zorak8 has quit [Read error: Connection reset by peer]
zorak8 has joined #jruby
pietr0 has quit [Quit: pietr0]
mrmargolis has joined #jruby
pjammer has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
marr has quit []
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] enebo pushed 1 new commit to master: http://git.io/vfFvf
<JRubyGithub>
jruby/master a5c372d Thomas E. Enebo: Readline somewhere does STDXXX.fileno and we redefine std{in,out,err}. Also change constants in addition to globals
JRubyGithub has left #jruby [#jruby]
bjfish2 has quit [Quit: bjfish2]
camlow325 has quit []
mrmargolis has quit [Remote host closed the connection]
<enebo>
headius: can you check to see if updated jnr-posix fixes this for jruby-1_7?
<enebo>
headius: I am retesting a bit because I am paranoid
mkristian has joined #jruby
benlovell has joined #jruby
rsim has quit [Quit: Leaving.]
<headius>
ok
<headius>
oh yes
<headius>
oh man, Windows?
<enebo>
HAAHAH
<enebo>
headius: sorry I did not notice that :)
<enebo>
It is fixed for master so I can close it so it shows up in fixed issues list
<headius>
I think I already fixed and confirmed this and it was just waiting on the release
<enebo>
headius: you comment made it sound a little unclear
<headius>
1.7 is still on previous jnr-posix but I'll update that
<enebo>
headius: I will jot a note down and try it on windows after 9k.pre2 is out
arturaz has quit [Ping timeout: 245 seconds]
<headius>
oh right I see my comment about not testing it on 1.7
<headius>
I don't see why it should be different though, they both call stat through posix
<enebo>
headius: it can only be different if we handle the string we pass to jnr-posix differently
mkristian has quit [Ping timeout: 265 seconds]
<headius>
I didn't touch any of that stuff during my 9k rework
<headius>
but yeah, we will retest
<headius>
I will update pom
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius closed issue #2424: Error in File.stat http://git.io/2lH8dg
JRubyGithub has left #jruby [#jruby]
<enebo>
headius: yeah I am not saying you would have changed anything but it is possible we just do something different with path strings on 1.7 (although I don’t know why we would)
<JRubyGithub>
[jruby] headius reopened issue #2424: Error in File.stat http://git.io/2lH8dg
JRubyGithub has left #jruby [#jruby]
bjfish2 has joined #jruby
<enebo>
headius: I want it closed so the issues script I run shows it resolved
<headius>
it's not marked for pre2 now
<enebo>
headius: I could manually make that line I guess
<headius>
I moved it to 1.7.20
<enebo>
headius: oh ok. yeah it was actually reported originaly against 1.7
<headius>
right, I just confirmed and fixed on 9k, but it's probably fine for 1.7 after confirmation
<headius>
github's fault for not letting us set multiple milestones
<headius>
it and other bugs fixed in both 9k and 1.7.20 won't show up for 9k bullets
pjammer has joined #jruby
mkristian has joined #jruby
rcvalle has joined #jruby
benlovell has quit [Ping timeout: 276 seconds]
<cpuguy83>
For the ruby image in Docker they've started running tests to make sure the things like the everything in the stdlib can all be loaded... naturally they started running the same test on the jruby image, and noticed that `dbm` is not included
<headius>
cpuguy83: all the *db* stdlibs are C exts
<cpuguy83>
ah
<headius>
for slightly different variants of db
<headius>
I don't know why they're even in stdlib anymore
<cpuguy83>
headius: Thanks
zorak8 has joined #jruby
camlow325 has joined #jruby
codefinger has joined #jruby
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] bjfish pushed 1 new commit to master: http://git.io/vfxOf
JRubyGithub has left #jruby [#jruby]
<JRubyGithub>
jruby/master e75cd2d Brandon Fish: [Truffle] Add json to stdlib.
shellac has quit [Quit: Computer has gone to sleep.]
<rtyler>
what's left on the docket for it?
<chrisseaton>
is DonV on here?
fivebats_ has joined #jruby
rsim has quit [Ping timeout: 265 seconds]
fivebats has quit [Ping timeout: 252 seconds]
bbrowning is now known as bbrowning_away
<enebo>
rtyler: we have 17 open issues marked against it
<enebo>
rtyler: about 8 of them are LOAD_SERVICE problem so it may not be as big a number as it seems
pjammer has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<rtyler>
what about this rubygems 2.4.5 upgrade work?
subbu|brunch is now known as subbu
<rtyler>
I'm trying to figure out what from this list I should try to schedule onto mkristian's Lookout backlog, since we're blocking a number of internal releases based on fixes in 1.7.20
pjammer has joined #jruby
<rtyler>
that + sign one is particularly frustrating :/
camlow325 has quit [Remote host closed the connection]
ir2ivps9 has quit [Ping timeout: 240 seconds]
<enebo>
rtyler: headius just landed it again today…third time will be for all the glory
<rtyler>
perhaps related, would it be useful to have a google hangout to go through some of this stuff and see what we (Lookout) can help do here?
