<headius>
and the alias_method_chain issue I thought was prepend-related was because it used arjdbc 1.x
<enebo[m]>
Monitor is no doubt a new kwarg supported
<headius>
so it gets farther... Monitor blows up the db:migrate
<headius>
so that's how far
<headius>
I'm finishing up fixes on the prepend change in 3.1 anyway and then I'll fix Monitor
<enebo[m]>
monitor will be interesting loooking that that code
<headius>
probably kwargs for timeout
<enebo[m]>
yieldSpecific is callign with 0 args so it is not monitor at all
<enebo[m]>
oh yeah it will be whatever is in the block
<enebo[m]>
I just saw new specs coming it which exposed :in for some time/date thing
<headius>
lovely
<enebo[m]>
it is likely it is already something in MRI but we are not too far from passing most things
<enebo[m]>
I just found something in warn as well which would trip over kwargs but I will land that soon
<enebo[m]>
as well as fix some more warn issues
<headius>
enebo: I mistakenly blamed Monitor because it was the top of the stack for this particular cropped output... it looks like it's a zsuper issue with anon kwargs
<headius>
took me a while to wrap my head around it but if simply ignores includes when searching for existing prepends
<headius>
(assuming I have interpreted this right)
<enebo[m]>
I think Ruby conceptually gets harder and harder to grok
<enebo[m]>
In ways you don't realize until something does not work (like mixing features together)
<headius>
they also run both include and prepend through the same logic with different inputs, and we have that method split up into pieces for prepend pre-check, do prepend, and do include
<enebo[m]>
yeah that method is weird
<headius>
it fixed two WIP, one in addition to the one from the CRuby issue
<enebo[m]>
like saving duplicating 60% of that method is worth trying to parse the weird duality of it
<headius>
yeah it is so big
<headius>
most of the conditionals only ever affect either includes or prepends too
<enebo[m]>
yeah different subsets of conditions for prepend or include
<enebo[m]>
I would use it as a class example of how you can make something more confusing by sharing it
<headius>
indeed
<enebo[m]>
well my warning work I might table this afternoon and look at the zsuper thing
<headius>
I'm going to hack the rails source to forward this differently so I can get past it for the moment
<enebo[m]>
I will properly write out warnings and pass things I throw at it but I get something messing up with how they are mocking warnings
<headius>
ah yeah
<headius>
that is a little weird
<enebo[m]>
Warnings is some hot garbage (not being in a charitable mood)
<enebo[m]>
anyways it involves needing to call non-dyncalled stuff which ends up dyncalling but then also needs to keep kwargs data along so it can be remade later
<enebo[m]>
It is just weird
<enebo[m]>
headius: so you got it to work by working around or something else?
<headius>
crud works
<headius>
I am relieved
<enebo[m]>
great
<enebo[m]>
fun: snippets/kwargs134.rb
<headius>
yeah fixed by changing that anon kwargs to have a name and forward explicitly
<headius>
in Rails code
<headius>
so if that were fixed Rails 7 can pass the smoke test
<enebo[m]>
yeah I figured it would involve '**'
<headius>
just don't install docs
<headius>
rails server does not start under 19, that's interesting
<headius>
might be some undiscovered issue with native fibers