While reading this post on what we lost with the abandonment of XML this passage caught my Lisper eye:
XML's structure is immediately familiar. It is essentially s-expressions with angle brackets instead of parentheses. An element is a tagged list; attributes are metadata; nesting is composition.
@amoroso iirc Henry Baker had some presentation at an Internal Lisp Conference about how Lisp dropped the ball on XML. There was also XMLisp includes a custom reader to read XML tags directly into classes
@amoroso Found the slides
https://international-lisp-conference.org/2005/speakers.html#henry_baker
https://international-lisp-conference.org/2005/media/baker-slides.pdf
```
The “XML Question”
• By rights, Lisp should own XML
• Lisp should immediately embrace XML
• Lisp needs to quickly develop standard
XML readers & printers
• Lisp needs to utilize XML as alternate
syntax
```
@PuercoPop It's interesting as it covers quite a lot, thanks.
@amoroso XMLisp is really cool as well, I remember finding the code somewhere, but I could only find a paper right now. Today it might be useful for parsing XML wayland descriptions directly 🤔
@amoroso XML with all its components and standards around has always been a parallel universe. XML has been widely misused for purposes that shouild have never seen any XML at all. In addition to that, techniques like XSLT or XPath have been widely considered as academic and non-approachable. XML knowledge has always been dominion knowledge. I've been to several XML conference (XML Prague, #XML London,Markup UK) and the high academic speaks for itself. XML & Co. appeared stimes like brainfuck.
@zopyx Yes, the post acknowledges this at least in part.
@amoroso This is what I've always said. XML is simply a prolix way of writing S-Expressions; toolkits like @weavejester's Hiccup, which translate between #Clojure and #XML representations of S-Exprs, make this very clear.
In my experience, this makes XML (and XSL) very comfortable to work with for Lispers.
@simon_brooke With proper indentation and enough screen real estate XML doesn't look too bad.
@amoroso I honestly mostly agree with this article, having now read it in full. I don't love the textual representation XML format, but, in particular, when I wrote things that used XML data exchange, I _always_ wrote a schema, and that schema frequently caught bugs (both in my understanding of the data being exchanged, and in the software doing the exchange). There were tools to write very short, readable schema for simple data (like RELAX NG).
@amoroso I actually like sxml for encoding HTML documents reasonably well. The attribute syntax is a little annoying and I often get it wrong (there's nothing wrong with it and it's a reasonable encoding, but for whatever reason I often miss parens or such), but it's much less verbose and it plays nice with lispy editing habits.
@elb I never used XML much but of course sexp-based notations come natural to me.
@amoroso XML was not without its detractors back in the day - I am minded of Erik Naggum's famous post https://www.schnada.de/grapt/eriknaggum-xmlrant.html
@dan Yes, I remember he wasn't a fan of XML.
