<headius> yay
<headius> productive day
joast has quit [Ping timeout: 244 seconds]
sid4321 has joined #jruby
<sid4321> Hello jruby users
<sid4321> Is it possible to get the jruby 1.7.1.tar.gz file ??
<sid4321> The AWS links to the same are down
sid4321 is now known as kill91
kill91 has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client]
sid4321 has joined #jruby
sid4321 has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client]
jmalves has joined #jruby
rdubya has quit [Ping timeout: 252 seconds]
xardion has quit [Ping timeout: 252 seconds]
xardion has joined #jruby
bga57 has quit [Ping timeout: 250 seconds]
rdubya has joined #jruby
rdubya has quit [Ping timeout: 260 seconds]
rdubya has joined #jruby
shellac has joined #jruby
<jmalves> Hey all! is there a better way to use a Map<Integer, X> coming from java in JRuby besides having to cast the key to java.lang.Integer?
<jmalves> Example `super_important_map[5863.to_java(java.lang.Integer)]`
shellac has quit [Quit: Computer has gone to sleep.]
<jmalves> (ok the map is not that important, was mostly just for dramatic effect!)
shellac has joined #jruby
joast has joined #jruby
<kares> jmalves: unfortunately Ruby Integer casts to Java long by default
<kares> and since we do not have type-info on the map's generics JRuby can not know what
<kares> ... you can shorten to 42.to_java(:int)
<headius> sid4321: yeah we restored some but not that...enebo may be able to get it uploaded, I'm not sure
<headius> our host accidentally wiped out all our release files
<headius> enebo: can you upload your whole collection of releases somewhere so I can make a backup?
<headius> maybe we should upload them to sf.net 😀
<enebo> headius: yeah definitely. In fact I will add all back to S3 and you can then use script to grab them
<enebo> I will add a note and try and do that today
<headius> jmalves: there's not really any way to tell it was generified that way at runtime due to erasure
<headius> jmalves: I believe to_java(:int) works as well for that thoughhuh
<headius> huh, I just noticed the default toJava does a slow lookup of data wrap struct stuff
<headius> enebo: ok
<headius> ugh, and all the numeric converters do a hash lookup of the converter too
<headius> dark matter
<headius> enebo, kares: we should chat about what we need to finish for 9.2.1
<enebo> headius: on simplest level we just need to realize which marked issues left are important enough
<enebo> I would say only one obvious to me is the Dir/glob one
<kares> +1 for the Dir glob
<kares> and it would be great to have a semi-green build on 11
<kares> but that is not hold till the maven plugins get released
<kares> also 1.7 is "back" https://github.com/jruby/jruby/pull/5359
<headius> 1, 2, 3, NOT IT
<headius> kares: oh boy
<headius> I'd be happy to work on 11, less happy to work on glob
<headius> but I support none of us want to work on glob
<headius> suppose
<kares> I do not mind but also should finish up some of my stale PRs
<kares> oh yeah smt for you guys to review: https://github.com/jruby/jruby/pull/5219
<kares> ^^ this one isn't stale but a change on how we inspect all Java objects // headius enebo
<headius> oh yeah
<kares> problem is atm all inspect gives on Java types with toString is hashy
<kares> irb(main):011:0> java.util.Date.new.inspect
<kares> => "#<Java::JavaUtil::Date:0x2aece37d>"
<kares> so e.g. with rspec you basically have no idea what's wrong
<headius> yeah I don't like those useless inspects either
<headius> can you show me example of your patch without me trying it?
<headius> my concern was always that toString for a large object would be way too much information for inspect typically
<headius> Java doesn't really have an inspect concept
<headius> I suppose you have the same issue in Ruby if you inspect a large nested array or hash
<kares> like what you'd like to see output?
<jmalves> @kares @headius, thx for you anwsers!
