IRC channel logs
2026-10-21.log
back to list of logs
<jfred>but I'm getting ahead of myself there, still learning my way around it haha <dthompson>the next error is in srfi-13. it's non-declarative. it has a re-export form outside the define-module form. so the fix is to add a srfi-13 to hoot that does it the "right" way <jlicht>dthompson: would fixing this on guile's side not be the "better" way? <jlicht>having not looked at any of the code yet, fwiw <dthompson>we could fix it in guile, too, but fixing it in hoot would be the fastest way <dthompson>I haven't had luck getting anything into guile lately <dthompson>hopefully guile can be convinced that it's better to be declarative than not so even though the change is not strictly necessary for guile it's beneficial <jlicht>doesn't it allow for better/more optimisations as well? <jlicht>at least that was the meme w.r.t. "sufficiently advanced compilers" back in uni :-) <dthompson>I don't know if it works out that way in this specific case but maybe <dthompson>jlicht: I'm seeing if I can quickly make a srfi-13 module for hoot <jlicht>dthompson: is there some reading material about how code expansion works on hoot? I only know how the Clojure/ClojureScript folks do it, but would love to learn more <dthompson>we use guile's macroexpander, if that's what you're wondering <dthompson>jlicht: I pushed a hacky srfi-13. the next issue is that (sxml ssax) has a macro that calls %search-load-path which hoot doesn't implement <dthompson>maybe we can implement it for expansion time code <jlicht>dthompson: yeah that's why I asked ;-)