[0x3b0b]I think I've only ever used XSLT for two things. One of them I don't remember, and the other was for a subset of conversion of WDDX to JSON because I was dealing with a server on which I didn't have sufficient influence to get someone to install whatever it was PHP needed in order to deliver JSON pretty-printed.
[Joe_Crawford]XSLT is never going to be "awesome" but if you are obliged to speak XML it's pretty great. MP3 dot com had what was essentially a static site generator that used XML files for storage in its early days.
[Joe_Crawford]Structured storage for hundreds of thousands of songs, artists, albums, etc and used XSLT in Perl (I want to say Perl::mason) rather than having to buy licenses the only performant database anyone wouldn't sniff at - Oracle.
sebbuwell, xslt is great in the sense that it can be transparent, and you can offer multiple xslt for the same xml document, even using xslt-fo for non-xml/non-html outputs
sebbui used xslt to do some stuff xpath don't allow me to do (like something i could code in a 30 lines python program using 4 or 5 xpath queries and some loops/branchs)
[tantek]alright, I'm thinking today I start drafting a bunch of updates to rel-author, in case folks have thoughts, outstanding issues etc. obviously this touches on many aspects like /authorship, and in particular I want to specify precise enough functionality to bridge HTML post permalinks (regardless of platform/software) and both HTML author pages and JSON AS2 profiles (that presumably have AP inboxes).