<GitHub97>
[jruby] phluid61 opened pull request #4297: Feature#12217 Enumerable/Array sum (ruby-2.4...feature/12217-enumerable-sum) https://git.io/vX1IV
thedarkone2 has quit [Quit: thedarkone2]
dido has joined #jruby
dido has quit [Client Quit]
sam2000 has joined #jruby
sam2000 has quit [Ping timeout: 246 seconds]
olle has joined #jruby
donV has joined #jruby
<donV>
Hi all!
<donV>
headius kares enebo : I see you are working on ARJDBC, and you have mentioned pg_jruby. Any plans for pg_jruby? I’d like to contribute.
<kares>
donV: hey! not really working myself just some fixes to stable atm
<donV>
Everything helps! :) Thanks you! We are using ARJDBC a lot and rely on it. Thanks!
<kares>
believe pg_jruby needs work/support/help - just like ARJDBC
<donV>
kares: Do you know anything about pg_jruby?
<kares>
donV: know it exists and that its not compatible with latest pg features ... 0.13 I guess
<kares>
think its built on top of the official driver code ... so its still JDBC (not that low level as pg)
<donV>
Since PostgreSQL support is included in ActiveRecord, it seems simpler/cheaper to implement just the connection driver than the whole adapter.
<kares>
personally would choose ARJDBC if I had to on any Rails on Postgres app :)
<kares>
donV: yes definitely that is the course ARJDBC is currently heading ...
<kares>
but I fear it will be much more maintenance to handle ... we will see
<donV>
kares: Really?! Sounds good! Could you elaborate a bit?
<kares>
Tom mentioned for next Rails 5 release there will be much more of sharing with AR
<kares>
I believe he managed to get some hooks into AR core so that it's easier
<donV>
Nice!
<kares>
PG is going to be the hardest piece I believe
<kares>
I mean AR PostgreSQL adapter
<kares>
MySQL should be simlper and for the rest in ARJDBC I do not know the plan
<donV>
I’ll try to ave a chat with Tom about it :)
<donV>
kares: Do you think they plan do implement this inside ARJDBC or have something separate like pg_jruby?
<kares>
donV: what do you mean?
<donV>
You mentioned that they want to share more with AR and just implement the low-level connection logic.
maclover7 has joined #jruby
<donV>
Do you think they will do this inside the ARJDBC project?
<kares>
donV: I meant there's changes in AR itself that should enable more sharing
<kares>
... which is good but I have my doubts (hopefully to be proven wrong!)
dido has joined #jruby
<kares>
SQLite adapter should be done, I wonder how PG will do since there's a lot of code
<donV>
OK, so any changes from our side will probably be done inside the ARJDBC project.
maclover7 has quit [Client Quit]
<donV>
Having better interface with AR would be good.
<kares>
donV: not really about the iface but more about the specific adapter code
<kares>
SQLite is done and should be less code with a future AR 5 release
dido has quit [Client Quit]
<kares>
PG is complex and lot of code so I wonder how will that go ... not saying it can't be done or that its bad - it very good I just fear the Rails team breaking ARJDBC in minor releases
<kares>
... still seems (to me) like they refuse to run CI against JRuby cause its slow :(
pawnbox has joined #jruby
claudiuinberlin has joined #jruby
blaxter has joined #jruby
zacts has quit [Ping timeout: 260 seconds]
vtunka has joined #jruby
shellac has joined #jruby
blaxter has quit [Quit: foo]
zacts has joined #jruby
vtunka has quit [Quit: Leaving]
vtunka has joined #jruby
prasunanand has quit [Ping timeout: 246 seconds]
sam2000 has joined #jruby
sam2000 has quit [Ping timeout: 246 seconds]
shellac has quit [Quit: Computer has gone to sleep.]
prasunanand has joined #jruby
vtunka has quit [Quit: Leaving]
vtunka has joined #jruby
olle has quit [Quit: olle]
shellac has joined #jruby
olle has joined #jruby
drbobbeaty has quit [Ping timeout: 260 seconds]
drbobbeaty has joined #jruby
yopp has quit [Ping timeout: 252 seconds]
dido has joined #jruby
dido has quit [Client Quit]
drbobbeaty has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
shellac has quit [Quit: Computer has gone to sleep.]