<enebo>
rtyler: so long as I am concerned we can punt all non-fixed stuff by thursday so long as this rg stuff + LOAD_PATH stuff is finally fixed
<enebo>
rtyler: looking at stuff a bunch are long-standing broken stuff which could always just stay broken a bit longer
<rtyler>
are you guys planning on tackling any of that this week?
<enebo>
rtyler: I am going to look at some of these windows issues
<enebo>
rtyler: but I think the LOAD_PATH + rg stuff should be done now barring a discovery
<rtyler>
comparing and contrasting internal versus external backlogs is a PITA
gazay has quit [Read error: Connection reset by peer]
x1337807x has joined #jruby
gazay has joined #jruby
codefinger has quit [Remote host closed the connection]
codefinger has joined #jruby
djellemah has quit [Read error: Connection reset by peer]
iamjarvo has joined #jruby
<headius>
rtyler: kates confirmed 2.4.6 looks ok on 1.7.20
<headius>
rtyler: you've been testing on 1.7.20 for a while then, yes?
<headius>
if you can kick all your builds to run with RG 2.4.6 it would be good confirmation
<rtyler>
I have not
<rtyler>
without 1.7.20 builds going to a snapshots repo, it's a lot more work than I have time to get 1.7.x pre-release candidates tested effectively here :/
<headius>
ahh righto
* rtyler
could set up Jenkins internally to start publishing to our internal artifactory with snapshots
baroquebobcat has quit [Quit: baroquebobcat]
<rtyler>
not the worst idea actually
cultureulterior1 has joined #jruby
<headius>
well if mkristian needs something to do, he might be able to help us get automated snapshot nightlies working
<rtyler>
got ticket?
<nirvdrum>
cpuguy83: I'm finally checking out the JRuby docker images. I see the 'dev' tag from 2014-12-02. I take it that's not updated on pushes. Is that intended behavior? Or did something break?
<headius>
not yet...it's your feature :-)
<rtyler>
heh
<rtyler>
alright alright
<headius>
nirvdrum: dev tag? I've never done that
<rtyler>
FINE DAD
ir2ivps9 has joined #jruby
baroquebobcat has joined #jruby
<nirvdrum>
headius: A docker tag. Not one in our repo.
<headius>
oh ok...nope, I don't know anything about it
<mkristian>
headius, hopefully you are talking about a maven snapshot repo
pierreatkillbill has quit [Ping timeout: 276 seconds]
<headius>
mkristian: yes
<cpuguy83>
nirvdrum: yeah, dev was tracking master, though I still had to manually update the commit hash for it.
cultureulterio-1 has quit [Ping timeout: 250 seconds]
<mkristian>
cool
<headius>
so that when releases lag folks aren't as blocked
mister_solo has joined #jruby
<cpuguy83>
nirvdrum: This was too difficult to do since the build artifacts for jruby would disappear and it wouldn't be buildable anymore.
<headius>
cpuguy83: hmm?
<cpuguy83>
nirvdrum: So now we just do the pre-1... working to get pre-2 up.
<headius>
cpuguy83: can it use our nightly snapshots?
<headius>
ci.jruby.org
<headius>
those filenames should be consistent and they build every night
<nirvdrum>
cpuguy83: The builds at http://ci.jruby.org/ have been stable for a while. Although the tarball is overwritten nightly.
<cpuguy83>
headius: Not without me updating it manually
<headius>
because it needs a hash too?
lance|afk is now known as lanceball
<cpuguy83>
It needs something that changes or else the build is cached.
<nirvdrum>
Ahh.
<cpuguy83>
And if the URL we are pulling doesn't change there's not much to change.
<headius>
well let us know if there's something we can do better
<cpuguy83>
I guess I could pre-compute the sha and use that
<cpuguy83>
I think sticking to the pre-N builds should be good. If someone wants master or a nightly they can adjust our dockerfile to point there.
<headius>
ok, no problem then
shellac has joined #jruby
<nirvdrum>
cpuguy83: Does it need to be something that changes in the filename? The ETag HTTP response header will update.
<cpuguy83>
Correct, I think all is good.
<nirvdrum>
I'm just curious at this point.