<kares> mostly I am not sure about draw backs I might be missing by going inspect -> toString
subbu is now known as subbu|afk
<kares> I do not fear its too much info, would actually like to add a reflective inspect later -> to be able to do smt similar to apache's reflectiveToString builer
<kares> * builder
<kares> this is not part of the PR as it is now
<headius> jmalves: there are some type-specific collections out there that should work transparently both ways...but this is a down side of erasure obviously
<headius> kares: might be interesting to add a way for us to runtime generify a collection
<headius> we can see the generic parameters for a type...so you might tag a proxied Java collection with some fill in the blank values for those type parameters
<headius> kares: so this just connects inspect to toString, ok
<headius> at one point we had proposals for inspect that does reflectively crack the object but that was always a little problematic too
<headius> if only JavaBeans had really taken off, then we could just use bean inspection for everything :-D
<headius> kares: I don't have any objection
<headius> the inspect output as it is now is basically useless
<jmalves> headius: Yeah, but this is not very bothersome it is for a spec and besides there the Map stays in Java
<headius> yeah
<headius> ok
<headius> jmalves: fwiw to_java(:int) works too
<headius> for all the standard primitive types and :string
<headius> yay, mkristian has patches to fix the property ordering and builds on 11
<headius> chatting on gitter
<jmalves> :+1: I adopted that suggestion
<headius> oh yeah I said that above :-)
<headius> coffee hasn't kicked in
<headius> kares: once we can build on 11 it will make the process smoother obviously
<headius> we still have those stupid ByteBuffer issues though
<headius> I really don't want to explicitly cast everything to Buffer but I'm not sure how else to do it
<headius> other than requiring Java 8 class library to build
travis-ci has joined #jruby
<travis-ci> jruby/jruby (mavengem:8689426 by Christian Meier): The build has errored. (https://travis-ci.org/jruby/jruby/builds/440661924)
travis-ci has left #jruby [#jruby]
niko has joined #jruby
subbu|afk is now known as subbu
bga57 has joined #jruby
subbu is now known as subbu|away
<headius> kares, enebo: I think this is a blocker for 9.2.1 also: https://github.com/jruby/jruby/issues/5267
<kares> nice, seen that one before
<kares> there was a kword fix before ... maybe its all good now
Puffball has quit [Remote host closed the connection]
Puffball has joined #jruby
<headius> I'm hoping yahonda can help us narrow it down because I have no idea how to get an Oracle instance up and running for that thing
travis-ci has joined #jruby
<travis-ci> jruby/jruby (mavengem:6a18401 by Christian Meier): The build passed. (https://travis-ci.org/jruby/jruby/builds/440663866)
travis-ci has left #jruby [#jruby]
<headius> nor do I really want to pollute my system with an Oracle install
<headius> hmm there's docker images
travis-ci has joined #jruby
<travis-ci> kares/jruby (master:388bbbb by kares): The build has errored. (https://travis-ci.org/kares/jruby/builds/440670046)
travis-ci has left #jruby [#jruby]
<headius> kares: can you build on 11 with mkristian stuff?
<headius> I'm still getting the jossl StringIndexOOB
<headius> oh wait maybe I was behind?
<headius> yeah tain't working
travis-ci has joined #jruby
<travis-ci> jruby/jruby (master:5448a3f by Charles Oliver Nutter): The build has errored. (https://travis-ci.org/jruby/jruby/builds/440685880)
travis-ci has left #jruby [#jruby]
<headius> what's going on with travis lately
<headius> sigh
<lopex> what a run
<lopex> are there any maven not dead yet plugin forks from torquebox ?
<kares> CI has issues with a shallow git check-out
<kares> if you push to builds on master the first one will always fail
<kares> should be an easy fix
<headius> travis seems to be having problems with merged branches
<headius> like it's trying to check out head of the branch
<headius> kares: bleh...maybe I should remove the shallow clone again
<headius> it defaults to 50 normally but I choked it down too much maybe
<kares> yep that should be it
<headius> I'll bump it up to 10, that should be good enough for branch commits to not go all weird
<headius> merge commits
<headius> @kares that issue is fixed in jossl right?