shellac has joined #jruby
yopp has joined #jruby
sam2000 has joined #jruby
dido has joined #jruby
dido has quit [Client Quit]
olle has quit [Quit: olle]
sam2000 has quit [Ping timeout: 246 seconds]
olle has joined #jruby
dido has joined #jruby
drbobbeaty has joined #jruby
olle has quit [Quit: olle]
maclover7 has joined #jruby
maclover7 has quit [Client Quit]
<GitHub190>
[jruby] rovf opened issue #4298: CSV.parse tries to close a file https://git.io/vX1P1
<GitHub119>
[jruby] pitr-ch pushed 2 new commits to truffle-head: https://git.io/vX11a
<GitHub119>
jruby/truffle-head 60a5888 Petr Chalupa: [Truffle] fix fix in c39b5de4 fixing d5949818
<GitHub119>
jruby/truffle-head b88b189 Petr Chalupa: [Truffle] let rubygems insert gem libs at correct place
<GitHub120>
jruby/master 978ffc2 Karol Bucek: Merge pull request #4299 from herwinw/math_lgamma_negative_zero...
<GitHub120>
jruby/master 1605ff0 Herwin Weststrate: Fixed Math.lgamma with input -0.0...
<kares>
enebo: hey! the 9.1.6.0 milestone is still open on GH ... one issue was re-opened but it still should not show up as so
<kares>
... great release btw.!
<enebo>
kares: closed…thanks. I did notice this yesterday and forgot about a minute later :)
<enebo>
kares: and thanks for working on it as well :)
<kares>
ok - wasn't sure to close ... thought there might be some spooky reason
<enebo>
kares: Do you know if anyone used the specific isolation levels for sqlite3? AR itself does not support that
<enebo>
kares: so atm on rails-5 it just returns false
<kares>
enebo: nope - du not know
<kares>
sure wasn;t added like someone wanted it badly
<enebo>
kares: ok I am removing it because it causes some impedence against fitting into Rails code more easily
<kares>
ok
<enebo>
kares: I guess if it is something someone wants they can push for it in AR itself first
<enebo>
kares: and that makes sense since sqlite does support a couple of types of isolation
<enebo>
kares: but perhaps no one uses it for anything past development and doesn’t care
<kares>
enebo: meeeh do not get those few ARJDBC contributors ... contribute even less!
<kares>
doubt someone uses it
<kares>
but there sure are excuisite jruby sqlite users :)
<enebo>
kares: yeah. I doubt it too since it is fairly low level call
<kares>
the underlying JDBC driver always felt like a hack ... and there are hacks
<enebo>
kares: but then again my JDBC/DB knowledge is paged out a lot over the last half a decade
<kares>
to be fair none of the embed 'drivers' are much to expect in terms of compliance
<enebo>
kares: yeah my big frustration was that one path would pass binds in as java.util.List and it took me hours to realize it was performing coercion
<kares>
enebo: sure not too vary - I used to drink heavily while doing ARJDBC - so that I do nto remember any of it
<kares>
* not to worry
<enebo>
kares: hahaha
<enebo>
kares: yeah I am amazed you were able to keep all these versions running at the same time. You have my respect :)
<enebo>
kares: so I hope to get rid of the extra version logic and hopefully when it is done it will not seem like a ton of code to support
<enebo>
kares: we can also be hopeful that AR will continue to converge on less and less changes now that it has reached a level of maturity
<kares>
enebo: REALLY hope so - otherwise this will be madness
<kares>
... and that they finally let JRuby run on travis
<enebo>
kares: but I wish there was a DBI/JDBC later which make the native adapter bits a bit more uniform…although it is moving that way more or less
<kares>
although that still won't be anough to test against ARJDBC regressions (due the sharing you're planning)
<enebo>
kares: sqlite3 for Rails 5.1 should be 75% less Ruby than it is for Raiols 5.0
<enebo>
kares: yeah but we can add regression tests locally and hopefully some of those will be regressions for them too
<kares>
enebo: ... just afraid they my broke smt for ARJDBC easily
<enebo>
kares: sorry I am not parsing that statement
<kares>
* they may broke ARJDBC with a minor point release touching the native adapter - not realizing its reused
<kares>
hopefully not
<kares>
enebo: you should see the enormity of sharing possible for PG
<kares>
actually a lot of is shared ... but due older versions older .rb parts are still there
<kares>
enebo: the thing is as I did the more sharing on PostgreSQL adapter on AR 4.2 it was fine
<enebo>
kares: yeah sqlite3 may have been a poor first adapter but I have a hard time seeing how most of the newer shared code breaking stuff
<kares>
and than they broke it in a point release
<enebo>
kares: ah I see
<enebo>
kares: yeah it is an issue definitely
<kares>
... by putting in native pg stuff 'for optimization'
<kares>
was unable to get them convinced its bad for ARJDBC
<kares>
simply lost hope for more and more sharing a long time ago
<kares>
but I glad someone else did not :)
<enebo>
class SQLite3Adapter < AbstractAdapter
<enebo>
include ArJdbc::CommonJdbcMethods
<enebo>
include ArJdbc::SQLite3
<enebo>
This is newer layout for adapters
<enebo>
that third line in Rails 5.1 will come from AR itself which is why we will drop 75% of our code in this adapter
<kares>
yeah i've seen it -- looks great :+100
<enebo>
the class referenced will be where we put overrides so even if they break a method we can override it
<headius>
kares: I'm hopeful that as we boil down the jruby-specific bits of arjdbc the AR maintainers may let us just merge it into rails proper...or at least work with us to maintain it
vtunka has quit [Quit: Leaving]
<GitHub164>
[jruby] headius pushed 1 new commit to travis_hang: https://git.io/vXMqh
<GitHub164>
jruby/travis_hang 4b9509d Charles Oliver Nutter: ...
