounds like messengers with one to many communication feature (e.g. Telegram) fit this requirement at least partially.

It might be not as interactive as mostly of websites, but it allows you to get data through API or develop own client.

Re:The Page with No Code by simonw on Saturday January 21, 2023 @11:18AM (#34467454)
Web pages are programmable - way more so than desktop or mobile apps.

You can open up the DevTools and start poking around - including running your own JS against the page DOM using the console.

You can write bookmarklets, or user scripts, or even fill browser extensions that run your own custom scripts.

You can even run a headless browser to automate this stuff further - a trick I use a bunch with my shot-scraper CLI tool, see https://simonwillison.net/2022/Mar/14/scraping-web-pages-sho... and https://til.simonwillison.net/shot-scraper/scraping-flourish

Re:The Page with No Code by ksec on Saturday January 21, 2023 @11:25PM (#34474268)
Basically remaking something that replaces HTML.

We could have tested those idea in a similar fashion as React and next.js in Javascript, and incorporate what works as native HTML. Before sort of refactoring HTML into something else.

But Google doesn't want that. Google wants everything on the web to be Javascript. So yes. Blame Google.