pedestal

hlship 2025-03-28T00:05:50.606439Z

0.8.0-alpha-1 • Fixed the double-called handler function bug • Tweaked a bunch of things related to testing with a Pedestal connector • Updated the embedded template to use http-kit • Updated the embedded template documentation Just pushed it all up to clojars.

💯 3
🎉 8
lread 2025-03-28T19:09:51.708389Z

Trying it out for cljdoc. I think maybe I'll need to study the new routing!

lread 2025-03-29T21:06:32.781209Z

Yep. Cljdoc is doing something custom for routing (not sure why, before my time). I will report back if I figure out what I need to change for 0.8 compat.

hlship 2025-03-29T22:06:31.409019Z

Routing assets vs. endpoints is somewhat complicated; I prefer using routing table for both but that isn’t always practical, which is why still supporting doing it with interceptors. It does make me wonder if we need a way for a router to prefer static over dynamic when routing (rather than making it a collision, the way sawtooth does).

lread 2025-03-29T22:18:30.677969Z

I can probably clean this up a bit more... to conform to the way pedestal works now... not sure.

lread 2025-03-29T22:22:10.159409Z

The PR for cljdoc upgrade to pedestal 0.8 is minimal: https://github.com/cljdoc/cljdoc/pull/1012, so thanks & congrats on all your hard work! Cljdoc tests use the io.pedestal.test ns, which, after your rewrite, continues to work great!

1
lread 2025-03-29T22:30:45.413109Z

I guess I should wait for at least a beta release before pushing to production, eh? simple_smile

hlship 2025-03-29T22:31:12.662229Z

YOLO

1
hlship 2025-03-30T01:01:05.709119Z

Or you could go big and switch to io.pedestal.connector and io.pedestal.http.http-kit.

lread 2025-03-30T02:26:46.732289Z

Oh! That would be going big and small!