<headius>
I want to accept a PR to add a missing bit, but it's in Ruby so it shows up in :return traces rather than :c_return
<headius>
no idea if that matters
<enebo>
headius: seems semantically broken by design no?
<enebo>
headius: I mean it is correct to show something impld in Ruby to be :return but yet if we don't match MRI then a tool may complain
<enebo>
perhaps though RT has ways of emulating this but I question whether that is true to what this API is for
rdubya has joined #jruby
<rdubya>
enebo: Saw your note on the PR
<enebo>
rdubya: I know kares looked at that PR and did not seem to say anything major but if there is something we can back it out
<enebo>
rdubya: he is also on this channel so he may pop in and say something too (although he is 7 hours later so maybe not too)
<enebo>
rdubya: In finder_test: def test_hash_condition_local_time_interpolation_with_default_timezone_utc
<enebo>
it gets a AR object and asks for a Time object and then calls getlocal on that and stuffs that into a where clause to try and refind the same instance
yahonda has joined #jruby
<enebo>
but the find fails because the local version of the Time instance cannot find the time which I am guessing is stored as UTC in db
<enebo>
I was curious if you pass that test or not with postgresql
<rdubya>
yeah I saw that and meant to check it out, at this point the only time test failing for postgres is the one dealing with BC dates/times
<chrisseaton>
headius: yes we have TracePoint
<chrisseaton>
It's not harder than STF
<chrisseaton>
But I'm not sure we have full coverage on all the events
<rdubya>
enebo: I had to override how those were handled for sqlite in prepared statements, so those overrides may need to handle a wider range of values
olle has joined #jruby
<enebo>
chrisseaton: but you have some things MRI would c_return as return since it is in Ruby right?
<enebo>
chrisseaton: and more overall events because impl is also Ruby
<enebo>
rdubya: both prepared and non-prepared seem to fail for this test so I think maybe I need to figure out more about how this works
olle has quit [Quit: olle]
<rdubya>
enebo: ah ok
<enebo>
rdubya: ah I see something perhaps
<enebo>
rdubya: I use AR quoted_date and you still provide an impl for it
<enebo>
not saying you should not have your own or not but that is an obvious difference
rdubya has quit [Read error: Connection reset by peer]
rdubya has joined #jruby
<rdubya>
enebo: I've actually had that commented out in the postgres adapter for a while
<enebo>
rdubya: ah cool
<rdubya>
just haven't committed it yet because I was focused on the prepared statement stuff
<enebo>
rdubya: so perhaps not the issue
<rdubya>
i have a meeting but i'll be back in an hour or so