<GitHub196>
jruby-openssl/master b88c206 kares: at least (internal) debug when BC provider fails to instantiate
sidx64 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
sidx64 has joined #jruby
olle has joined #jruby
sidx64 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
sidx64 has joined #jruby
sidx64 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
sidx64 has joined #jruby
ralberto has joined #jruby
<ralberto>
Is there a place where I can find the EOL of JRuby versions?
<kares>
ralberto: right here :)
sidx64 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<enebo>
kares: heya
<enebo>
kares: so we changed the gem name to add -java then right?
<kares>
yep seems so
<enebo>
kares: will that push to same location because that is just an architecture flagy thing
<kares>
you mean the gem push?
<enebo>
kares: well I mean it will go to same rubygems.org entry
<kares>
yep it will be fine
<kares>
no worries
<enebo>
kares: I realized as you noticed just recently about the dep locking
<kares>
enebo: we did the same for jossl years ago
<enebo>
kares: I was testing a bug and did 51.0 and it pull the 51.1 jruby-adapter
<enebo>
kares: so that was unexpected so I half think we should lock them together as we have no expectations we can mix them
sidx64 has joined #jruby
<kares>
enebo: yy that makes sense ... for 1.3 it wasn't much of an issue
<kares>
users were used to using the base gem
<enebo>
kares: yeah I still would like to figure out how to decouple this central piece so any point release for a particular adapter does not mean revving all of them...but that is perhaps a much more distant desire
<enebo>
kares: in Rails 6 they will also be allowing multiple db connections so we need to see how that all plays out too
<kares>
enebo: what does that mean?
<kares>
multiple what in the same thread or what?
<enebo>
kares: sounds like you will have n dbs you can configure and it will load them all
<kares>
that works since like 2.3 or 3.0
<enebo>
kares: so likely a single controller will be able to ask from multiple data sources
<kares>
that isn't new really :)
<enebo>
well they made some point about making it a more significant feature so I don't know
<enebo>
I was not there :P
<kares>
its just that they pbly centralize it up a bit
<kares>
every model can connect wherever it wants
<kares>
maybe its about allowing 2 data-sources for one model
<kares>
that would be smt new in core
<enebo>
trying to figure out what was meant now
<kares>
I wouldn't worry
<kares>
by than arjdbc will be part of core anyway :)
<enebo>
heh
<enebo>
ok. so I think we can release 50.1 and 51.1
<enebo>
we did talk about this before but unless you noticed anything new to worry about (perhaps being more explicit on the linking between the two gems)
<kares>
YAY
<kares>
you can put in the lock for gem so it grabs 50.1 with jdbcmysql-adapter
<enebo>
I will try your gist but if you want to actually release that would be helpful. I am not sure though if you are still on vacation
<kares>
I am back - working on jossl - trying a quick turn-around for a release as we speak
<enebo>
oh ok I will do this then
<kares>
before I get into finding a job :)
<kares>
np I can release
<enebo>
oh ok :P
<kares>
in like 3h ... will need a break anyway
<enebo>
kares: I am going to make 5.2 branch since I think I may have put that in an abstract
<enebo>
Even if I didn't I think I want to see how much work it will take
<kares>
okay so releasing as is from 50-stable + 51-stable which is master
<enebo>
right
<kares>
do you need 52-stable ... it can be master for now?
<enebo>
kares: if 52 work goes really well we may need to make master 51-stable
<kares>
enebo: well go ahead and do that, maybe?
<kares>
esp. if you're planning 5.2 work
<kares>
or whatever feels best
<enebo>
kares: yeah 51-stable should be made and master should be 52 work then
<kares>
makes sense
<enebo>
I will throw 51-stable branch right now from master
<enebo>
ok
<kares>
great we'll release from stable branches than
<enebo>
ok should be pushed
<enebo>
kares: this is like a birthday dream (one day early) :)
<kares>
ah so more beer testing tomorrow ... happy-happy than :) !
<GitHub21>
jruby/master 30a6ae2 Charles Oliver Nutter: Update jnr lib versions to latest.
<headius>
kares: my 10 might be a prerelease
<GitHub33>
[jruby] headius closed issue #1408: Jruby class implementing java interface with "initialize" method as a part of the interface fails to compile with jrubyc https://git.io/vpFw4
<GitHub88>
[jruby] headius closed pull request #3807: [ji] allow for initialize() to be a valid Java method (master...jrubyc-initialize) https://git.io/vwm6g
<headius>
apparently I don't understand class vars again
<headius>
enebo: I'm not fixing that BS
<headius>
hmm well it might be possible if we just BREAK class variables to always search for the highest one
<headius>
on every get and set
<headius>
pretty sure that's what MRI does
<headius>
enebo: yeah that works
<headius>
basically it will now always search for the highest one
<enebo>
hmm
<enebo>
I wonder if people depend on this
<headius>
nobody should use class variables period
<headius>
but this upends my understanding of them once again
<headius>
I thought child classes could override them
<headius>
I can't think of any language that has a feature like this
<GitHub195>
[jruby] headius opened pull request #5175: Always search full hierarchy for highest definition of a cvar. (master...stupid_cvars) https://git.io/vpFyf
<GitHub125>
[jruby] headius closed issue #3851: [JRuby 1.7.25] YAML.load_file() fails with IOException after a number of iterations https://git.io/vwFvP
<GitHub114>
[jruby] headius pushed 1 new commit to new_reify_classes: https://git.io/vpFj9
<GitHub114>
jruby/new_reify_classes 39a76a3 Charles Oliver Nutter: Merge branch 'master' into new_reify_classes
<headius>
enebo: so I guess I forgot I spiked generating reifiy vars classes
<headius>
intending to continue on and add the reify classes logic on top of it and never got back to it after the work last year
<headius>
this eliminates the RubyObjectVar0-9 classes and generates all shapes
<headius>
and it's passing CI
<headius>
enebo: you want me to fix those MRI failures on master? that and another failure looking for rake in stdlib are the only things red on master
<headius>
and how much does the reification scare you