<headius> the index OOB
<headius> builds fine on 9
xardion has quit [Remote host closed the connection]
<headius> kares: you are on linux yeah?
xardion has joined #jruby
travis-ci has joined #jruby
<travis-ci> jruby/jruby (master:920193e by kares): The build has errored. (https://travis-ci.org/jruby/jruby/builds/440689724)
travis-ci has left #jruby [#jruby]
<kares> headius: yes
<kares> issue was fixed with 0.10.1
<kares> which is shipped both with 9.1.17 and 9.2
<kares> the thing is it was only failing on 9/10 'pre' releases
<kares> on final 9/10 it was fine
<kares> its picking up 9.1.8.0
<kares> so that must be the issue
<kares> jruby-core/9.1.8.0/jruby-core-9.1.8.0.jar
<headius> it's working for me now 😳
<headius> I didn't do anything
<headius> I'm guessing there's some soft-versioned dependency that had not propagated in maven
<kares> hmm
<kares> okay I will try to cleanup Travis' cache
<kares> btw. I also had it working locally while having ~ same Java as CI
<kares> it seemed really weird - but I was definitely able to run some tests
<headius> we'll see how your latest 11 CI build goes
<headius> oh yeah it failed
<headius> same issue
<headius> I literally changed nothing locally and suddenly 11 worked so I'm really confused
<headius> I also confirmed with a clean checkout
travis-ci has joined #jruby
<travis-ci> jruby/jruby (master:920193e by kares): The build has errored. (https://travis-ci.org/jruby/jruby/builds/440689724)
travis-ci has left #jruby [#jruby]
<kares> headius: going to clean CI cache and restart build to see
<headius> kares: I'm wondering if it's caching the gems for lib build and then 11 works from then one
<headius> I failed on 11, succeeded on 9 and 10, and then succeeded on 11
<kares> oh yeah that might explain it
<headius> mkristian is on gitter...I suggested that gems are caching so it doesn't try to fetch again, but he thought it should always still be loading jossl
<kares> but as I was talking with Kristian its a loading issue
<headius> yeah
<kares> jossl isn't really used with maven plugins
<kares> it just gets loaded due rubygems doing a require
<headius> right
<kares> and now there actually should be a dummy openssl.rb
<headius> hmm
<headius> it does beg the question why we can't update the plugins to 9.2
<kares> headius: I might have been the one against it
<kares> there's this issue with Gem.path not working ... which I am not sure if is a real problem
<headius> hmm
<kares> but if it is it would go bad for sure with the embedded maven setup
<headius> At least bump to a version of JRuby that we know runs properly on 9+
<headius> 9.1.8 is not that
travis-ci has joined #jruby
<travis-ci> jruby/jruby (ci-jdk-11:46606c0 by kares): The build is still failing. (https://travis-ci.org/jruby/jruby/builds/440689907)
travis-ci has left #jruby [#jruby]
<kares> headius: that is just wrong ... an upgrade was done to 9.1.17
<kares> so this is smt not picking up properly
<kares> it failed with a clean cache on CI as well
<kares> will try locally
<headius> I just reproduced again locally by wiping out .m2/repo/rubygems
<headius> so CI could work once cache is populated, but that's obviously not working right
emerson has joined #jruby
<headius> grr
<headius> so close
<kares> failing for me as well
<kares> thinking it must be due polyglot maven stuff
<kares> since it happens early on - before downloading gems
<kares> guess I might have confused that during the bug report ;(
<kares> oh no wipe that
<kares> [INFO] Downloading from mavengems: mavengem:http://rubygems.org/com/github/jnr/jffi/maven-metadata.xml
<kares> LoadError: load error: jopenssl/load -- java.lang.StringIndexOutOfBoundsException: begin 0, end 3, length 2
<headius> same
travis-ci has joined #jruby
<travis-ci> jruby/jruby (master:b2f694c by Charles Oliver Nutter): The build has errored. (https://travis-ci.org/jruby/jruby/builds/440700877)
travis-ci has left #jruby [#jruby]
shellac has quit [Ping timeout: 252 seconds]
travis-ci has joined #jruby
<travis-ci> jruby/jruby (master:b2f694c by Charles Oliver Nutter): The build passed. (https://travis-ci.org/jruby/jruby/builds/440700877)
travis-ci has left #jruby [#jruby]
<lopex> headius: I'm going to spin another jcodings release since two code range tables were badly named
<headius> kares: what version of jruby includes jossl that works on 11?
