Hacker News Reader

Comment by gred | Hacker News Reader

This is a capstone of sorts to a process begun by the W3C's XHTML focus 20 years ago. Once again, incremental improvements to a flawed technology (HTML) were more successful than a perfectionist rewrite (XHTML). We so often ignore or downplay costs associated with adoption and migration when we build our "castles in the air".

> The programmer, like the poet, works only slightly removed from pure thought-stuff. He builds his castles in the air, from air, creating by exertion of the imagination. Few media of creation are so flexible, so easy to polish and rework, so readily capable of realizing grand conceptual structures. (Fred Brooks)

The W3C HTML working group was even further removed from users, imagining castles but then delegating the building to others.

View on HN

You're not wrong. The W3C was focused on DTDs and semantic markup and whether HTML "validated" or not. They were focused on moving folks from "transitional" XHTML to "strict" XHTML with no perceivable benefit. Meanwhile the WHATWG was interested in things like media tags and making interactive content be things other than Flash and Silverlight and Java.

They got too lost in the sauce with what was essentially an academic exercise that didn't really benefit the end user.

> The W3C was focused on DTDs

No, imposing structure onto markup using DTDs had always been a useful and elegant technique.

What W3C was busy with after subsetting XML from SGML as a canonical subset not requiring markup declarations/DTDs for mere parsing (which while useful was already possible and practiced using SGML profiles) is inventing ill-conceived modularization and meta conventions (namespaces, xsi, xinclude), entirely new schema and transformation/query languages, unrealistic and unproven design-by-comittee formatting and browser interactivity systems, and enormous enterprise messaging system ecosystems.

And yes, semantic web stuff; graph DBs were a love affair of TBL's predating even the web (cf. https://en.wikipedia.org/wiki/ENQUIRE).

> a useful and elegant technique

Maybe, but who does it benefit? What is the outcome for people using the web? How were the developers making websites benefiting? It wasn't a bad effort but it was focusing on something that didn't actually solve the biggest problems with the web. XHTML wasn't built for anyone real, it was built for an idealistic outcome.

Private companies (Adobe, Sun, Microsoft, Real, Apple) had a literal monopoly on video and animation on the Web. A lack of interface between JavaScript and the DOM led to companies like Microsoft freely inventing their own stuff.

If you want to be the steward of a technology, you have to care about the way people are actually using the technology. Otherwise someone else who actually does care will (and did!) come along and do something meaningful.

Users being Google.

Users being anyone using the web. HTML 5's initial draft, which was the beginning of the end for the W3C, was published in January 2008. Chrome was released in September 2008.

HTML 5 became a W3C recommendation in 2014, when Chrome had less than a 40% market share.

I think you meant ChromeOS.

No I don't. Google didn't even publish the Chrome OS source until the end of 2009, and it wasn't a consumer product until June 2011.

In 2025 it is going definitely to be ChromeOS.

Web is driven by Google nowadays, so a standards body that only puts a signature into whatever Google decides, is naturally worthless.

The whole point is that the W3C has been impotent when it comes to HTML for a decade now. They're not publishing anything Google wants in that domain because they're not publishing anything. HTML was taken out of the W3C's control because of their mismanagement it well before Google had a browser majority (let alone a monopoly).

Because browser vendors refused to implement XHTML properly, and then Mozzilla folks pushed for HTML5 outside W3C, Web freedom, what an irony than now Google plays Simon says.

That's false. Browser vendors implemented XHTML so the Web wasn't a broken mess. You can't force developers to write perfect markup. XHTML had lots of undefined behavior and bad failure modes. Implementing XHTML "properly" meant breaking the web, but that's not what the W3C cared about, they cared about standards for standards sake.

Let me be clear: XHTML didn't materially benefit anyone at all except the people who were putting their names on a standard. HTML 5 actually fixed real problems. It made parsing HTML sensible and exhaustively defined. It added open features that companies like Macromedia, Apple, Real, Sun, and Microsoft had a monopoly on for a decade prior. Unlike the W3C, the WHATWG made actual, real life people's lives meaningfully better.

Moreover, most of the things Google is pushing for aren't HTML. Protected Audiences—the ad targeting API Google wants—is sitting in front of the W3C folks as part of the WICG.

Developers will do whatever browser vendors push them to do, and they clearly decided they did want to push HTML5 agenda instead.

Now, Firefox is for all purposes dead, just fairing a bit better than The Year of Linux Desktop, Safari with Apple doing its own thing thanks iDevices market share, and everyone else is basically a reskinned Chrome, or Electron garbage, solidifying ChromeOS as the future we all deserve to have, as the IE lesson is long forgotten.

Reply on news.ycombinator.com