<cpuguy83>
nirvdrum: Something would need to change in the Dockerfile
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] chrisseaton pushed 2 new commits to truffle-head: http://git.io/vfpXU
<JRubyGithub>
jruby/truffle-head 6044f60 Chris Seaton: Merge branch 'master' into truffle-head
<JRubyGithub>
jruby/truffle-head 320f656 Chris Seaton: [Truffle] Reorder priorities when searching for Graal.
JRubyGithub has left #jruby [#jruby]
<cpuguy83>
nirvdrum: We could use `ADD <url>` and I think that would work, but then we end up with an extra image layer with the tarball in it which is not needed (and takes up space)
<nirvdrum>
Gotcha.
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius deleted test-travis-fail at 6bdfd35: http://git.io/vfp19
JRubyGithub has left #jruby [#jruby]
<headius>
if anyone has branches pushed to github they're done with, please remove them
gazay has quit [Quit: gazay]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] nirvdrum deleted lexer_source_encoding at 8accc1d: http://git.io/vfpDh
JRubyGithub has left #jruby [#jruby]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] mkristian deleted test-load-service at 0ee53c5: http://git.io/vfpyf
JRubyGithub has left #jruby [#jruby]
<nirvdrum>
headius: There are a lot of stale branches. Might just have to clean house on some of those.
<JRubyGithub>
[jruby] mkristian deleted pack200 at 42f65bc: http://git.io/vfpyF
JRubyGithub has left #jruby [#jruby]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] mkristian deleted classloaders at 3ce1329: http://git.io/vfpSy
JRubyGithub has left #jruby [#jruby]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] mkristian deleted uniform-jruby-home at 2a159b6: http://git.io/vfp9p
JRubyGithub has left #jruby [#jruby]
x1337807x has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] mkristian deleted no-duplicated-yaml at 2caa923: http://git.io/vfpQ8
JRubyGithub has left #jruby [#jruby]
<headius>
nirvdrum: after a release is as good a time as any
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius deleted test-proxy-leak at 2bdfb5a: http://git.io/vfp7J
JRubyGithub has left #jruby [#jruby]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius deleted method_audit at bac44a6: http://git.io/vfp7L
JRubyGithub has left #jruby [#jruby]
<headius>
nice, the branch view shows the branches that you made (presumably branches that only differ by your commits from master baseline
<headius>
)
<headius>
or branches that have ANY of your commits different from master, I suppose
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] mkristian closed pull request #2509: improving building of distribution archives and artifacts (master...test-build-distribution) http://git.io/VU7Oaw
JRubyGithub has left #jruby [#jruby]
<headius>
so many experiments
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius deleted gh-pages at d620f56: http://git.io/vfpdk
JRubyGithub has left #jruby [#jruby]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius deleted prepend at f1ff8dd: http://git.io/vfpdc
JRubyGithub has left #jruby [#jruby]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius deleted new_transcoder at 913a4c4: http://git.io/vfpdg
JRubyGithub has left #jruby [#jruby]
<nirvdrum>
"enebo_lexical Updated 9 years ago by headius"
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius deleted memory-opto at 10d2ce1: http://git.io/vfpFs
JRubyGithub has left #jruby [#jruby]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius deleted dynbuf_io at e564e31: http://git.io/vfpFl
JRubyGithub has left #jruby [#jruby]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius deleted test-fix-2221 at a9fa229: http://git.io/vfpFz
JRubyGithub has left #jruby [#jruby]
<headius>
nirvdrum: pretty sure that's done but enebo can get it when he's back from lunch
<headius>
mmm cleaning house
<nirvdrum>
There's a few from Ola, too.
<nirvdrum>
9 years is a long time :-)
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius deleted test-instance_eval-self-fix-2301 at 7178caa: http://git.io/vfpFX
JRubyGithub has left #jruby [#jruby]
<headius>
anything from Ola can be nuked
<enebo>
hmmm
<enebo>
I have been back from lunch for almost 2 hours
<headius>
[enebo] Updated 8 years ago by headius
<enebo>
I just ate a sandwich
<headius>
enebo: a really BIG sandwich
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] nirvdrum deleted fixShiftLeft at 52de8f1: http://git.io/vfpbq
JRubyGithub has left #jruby [#jruby]
<headius>
serializable_procs...experiment I did for avibryant to show procs serializing
<headius>
I think he wanted to do something with hadoop
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius deleted serializable_procs at 3f63adc: http://git.io/vfpbl
JRubyGithub has left #jruby [#jruby]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius deleted one-time-warning at 70b90c3: http://git.io/vfpbE
JRubyGithub has left #jruby [#jruby]
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] nirvdrum deleted GH-1551 at 05865c1: http://git.io/vfpbw
JRubyGithub has left #jruby [#jruby]
<headius>
sigh
<headius>
clojure branch still has my transactional constants/classes/modules
<nirvdrum>
I think this tmpdir-failure branch from nahi was for an issue I reported. And that's since been fixed.
