|
|
Subscribe / Log in / New account

XSLT is kinda of horrible

XSLT is kinda of horrible

Posted Aug 27, 2025 23:12 UTC (Wed) by roc (subscriber, #30627)
Parent article: The tangled web of XSLT browser support

XSLT is awful to use. Using XML as a syntax for stylesheets and code is nasty. Also, ultimately you want to write transforms in a real programming language, and extending XSLT into more of a real programming language sounds horrendous.

Of course, being technically horrible isn't the only thing that matters, but it partly explains why XSLT isn't much used and why browser vendors aren't more enthusiastic.


to post comments

XSLT is kinda of horrible

Posted Aug 27, 2025 23:46 UTC (Wed) by basscomm (subscriber, #125799) [Link]

I see this sentiment a lot and I have to admit that I don't understand it all. I admit that I'm some kind of outlier here, but I've been been refreshing my XSLT over the past few years to convert some XML to HTML on a hobby site, and I don't find it particularly hard to understand or use. Sure, that opinion might change if I ended up doing anything more complex, but for basic stuff, it's really not that bad.

XSLT is kinda of horrible

Posted Aug 28, 2025 1:22 UTC (Thu) by raven667 (subscriber, #5198) [Link] (1 responses)

The aesthetics aren't really part of the argument, XSLT is an open standard with multiple implementations which exists, and is built for a different vision of the web built out of information sharing than the dominant advertising and surveillance supported one built by big tech companies. Using XSLT to build a web view client-side means that the client has access to the rich XML data too, with the great example of RSS, which is not tied to any particular vendors software stack (get your podcasts *wherever* you get them), which is a very radical feature given how much of the modern web is driven by FAANG, which uses web technologies, like JSON and JavaScript, but tends to be locked in bespoke APIs and difficult to maintain frameworks which limit integration to large companies which can afford it. This isn't a technical argument about the merits of XSLT, it's mourning of a way of being. RETVRN ;-) There is some technical meat though, if the XSLT engine had been maintained, if Apple and Google or MS had thrown a few FTE to the project then maybe it could have been updated to XSLT v2 and v3 which supports templating JSON data as well, and that have might lead to a very different kind of design for PWA and SPA than what React (Facebook) and Angular (Google) evangelized, none of which works toward the strengths of the browser engine or the web as an information-sharing platform, SPAs tend to obfuscate the underlying information and do not lend themselves to custom user-agent behavior in the way RSS does.

This whole argument is difficult to articulate because it's an emotional, vision and alignment discussion and not primarily a technical one

No REST for the wicked

Posted Aug 28, 2025 7:19 UTC (Thu) by tux3 (subscriber, #101245) [Link]

There isn't exactly a bijection between XML and JSON, but I don't find that the way it's rendered makes that much of a difference on access to the raw data

I usually have little trouble getting raw data in the modern JSON/Javascript Zeitgeist. And it'd be writing very similar Python scrapping snippets in an XML/Javascript or XML/XSLT world

Either way if I'm going to be doing anything interesting with the data, it won't be by manually copying bits of data from my browser's debug view, I'll be calling some bespoke API. If anything, the olden XML APIs I'm forced to use sometimes have this distinctive flavor of over abstraction and unnecessary complexity.

Data's data. The presentation layer doesn't really change access, in my experience


Copyright © 2025, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds