<daviid>yep, 18000 commit in 5y! I 'pre-announced' here (that they would release 4.0 in 10 days - and they did :)) - https://blog.gtk.org/2020/12/16/gtk-4-0/ - the blog image in the annoucement is ... very 'clutter like', it feels like a clutter release :):) (which is perfect, all we need is clutter :):) ***apteryx is now known as Guest16789
***apteryx_ is now known as apteryx
<rlb>hmm, is there a way to "rethrow" an exception caught by catch, preserving the original backtrace? <jmarciano>which module can process skribe markup on command line, is it skribilo? <wingo>civodul: idea: wdyt about "load-extension" causing a module to be marked as non-declarative <wingo>or causing a warning, perhaps <civodul>wingo: you think extensions typically rely on being able to access the module's top-level bindings? <civodul>i don't remember encountering issues of that sort, but i can imagine it happening <wingo>dunno, i was thinking about reasoning about top-level bindings; if there are definitions the compiler can't see, it doesn't know if any given reference is a local definition provided by load-extension or something, or an import <wingo>context is that i have a rewritten pass that combines macro-use-before-definition and unbound-variable analysis, adding detection for those weird (define old-+ +) (define + -) cases <wingo>and it would be nice if it could reason more precisely <civodul>yeah in a general sense, the compiler doesn't see the big picture when load-extension is used <civodul>so it'd be safer to make modules non-declarative in that case <mwette>extensions should convert volatile top-level bindings to fluids? <mwette>I just upgraded to ubuntu 2.10. Packages include "guile-3.0" and "nyacc". <rlb>RhodiumToad: ahh, thanks. I'll take a look.