<nirvdrum>
That was 3 years ago, so I may not be recollecting that properly.
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius deleted clojure at 49da36e: http://git.io/vfpNd
JRubyGithub has left #jruby [#jruby]
<headius>
nirvdrum: if you think you remember and it's ok, go for it
<headius>
nahi hasn't contributed in a good while
<codefinger>
when building my own pre2 tarball, i'm getting a `lib/target/classes/META-INF/jruby.home` dir that isn't in the tarball on jruby.org. looks like i can just remove this?
<subbu>
oh .. hmm.. delete it. i have a local copy here anyway .. unless enebo needs it.
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] nirvdrum deleted tmpdir-failure at 716166f: http://git.io/vfpA1
JRubyGithub has left #jruby [#jruby]
<headius>
codefinger: the tarballs we make are via some dist poms under maven/
<headius>
I believe mvn -Pdist builds all of them
<headius>
if you want to do your own that dir is not important...I think it's a build artifact that should be deleted (mkristian?)
<nirvdrum>
I suspect yokolet's startup-time branch is just obsolete with IR now. But maybe one of you want to spot check it.
<enebo>
subbu: headius: yeah I don’t care I don’t think
<headius>
enebo, subbu: you guys aren't giving me a lot of confidence
hone has joined #jruby
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] nirvdrum deleted avoid_seekable_byte_channel at e712294: http://git.io/vfpxX
JRubyGithub has left #jruby [#jruby]
<mkristian>
headius, codefinger, under the ./lib directory we build the jruby-stdlib.jar now - which is new for pre2. of course all target/** should be excluded from the tar
<enebo>
headius: codefinger: heh I am also talking to hone2 right now too about this…I don’t know why it is there but I am guessing it is somewhat transient data used for packaging
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] kares deleted test-embed-internal-fixes at a05035b: http://git.io/vfpxx
JRubyGithub has left #jruby [#jruby]
<hone>
hi
<enebo>
codefinger: I did not see this enter any of the dist bits I made
<mkristian>
codefinger, but I would have a better feeling if you would use ```mvn -Djruby.default.ruby.version=2.2.2 -T4 -Pdist``` and use the tar under
<enebo>
headius: don’t remove it then :)
<subbu>
headius, DELETE IT .. is that confidence enough for you? :)
<codefinger>
mkristian: yea makes sense
<enebo>
HAHAHAH
<mkristian>
the tar will be in ./maven/jruby-dist/target
<headius>
enebo, subbu: I'll leave it up to y'all :-)
<enebo>
I spent some amount of time trying to find any vestigial/shit extra files in our dists and found none
<JRubyGithub>
[jruby] mkristian deleted test-no-if-else-try-catch-logic-for-loading-jars at d87e197: http://git.io/vfhOR
JRubyGithub has left #jruby [#jruby]
<headius>
mkristian: so question about the 3.3.1 branch
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] mkristian deleted merge at e3d93f2: http://git.io/vfhOV
JRubyGithub has left #jruby [#jruby]
<headius>
with that change it removes pom.xml everywhere we can have a pom.rb, but then everyone *must* be using 3.3.1 or higher
<headius>
I know there are some stubs we could add to root of repo too
<headius>
rmvn.bat and sh scripts
<mkristian>
this is the build moved to maven-3.3.1 without any pom.xml - there is a mvnw like gradlew which we can include
<headius>
right ok
<mkristian>
and works OK on travis as well
<headius>
oh, I'll have a look at that
<mkristian>
Ithe current maven-3.3.1 needs mvn-3.3.1 and it is not up-to-date anymore
brightball has quit [Ping timeout: 256 seconds]
tcrawley is now known as tcrawley-away
<mkristian>
headius, about the 'do-not-change-jruby-commandline-context-classloader' - I still think it is needs to be merged and if someone really "depends" on the command-line only feature then a one-liner
<mkristian>
helps to let even work in servlets
<headius>
hey homebrew has 3.3.1
<mkristian>
headius, that is nice
<headius>
nirvdrum: what's the latest maven package for your env?
<headius>
are you on Windows or Linux now?
<headius>
mkristian: the commits on that branch appear to add rdoc and some other things to repo
<headius>
I think you sucked in some files you didn't want
<nirvdrum>
headius: I have both loaded.
<nirvdrum>
headius: Are you asking what version of maven I have installed?