<headius> lopex: ok sure
<kares> headius: >= 9.1.17
<kares> maybe even 9.1.16
<headius> 😡
<headius> so I updated polyglot plugin because mkristian thought the 0.3.1 version was pulling in 9.1.8.0
<headius> but I'm still getting the same error when it tries to fetch mavengem
<headius> mavengems
travis-ci has joined #jruby
<travis-ci> kares/jruby (master:920193e by kares): The build has errored. (https://travis-ci.org/kares/jruby/builds/440717604)
travis-ci has left #jruby [#jruby]
<headius> kares: frustrating
<headius> I don't understand where this error is coming from during mavengem fetch
<headius> I'm sure it's using 9.1.17 now
<kares> headius: yeah I just read your gitter chat
<kares> too much embedding and maven of course ;)
<headius> I know
<headius> I'd hate to have to switch another build tool but this has become a bottleneck
<lopex> polyglot itself ?
<headius> well it's a combination of polyglot and other jruby plugins
<headius> hopefully once we get this all working on 11 we'll be fine for a good long while
<headius> looks like we have it working now though
<headius> kares: 1.0.3 of maven-wagon released, seems to work
<headius> I'll push to master once I see it propagate
subbu|away is now known as subbu
<headius> kares: master should build on 11
<headius> by moving jruby itself into module-path many of the warnings go away
<lopex> magic
<headius> yeah wtf magic
<headius> I even asked mreinhold about it and he seemed confused
<headius> nobody knows how anything works
<lopex> and yet we're at the very beginning
<lopex> enebo: !!
<enebo> lopex: !!!!!
<lopex> enebo: my brother at one of his first larger scale beer brewing https://www.facebook.com/BrowarGaraz/photos/a.1818099261793971/2192405991029961/?type=3&theater
<lopex> enebo: doesnt it look cool ?
<lopex> he's all into it now
<enebo> yeah he looks "all in". That is a lot of fermentation
<lopex> into the beer thing
<enebo> lopex: I need to buy a new fermenter. I broke the one I liked
<lopex> how can you brake z fermenter ?
<lopex> *a
<lopex> er, wrong window
<headius> it was a plastic (?) conical fermenter and he drop kicked it in his kitchen
<lopex> oh
travis-ci has joined #jruby
<travis-ci> jruby/jruby (master:09fd1b7 by Thomas E. Enebo): The build was broken. (https://travis-ci.org/jruby/jruby/builds/440805115)
travis-ci has left #jruby [#jruby]
<enebo> lopex: headius: yeah dropped and bottom is plastic pieces threaded into each other...pushed the threads sideways and I don't think I can trust this thing if I re-thread them again
<lopex> ah
<enebo> It has about 3 gallons of portable brewers wash in it at the time. I was lucky I got it into the sink without needing a lot more cleaning equipment
<lopex> yeah re-threading is bad
<enebo> probably about .75 gallons on the floor
<enebo> lopex: I am looking at a stainless 7 gallon one now
<enebo> lopex: but I am debating how much fun it will be to carry and clean
<lopex> enebo: probably much costlier ?
<lopex> steel isnt cheap right ?
<enebo> yeah the plastic ones are ~100$ give or take and this one I think is 300 but now I cannot see it on their website
<enebo> It was there before
<enebo> and 300$ is actually pretty cheap for stainless but this is still a really nice unit
<enebo> oh drat 400$
<lopex> enebo: just toss it with metabisulfite ?
<enebo> Although I see this has domed cover so I am hopnig the one I saw will be on sale...maybe that is why I think it is 300
<enebo> I know they revved this unit from a review
<lopex> enebo: the other thing that came out as a serious cost was steerable gas valves
<lopex> enebo: you can buy rpi and whatever for what ?
<lopex> but the valves
<enebo> This thing empty is about 6kg and I brew ~3 gallon batches so it should only be carrying about 18-20kg down a flight of stairs
<lopex> enebo: do you boil on gas too ?
<enebo> lopex: I do on the stove
<enebo> lopex: I have a 7 gallon steel kettle
<lopex> lolz
<enebo> lopex: I brew in a bag
<enebo> so I don't even "properly" mash but it is technically all-grain brewing
<enebo> speaking of which my wet hop beer has been bottled for 16 days. I should put one in the fridge and try it
<lopex> enebo: what's wet hop ?
