Mason Freed, a Chromium/Blink developer employed by Google, opened a ticket a couple weeks ago on the WHATWG/html repository: Shall we remove XSLT from the web platform? Typical of the simultaneously high-handed and ignorant approach of the browser vendor cartels that have come to dominate the Web, Freed is proposing that client-side XSLT support be removed from the “web platform” (which is cartel code for “whatever We decide”). Like most attacks on the Open Web, this one is backed by concerns about “security”—as if the only way forward was to keep using libxslt without paying its developers.

Mason’s laughably ignorant proposal of a polyfill is completely unworkable, as such a polyfill would increase payloads by an order of magnitude—and it wouldn’t even work for most use-cases where the goal is to serve an XML document that is automatically rendered, rather than to serve an HTML document that embeds the result of transforming an XML document at load-time. It’s a non-starter. Why do people who know nothing have authority over people who know something? I guess it’s the way of the world.

Now, I understand that XSLT in the browser is not as commonly used as it was in the past. Moreover, real industrial users of XSLT are not using XSLT 1.0 but rather later more fully-featured versions like 3.0 via tools like Saxon, etc. But one very common use of client-side XSLT is to render RSS/atom feeds (for blogs and podcasts—yes, people still do those!) in a user-friendly way when viewed in the browser; of course Forester uses client-side XSLT as well. Since the writing has been on the wall for a while, I have been planning to change Forester to not depend on client-side XSLT, but I would prefer not to be forced into this by the browser cartels.

Many of the people whose comments have not been censored in the GitHub thread seem to be entirely unaware of the fact that RSS even exists. My guess is these viciously ignorant people think that podcasts are things that you subscribe to on Spotify, or watch on YouTube.

Anyway, I was thinking. I wonder if the health and security of the Open Web would be greatly benefited if Mason Freed and his colleagues in the other major browser vendors did not have the ability to change the Web without consent of its actual constituents—the people who publish websites. Using a tiny fraction of these vendors’ resources, we could fund the amazing team building xrust (a modern implementation of XPath 3.1, XQuery 3.1, and XSLT 3.0 in Rust), or any one of several other teams working to implement modern XML tooling in open source. I’d love to restore the “web platform” to its rightful owners—the PEOPLE. I don‘t know if Ladybird is the solution, but I’m glad to see some competition in this space from outside the cartel.

To Mason: I’m sure you’re a fine guy in real life. There is a sense in which you are only a vessel for these proposals, which would be made by someone else if you weren’t involved. But I think there are plenty of people for whom there would be no salary big enough to get them to let the cartel’s words flow through their lips so uncritically. I think those are the kind of people we need in stewardship over the Web.

See also this post by David Bushell.