<headius>
nirvdrum: asking if 3.3.1 is available for package install on your Linux
<nirvdrum>
I just use tarballs for each.
<headius>
just trying to get the lay of the land...if we get everyone using 3.3.1 we can ditch most pom.xml
<headius>
ok
<nirvdrum>
Distribution supplied Java is usually a minefield.
<mkristian>
headius, all lib/ruby/shared from 1_7 branch - looks like
<nirvdrum>
Although I do use Ubuntu's ant since that hasn't updated in years.
<headius>
the linux dists have really upped their game wrt java packages
<headius>
I think they all have some integration with maven now
<nirvdrum>
Ahh. Well, I still just use a tarball :-)
<headius>
mkristian: yeah
<mkristian>
ubuntu comes still with maven-3.0.5
<headius>
mkristian: the changes to make us 3.3.1 only aren't big, right? maybe you can re-push that branch based on current master
<nirvdrum>
I'd check with jeremyevans, too. I think he maintains packages for OpenBSD.
<nirvdrum>
I'm not sure if bumping maven affects people that want to build JRuby for distribution in their dists.
<headius>
nirvdrum: for any folks that need to build and don't have mvn 3.3.1 there would be build scripts in project root
<headius>
yeah that's a great question
<headius>
they complain about everything
<nirvdrum>
I guess on Windows I've installed via Chocolatey. But I could grab a zip instead easily enough.
<mkristian>
the for the IDE support we should keep the the pom.xml, core/pom.xml, truffle/pom.xml and lib/pom.xml - but you guys need to use mvnw or mvn-3.3.1 when changing the pom.rb
<headius>
mkristian: ok
<headius>
I updated my mvn so I'm good... enebo: you probably should brew update ; brew upgrade maven
<headius>
assuming you use homebrew
<headius>
mkristian: with 3.3.1 and both pom.rb and pom.xml, will it try to prefer pom.rb?
<headius>
I'm ready to join mkristian in the "I only update pom.rb" world
<jeremyevans>
headius: The OpenBSD jruby port just uses the binary distribution, we don't build from source anymore (other than the ffi code)
<headius>
mkristian: context-classloader change... passes all tests I presume
<headius>
jeremyevans: ok thank you
<headius>
as always let us know if we can do anything to make it easier
<enebo>
hmmm perhaps I should update to YO SA MIGHT this week
<enebo>
whoo hoo 3.3.1
<headius>
enebo: after 1.7.20 :-)
mkristian_ has joined #jruby
<headius>
I didn't do an upgrade so I have no idea what breaks
<headius>
occurred to me...if anyone working on jruby wants to use my IRC bouncer just let me know
yfeldblum has joined #jruby
<mkristian_>
headius, yes it will take the pom.rb instead of pom.xml
mkristian has quit [Ping timeout: 255 seconds]
<headius>
mkristian_: neato
<headius>
hmm, some red jobs on the RG 2.4.6 update
<mkristian_>
headius, it does break one spec which tests the setting of the classloader. and on spec which loads a reified class from the context class loader which is the jruby-classloader only when using bin/jruby
JRubyGithub has joined #jruby
<JRubyGithub>
[jruby] headius pushed 1 new commit to jruby-1_7: http://git.io/vfhCN
<JRubyGithub>
jruby/jruby-1_7 43d3006 Charles Oliver Nutter: Mo' pems, mo' bettah.
JRubyGithub has left #jruby [#jruby]
zorak8 has joined #jruby
<headius>
mkristian_: so both cases that explicitly go after context classloader
aramisbear has quit [Quit: Leaving...]
<mkristian_>
yes - as far I remember
<headius>
I wish we'd have had this exercise before pre2
<headius>
enebo: what do you think
<headius>
this is a change that removes our setting of context classloader for command-line only
<headius>
looks like I originally did it in 59719b4a76807b178c21338bf74595dac91af836
<mkristian_>
headius, well my point is that is not a good idea to have command-line only feature. the moment you put the "working" code into a servlet is stops working
<enebo>
So we change some behavior which will only break non-embedding scenarios
mrmargol_ has joined #jruby
<headius>
right, I get that
<headius>
I'm just trying to see why I did it originally
<chrisseaton>
nirvdrum: thanks - the problem is I have a few bits different in a large file from a third party application - no idea how to start even figuring out what object it comes from
<nirvdrum>
I think if the files load to equivalent objects in MRI and JRuby, you might be okay.
baroquebobcat has quit [Quit: baroquebobcat]
kfpratt has quit [Remote host closed the connection]
<chrisseaton>
yeah - I haven't actually tried that - good idea
zorak8 has quit [Read error: Connection reset by peer]