<lopex> without dry hopping ?
<lopex> enebo: we they might call it other ways in polish here
<lopex> so I might be mixing some terms
<lopex> enebo: dry hopping is just adding hops at the very end right ?
<enebo> lopex: yeah I did dry hop as well :P
<enebo> lopex: wey hop beer is using fresh hops right off the vine
<lopex> enebo: they call is cold hopping if you;d translate the beer sland
<lopex> *it
<enebo> lopex: so I picked hops from a vine in my yard and threw em in when I did a flame out
<enebo> lopex: but after I fermented a week I also added hops to dry hop it
<lopex> ok
<enebo> confusing terminology
<lopex> yeah
<lopex> and also the words themselves change
<enebo> lopex: I also see people refer to wet hop as fresh hop
<lopex> aaah
<lopex> ok
<lopex> now I got it
<enebo> lopex: but I see both phrases and I heard wet hop before fresh hop
<lopex> which is directly transletable as "cold hopping"
<lopex> enebo: trying to relate to english now
<lopex> just adding hops after just for the aroma
<lopex> enebo: but there's no cold hopping term in right ?
<lopex> en EN world
<lopex> *in
<lopex> enebo: how do you call fresh hops ?
<lopex> enebo: unprocessed
<lopex> same ?
<lopex> or that "wet"
<lopex> I'm becoming more confused now I have to admit
<enebo> wet as in unprocedded
<enebo> heh unprocessed
<lopex> ok, so I got that right
<enebo> yeah otherwise I would dry them out in an oven for later (which I have never done)
<lopex> so we dont even seem to have that term since most of the hops we have are imported
<lopex> and it's overseas mostly
<enebo> I largely buy mine as well and even this beer I used plenty of purchased hops
<lopex> enebo: so underpressurised ones are definitely processed right ?
<enebo> you mean whole flower?
<enebo> there is leaf, plug, and pellets I think
<lopex> enebo: so it's all grinded ?
<lopex> enebo: I dont know any of this
<enebo> picture of pellets
<lopex> ok
<lopex> and the second ?
<lopex> ok, sa we dont have access to pellets then here
<lopex> we have some "domestic" hops but that's just a few
<enebo> lopex: I have never used plugs
<enebo> lopex: pellets are the main thing since they are super small and don't suck up a lot of water
<lopex> enebo: they seem to only use plugs here it seems
<lopex> ok
<lopex> enebo: so you're lucky in that respect then
<enebo> lopex: yeah home brewing really has been a thing in the US for decades so I think we have a lot of options
<lopex> or they might be costly here
<lopex> enebo: in eu is just like a two or three "national" hops available per country
<enebo> lopex: but a lot of this also is connected to craft brewing popularity where US has had one a bit longer
<lopex> and you have plenty
<lopex> yeah
<enebo> lopex: we keep making new ones too but they are all proprietary
<lopex> I read a bit about Us beere
<lopex> enebo: and yet like 6 years ago I thought US was bad at beer
<enebo> US from non-US always gets this bad opinion of US beer but in spite of all the cheap adjunct lagers we drink we probably have the most diverse beer scene in the world now
<lopex> enebo: all propaganda :P
<enebo> lopex: an old reputation...like Java is slow
<lopex> enebo: but I remember some shortage of hops in the use in like 60ies so they use japansese ones like sorachi
<lopex> enebo: is there truth in that story ?
<lopex> *in US
<enebo> I have never heard it but it is possible since pacific northwest is major producer of hops
<enebo> Maybe they had a horrible year for hops
<lopex> enebo: in US ofc
<enebo> Wisconsin in 19th century used to be dominant state for hops...not sure how many hops even exist there now
<enebo> but Minnesota/Wisconsin also had a lot of German settlers
<enebo> lopex: there definitely are years where particular hops are hard to come by
<lopex> enebo: and they are still trying to mix all of them
<lopex> genetically
<lopex> enebo: I still love citra mostly for ipas
<enebo> I like citra
<lopex> and then some simcoe mix
<lopex> cascade is too weak for me
<enebo> I like amarillo and simcoe and cascade
<enebo> but I do definitely believe mixing some of these really helps
<lopex> but citra is very strong
<lopex> yeah
<lopex> enebo: that mosaic I dont realy remember what that was
<headius> there was a big hop shortage just a few years back
<lopex> enebo, headius I hate that there's some demand on gose now
<lopex> I dont really dislike it
<lopex> it's just not my thing
<lopex> I like most barleys
<lopex> but I'm stil an ipa guy
<headius> hah yeah so that's your current phase there?
