@meb has joined the channel
@ryanc Is it possible to fix the stxparse tests soon? currently all of our CI except DrDr is failing because of that, which limits our test coverage as a whole and makes it hard to tell if we regress things
@samth should be fixed now
@ryanc great
why does a plumber flush callback receive its flush handle as an argument? the docs say it’s so it can conveniently unregister itself, but why would it need to do that? is it to guard against the plumber being flushed more than once, or to deal with the case where flushing is unsuccessful and should be retried?
Yes, helping to avoid another flush was the intent. But I’m often unsure whether it’s better to make a callback receive a relevant argument or have the callback be a closure as needed.
is there a reason flushing the callbacks doesn’t automatically unregister them?
test-log is documented as displaying by default, but that’s not the case. I need to use (void (test-log #:display? #t)) to get actual output. cc/ @notjack (since you’re in there a lot of late)
@zenspider I think the docs are just wrong there. Based on the git history those two arguments have always defaulted to #f and been documented as defaulting to #t
doc patches welcome :)