<lopex>
hmm, this seems to fail [{1 => 2}].to_h{|r| [1, 2]}
ur5us has quit [Ping timeout: 260 seconds]
ur5us has joined #jruby
ur5us has quit [Ping timeout: 260 seconds]
nirvdrum has quit [Ping timeout: 260 seconds]
ur5us has joined #jruby
shellac has joined #jruby
ur5us has quit [Ping timeout: 260 seconds]
ur5us has joined #jruby
drbobbeaty has joined #jruby
_whitelogger has joined #jruby
ur5us has quit [Ping timeout: 260 seconds]
claudiuinberlin has joined #jruby
nirvdrum has joined #jruby
claudiuinberlin has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
claudiuinberlin has joined #jruby
<enebo[m]>
kares: did we delete rails-6 branch?
<enebo[m]>
err sorry rails-5
<enebo[m]>
I sort of remember something about this but someone says they were depending on it in an issue
<kares[m]>
what branch, from AR-JDBC?
<enebo[m]>
kares: yeah
<enebo[m]>
kares: there is an issue open
<enebo[m]>
kares: I mean on one hand I am sympathetic and we maybe should have just kept it but on the other it was a spike branch
subbu is now known as subbu|afk
<rdubya[m]>
Yeah, I saw that, too. I have a copy of the branch if we want to push it back up there
<rdubya[m]>
but I'm wondering why they wouldn't just switch to the released gem
<enebo[m]>
rdubya I say we push it... kares any objection?
<enebo[m]>
I would rather just end this person's problem than probe them on why
xardion has quit [Remote host closed the connection]
xardion has joined #jruby
subbu|afk is now known as subbu
<kares[m]>
well I do not see why not
<kares[m]>
how hard it is for them to use a stable release
<enebo[m]>
kares: if you want to engage them and push back you can otherwise we just push this and close it out with a message encouraging them to
<kares[m]>
if every oss project kept all long term branches after merge ... Branchpocalypse!
<enebo[m]>
kares: yeah but I don't know if we are hitting that point but my only goal here was to not spend time on this...if you want to spend the time then you are welcome to
<kares[m]>
yeah not, had JRuby plans this week but work support nailed me. maybe later, I'll check the issue ...
shellac has quit [Quit: Computer has gone to sleep.]
<headius[m]>
enebo: load service branch merge went well and it's green again
<headius[m]>
I am tempted to just go ahead and merge now and continue working on the remaining specs on master
<headius[m]>
it fixes dozens of things but there's a few critical ones tagged I will deal with, specifically around some parallel autoload + require cases
subbu is now known as subbu|lunch
<enebo[m]>
headius: ship it!
<enebo[m]>
headius: we may as well work through any issues we find just to get it on the board
<headius[m]>
zeitwerk has 4 failures with this branch
<headius[m]>
55 failures on master
<headius[m]>
one of the key remaining issues is atomicity of require and autoload locks... as in CRuby there's two separate tables so I need to unify them
<enebo[m]>
I don't recall but is require/load faster now?
<headius[m]>
well I did enough optimization to get a few key cases back to equal
<headius[m]>
I did not go further than that
<enebo[m]>
ok. I thought somehow MRI did end up faster at some point from caching
<headius[m]>
yeah they do but this impl is still half port
<headius[m]>
there's more work to improve it and speed things up but it's doing the right bits now
<headius[m]>
I had to port a lot of code but also keep existing logic for jar files and stuff working, so that's probably not as efficient as it could be
<headius[m]>
a big part of the perf improvements is maintaining an expanded list of load path entries, which is in there
shellac has joined #jruby
subbu|lunch is now known as subbu
<headius[m]>
ah yes I remember the race issue now
<headius[m]>
CRuby checks both the loaded features table and the require locks when searching... if we find something in the former of course it's already loaded, but if we find it in the latter we need to wait for it to finish loading
<headius[m]>
there's some racing involved because these are two separate tables
<headius[m]>
olleolleolle: it is but I think we need to find folks to help maintain... codefinger was doing it but I'm not sure he's on Ruby stuff at all now
<headius[m]>
these recent ones are pretty simple
<olleolleolle[m]>
I can probably fix the build (a Bundler detail).
<headius[m]>
can I add you and let you merge in this stuff?
<headius[m]>
not tapping you as official maintainier or anything... but we could use a bit of help there :-)
<olleolleolle[m]>
You could "let me in to merge"
<headius[m]>
yeah
<olleolleolle[m]>
Thanks for the disambiguation!
<headius[m]>
at least
<headius[m]>
olleolleolle: sent
<headius[m]>
there's an olle, an olleolle, and an olleolleolle on GH
<olleolleolle[m]>
hehe, a good group, a band, perhaps
<headius[m]>
hah yes
<olleolleolle[m]>
Alright, let the merging commence.
shellac has quit [Quit: Computer has gone to sleep.]
Antiarc has quit [Quit: ZNC 1.7.4+deb7 - https://znc.in]
shellac has joined #jruby
Antiarc has joined #jruby
<rdubya[m]>
kares: enebo sorry, I missed the end of the conversation. Should I push up this rails-5 branch or were we just going to push them to upgrade?
<olleolleolle[m]>
I sent an attempted patch to mkristian, too. Opening the can of works.
<olleolleolle[m]>
worms
<olleolleolle[m]>
Cheers, goodnight, all
<enebo[m]>
rdubya: kares said he will ask the guy about updating to a release but you could tell him on that issue you do have the branch if wants/needs it.
<enebo[m]>
Honestly it is surprising that branch works for him...I believe we went to master well before rails 5 was stable and working
<enebo[m]>
well perhaps it was a bit better than that but we definitely changed a lot after that
<rdubya[m]>
yeah i can't remember if that is the branch i started working off of or not, but it definitely wasn't in great shape when i started on it lol
<enebo[m]>
I think a lot of normal sorts of db operations did work but we ended up finding more cases with expanded AR unit tests
shellac has quit [Quit: Computer has gone to sleep.]
shellac has joined #jruby
<rdubya[m]>
now that I look at it, i'm not even sure if the copy I have is up to date with what that branch was
<rdubya[m]>
i had forked the repo originally, but then started committing directly to the repo so that branch probably wasn't even getting updated
<enebo[m]>
may make the whole discussion moot then
shellac has quit [Client Quit]
<rdubya[m]>
yeah, i don't think I want to even propose it, if its out of date who knows what effects it might have
shellac has joined #jruby
shellac has quit [Client Quit]
shellac has joined #jruby
shellac has quit [Client Quit]
ur5us has joined #jruby
<enebo[m]>
kares: headius AND ANYONE else using our orgs OSS intellij licenses...Hiro got the update. I just updated and we are good until next March
<headius[m]>
Cool
shellac has joined #jruby
shellac has quit [Quit: Computer has gone to sleep.]