<headius>
travis docker image is 1.3GB...sigh, guess I'll make coffee
shellac has quit [Quit: Computer has gone to sleep.]
bbrowning is now known as bbrowning_away
<GitHub39>
[jruby] eregon pushed 7 new commits to truffle-head: https://git.io/vXMKJ
<GitHub39>
jruby/truffle-head 4321d0e Benoit Daloze: [Truffle] Remove a useless indirection.
<GitHub39>
jruby/truffle-head 45e9e5b Benoit Daloze: Move Solaris File#flock specs in a different describe...
<GitHub39>
jruby/truffle-head d5a1e62 Stefan Anzinger: Fix UDPSocket.new spec, Solaris reports a different error on unknown protocol family.
<eregon>
nirvdrum: maybe we should do a TODO-week :) I'm sure we'd find some interesting bits.
maclover7 has joined #jruby
<headius>
yay, reproduced it
maclover7 has quit [Client Quit]
maclover7 has joined #jruby
maclover7 has quit [Client Quit]
shellac has joined #jruby
claudiuinberlin has quit []
dido has joined #jruby
bbrowning_away is now known as bbrowning
olle has joined #jruby
tcrawley is now known as tcrawley-away
djellemah has joined #jruby
<GitHub157>
[jruby] nirvdrum pushed 4 new commits to truffle-head: https://git.io/vXMDc
<GitHub157>
jruby/truffle-head 2a96185 Kevin Menard: [Truffle] Cleaned up StringOperations by inlining some methods only needed by StringCodeRangeableWrapper.
<GitHub157>
jruby/truffle-head 57e6a1e Kevin Menard: [Truffle] Introduced UTF8Operations to organize code a bit better.
<GitHub157>
jruby/truffle-head a4310d4 Kevin Menard: [Truffle] Removed obsoleted TODOs.
<prasunanand>
I think it's a great place to start :smile: .
<chrisseaton>
prasunanand: where are you starting from? What do you know about Ruby and Java?
<prasunanand>
I worked on my GSoC 2016 project 'Port NMatrix to JRuby'
<prasunanand>
with Sciruby
<prasunanand>
I am interested in Sulong+Truffle
<chrisseaton>
Great!
<prasunanand>
To get started I will work with Begineer issues.
<prasunanand>
Thanks :)
<chrisseaton>
That list is about normal JRuby, and I wouldn't want to steal you to work on Truffle instead, so you could just pick anything there and start by reading the MRI issue to understand how the new feature works.
<prasunanand>
chrisseaton: I have read your slides. Would love to work on Truffle
<prasunanand>
Okay :)
<chrisseaton>
prasunanand: if you came here already wanting to work on Truffle then that'd be amazing - I just don't want to look like I'm poaching people who came here to work on classic
<chrisseaton>
Either way, we need to fix all the issues on that list as well, so you can still start by reading the MRI issue
<chrisseaton>
What could be fun would be implementing the feature for both classic and Truffle
<chrisseaton>
Could be fun to compare how easy it is and how fast it is on each
<chrisseaton>
When you've read the MRI issue the next step could be to write some specs
<chrisseaton>
Ask if you need any help at all!
<prasunanand>
Thanks chrisseaton . Starting with classics will definitely help me to understand the codebase.
maclover7 has joined #jruby
kith has quit [Read error: Connection reset by peer]