<flatwhatson>please let me know if you notice something that could be handled better, error parsing is very ad-hoc as guile is not always super consistent, the checker is basically a collection of regexes to parse out useful info from the various forms ***wingo_ is now known as wingo
<civodul>somewhere the right exception handler is ignored <antipode>civodul:I think I've seen that one before ... <antipode>According to the follow-up mail, this is actually documented. <antipode>the 'handler' needs to be run in the right context for backtraces (when using #:unwind? #t) or for the outer parametrisations (when using #:unwind? #false), but fiddling with which handler is run seems weird to me. <antipode>Wasn't there something about Guile's implementation of 'guard' being incorrect because it runs the clauses in the wrong dynamic environment, but a proper implementation would be inefficient (due to it requiring continuations?)? ***nikken_ is now known as nikken
***wklew_ is now known as wklew
***arrdem_ is now known as arrdem
***fluffyballoon_ is now known as fluffyballoon
***ane__ is now known as ane
***\f_ is now known as \f
***unmatched-paren_ is now known as unmatched-paren
***cylb_ is now known as cylb
***reyman_aw_ is now known as reyman_aw
***matta_ is now known as matta
***_ctb_ is now known as _ctb
***jpoiret7 is now known as jpoiret