<lopex> headius: no, I'm ipa since four years or so
<headius> I mean in Poland, lots of goses?
<headius> it's all NE IPA here right now, to an annoying degree
<lopex> headius: heay, now it seems
<headius> enebo: those bash changes seem ok?
<lopex> headius: but after so many ipas I can really give a value to a good lager
<headius> I am now wondering if there's a security concern about turning off bytecode verification for everything
<lopex> good lager can be good
<headius> but I don't know how to turn it off just for boot classes
<lopex> headius: and what about windows bats ?
<headius> ugh
<headius> we ship a .exe so I dunno about the scripts
<lopex> should be one to one right ?
<headius> yeah we still have a .bat
<headius> I guess this could be powershell on pretty much all extant Windows, no?
<lopex> headius: and ppl will want to set envs for a process on windows still
<lopex> ugh
<lopex> I wonder what's ps adoption really
<headius> oh whew
<headius> jruby.bat just calls jruby.exe
<headius> so it's just there for backward compat or something
<lopex> and how it interacts when executed from ps script ?
<lopex> wrt env vars
<lopex> does any one know ?
<enebo> headius: the ones we were talking about earlier with the java9 -d detect thing?
<headius> yeah
<headius> hmph
<lopex> er
<headius> I don't like the disabled verification
<lopex> java9 is a label for >8 now right ?
<enebo> headius: no verify is something that may be nice to disable for testing at least
<enebo> I am less worried about our released jars causing strife from no verification and I guess if you are using lots of other released jars in jruby hopefully someone is verifying those in their test envs
<enebo> I am not really worried about security ramifications of it
<enebo> but we once released something which ran but would not verify in the past
<enebo> from what I recall we just would catch and interpret
<headius> well the vector would be someone figuring out a bug in our or someone else's bytecode generation and getting JVM to load bad bytecode that does something bad somehow
<headius> it does seem distant
<lopex> that ver is just boot cp or noverify ?
<enebo> if we test out jars with verify on then it is still possible we have an error but we run enough where it is pretty unlikely. Including random foreign jars would be my main question mark
<enebo> but most of those are from released projects which probably run with verify on normally
<enebo> I guess I don't know :)
<lopex> headius: you mean you fear an exploit of some sort ?
<headius> I guess
<headius> poking around online there's a few pages that say NEVER DO IT OMG
<lopex> but jruby was always on boot cp right ?
<lopex> well, almost always
<headius> right...but that's not going to work for module stuff in 11
<lopex> ah
<headius> or at least I think it violates something
<lopex> ok
<headius> the patch there basically switches from using bootclasspath + classpath to -Xverify:none + module path + classpath
<headius> getting us on module path is at least a minimum step to start doing module stuff above board
<lopex> the whole thing seems quite fragile, are they allowed to change anything through 11 line ?
<lopex> headius: there's just module path and no opotions for verification for it alone ?
<lopex> madness
<headius> well, I think the idea is that you use appcds or jlink or whatever other stuff and it prelinks all this stuff so verification is done
<lopex> ok..
<headius> with appcds I can leave verification on, it doesn't really impact things much
<lopex> yeah
<lopex> unless it doesnt work
<lopex> wrt persistance
<headius> don't know if I told you I put this gem out
<headius> oh heh, maybe I should do somethign with that readme
<lopex> and cds is being kept where ?
<lopex> can it be share across jvms ?
<lopex> shared
<lopex> or globally in some parts ?
<lopex> so many questions
<headius> better
<headius> I'm not sure about the sharing but I don't think so
<headius> that's why I'm making this utility for now, to make it easy for anyone to test it