Bug#893244: jruby FTBFS with openjdk-9
On Tue, Feb 26, 2019 at 03:05:48PM +0100, Markus Koschany wrote: > > [...] > > The FTBFS bug got fixed yesterday. I should complain more often. Andrej > uploaded version 9.1.17 to unstable. This is not the latest one but I > guess better than nothing? The original bug has not been closed yet. > Andrej, can we close it now and Debian bug #917702 too? Thanks a lot to Andrej for putting together a new upstream version upload at such short notice!! Now I think with that should be enough to fulfill the reverse build-depends to avoid having to drop packages from the upcoming release but I reiterate that jruby in its current state in Debian is only usable for very simple use cases. The output from many tests during build time is evidence of this. Any user expecting to run production workloads with Debian's jruby will be disappointed and they will require to use upstream artifacts, unfortunately. Thanks again for fixing the outstanding RC bugs on this package!! -- Miguel Landaeta, nomadium at debian.org secure email with PGP 0x6E608B637D8967E9 available at http://miguel.cc/key. "Faith means not wanting to know what is true." -- Nietzsche signature.asc Description: PGP signature
Bug#893244: jruby FTBFS with openjdk-9
Hi, On Tue, 26 Feb 2019 at 15:20, Emmanuel Bourg wrote: > Le 26/02/2019 à 15:05, Markus Koschany a écrit : > > The FTBFS bug got fixed yesterday. I should complain more often. Andrej > > uploaded version 9.1.17 to unstable. This is not the latest one but I > > guess better than nothing? The original bug has not been closed yet. > > Andrej, can we close it now and Debian bug #917702 too? > > Any hope to package JRuby 9.2.x? AFAIK it supports Java 11 better. I may look into it when I have time, but at the moment it’s lower on my todo list. We’re fighting with a lot of Java-related build failures in Apertis, so I’m basically doing what’s most important to unblock our further work. -- Cheers, Andrej
Bug#893244: jruby FTBFS with openjdk-9
Le 26/02/2019 à 15:05, Markus Koschany a écrit : > The FTBFS bug got fixed yesterday. I should complain more often. Andrej > uploaded version 9.1.17 to unstable. This is not the latest one but I > guess better than nothing? The original bug has not been closed yet. > Andrej, can we close it now and Debian bug #917702 too? Any hope to package JRuby 9.2.x? AFAIK it supports Java 11 better. Emmanuel Bourg
Bug#893244: jruby FTBFS with openjdk-9
Hi, Am 25.02.19 um 19:40 schrieb Miguel Landaeta: > Hi Markus, > > On Sun, Feb 24, 2019 at 08:39:48PM +0100, Markus Koschany wrote: >> JRuby is a mess. I guess we "just" need to package the latest upstream >> release to fix the FTBFS bugs. Nobody felt like doing that in the past >> twelve months, so I think it is unrealistic to believe we can make it >> happen within one week. Of course if the release team can be convinced >> to accept a new upstream release there might be additional time left. > > I agree, jruby is a mess, mostly because of me since I didn't have > almost any time during this release cycle to work on it. > > I think jruby should be dropped from buster and a new libspring-java > upload should be prepared shortly, to disable jruby support in it. > > It's not realistic to think that a new upstream release for jruby can > be prepared in a week and that will work to be well supported during > the next stable release life cycle. > > Cheers, > Miguel. The FTBFS bug got fixed yesterday. I should complain more often. Andrej uploaded version 9.1.17 to unstable. This is not the latest one but I guess better than nothing? The original bug has not been closed yet. Andrej, can we close it now and Debian bug #917702 too? Regards, Markus signature.asc Description: OpenPGP digital signature
Bug#893244: jruby FTBFS with openjdk-9
Hi Markus, On Sun, Feb 24, 2019 at 08:39:48PM +0100, Markus Koschany wrote: > JRuby is a mess. I guess we "just" need to package the latest upstream > release to fix the FTBFS bugs. Nobody felt like doing that in the past > twelve months, so I think it is unrealistic to believe we can make it > happen within one week. Of course if the release team can be convinced > to accept a new upstream release there might be additional time left. I agree, jruby is a mess, mostly because of me since I didn't have almost any time during this release cycle to work on it. I think jruby should be dropped from buster and a new libspring-java upload should be prepared shortly, to disable jruby support in it. It's not realistic to think that a new upstream release for jruby can be prepared in a week and that will work to be well supported during the next stable release life cycle. Cheers, Miguel. -- Miguel Landaeta, nomadium at debian.org secure email with PGP 0x6E608B637D8967E9 available at http://miguel.cc/key. "Faith means not wanting to know what is true." -- Nietzsche signature.asc Description: PGP signature
Bug#893244: jruby FTBFS with openjdk-9
JRuby is a mess. I guess we "just" need to package the latest upstream release to fix the FTBFS bugs. Nobody felt like doing that in the past twelve months, so I think it is unrealistic to believe we can make it happen within one week. Of course if the release team can be convinced to accept a new upstream release there might be additional time left. In case a new upstream release or a similar patch won't happen in time, here is the list of reverse-dependencies: reverse-depends -b jruby Reverse-Build-Depends-Indep === * libspring-java * mustache-java Reverse-Build-Depends = * jruby-maven-plugins * jruby-openssl * libfreemarker-java * ruby-psych * yecht I presume every package with "ruby" would be expendable. The rest needs a closer look but I hope jruby is always optional. Markus signature.asc Description: OpenPGP digital signature
Bug#893244: jruby FTBFS with openjdk-9
Source: jruby Version: 9.1.13.0-1 Severity: serious https://tests.reproducible-builds.org/debian/rb-pkg/unstable/amd64/jruby.html ... [ERROR] COMPILATION ERROR : [INFO] - [ERROR] /build/1st/jruby-9.1.13.0/core/src/main/java/org/jruby/management/BeanManagerImpl.java:[87,26] error: cannot find symbol symbol: class Agent location: package sun.management
