Bug#950272: buster-pu: package mariadb-10.3 10.3.22-0+deb10u1
Hi All, Paul Gevers ezt írta (időpont: 2020. febr. 6., Cs, 20:17): > > Hi all, > > On 06-02-2020 15:11, Paul Gevers wrote: > > Yes, and I am worried about that. If the MariaDB upgrade causes kodi to > > not run, that's a bad regression, don't you think? > > > > kodi has a pretty good history in Ubuntu on ppc64el, so I don't expect > > it to be flaky at first sight. > > I'm unhappy with the result, but the reference run with the old MariaDB > now fails as well, so, no regression. > > I'm not sure what this means for kodi on ppc64el though. Thank you for looking into the regression. I believe the user-base of Kodi on ppc64el is very small, thus it is not really worth the effort to fully triage this if this is not a symptom of a MariaDB regression. Cheers, Balint > > Paul >
Bug#950272: buster-pu: package mariadb-10.3 10.3.22-0+deb10u1
Hi all, On 06-02-2020 15:11, Paul Gevers wrote: > Yes, and I am worried about that. If the MariaDB upgrade causes kodi to > not run, that's a bad regression, don't you think? > > kodi has a pretty good history in Ubuntu on ppc64el, so I don't expect > it to be flaky at first sight. I'm unhappy with the result, but the reference run with the old MariaDB now fails as well, so, no regression. I'm not sure what this means for kodi on ppc64el though. Paul signature.asc Description: OpenPGP digital signature
Bug#950272: buster-pu: package mariadb-10.3 10.3.22-0+deb10u1
Hi Otto, On 06-02-2020 14:10, Otto Kekäläinen wrote: >> tail: cannot open '/home/debci/.kodi/temp/kodi.log' for reading: No such >> file or directory >> tail: no files remaining > > This is the only output in the test log, otherwise it is installing > packages and everything else seemed to go fine. > > Test source says it just starts the app, but now it does not seem to > output anything: > https://salsa.debian.org/multimedia-team/kodi/blob/master/debian/tests/gui Yes, and I am worried about that. If the MariaDB upgrade causes kodi to not run, that's a bad regression, don't you think? kodi has a pretty good history in Ubuntu on ppc64el, so I don't expect it to be flaky at first sight. Paul signature.asc Description: OpenPGP digital signature
Bug#950272: buster-pu: package mariadb-10.3 10.3.22-0+deb10u1
> tail: cannot open '/home/debci/.kodi/temp/kodi.log' for reading: No such file > or directory > tail: no files remaining This is the only output in the test log, otherwise it is installing packages and everything else seemed to go fine. Test source says it just starts the app, but now it does not seem to output anything: https://salsa.debian.org/multimedia-team/kodi/blob/master/debian/tests/gui
Bug#950272: buster-pu: package mariadb-10.3 10.3.22-0+deb10u1
Hi Otto, kodi maintainers, On 02-02-2020 11:56, Otto Kekäläinen wrote: > Ok, I will amend the changelog and re-upload today 1:10.3.22-0... to > buster updates. The autopkgtest of kodi fails on ppc64el (twice) with the p-u version of MariaDB, while it passes with the old MariaDB version. I'll retrigger it for the third time, but maybe you want to have a look already. Paul [stable result] https://ci.debian.net/user/britney/jobs?package=kodi=mariadb-10.3%5B%5D=stable%5B%5D=arm64%5B%5D=ppc64el [failing log] https://ci.debian.net/data/autopkgtest/stable/ppc64el/k/kodi/4212994/log.gz autopkgtest [20:24:43]: test gui: [--- tail: cannot open '/home/debci/.kodi/temp/kodi.log' for reading: No such file or directory tail: no files remaining autopkgtest [20:24:49]: test gui: ---] signature.asc Description: OpenPGP digital signature
Bug#950272: buster-pu: package mariadb-10.3 10.3.22-0+deb10u1
Ok, I will amend the changelog and re-upload today 1:10.3.22-0... to buster updates.
Bug#950272: buster-pu: package mariadb-10.3 10.3.22-0+deb10u1
On Sun, 2020-02-02 at 10:44 +0100, Otto Kekäläinen wrote: > Thanks for pointing out this mistake. > > Unfortunately it was already uploaded. It can still be reuploaded today. > There are frequently releases to unstable/testing so I don't expect > this to cause any bigger problems. It will cause problems next weekend unless a new upload had been uploaded to unstable and transitioned to testing before the point release on Saturday morning. Given that timescale, a second buster upload with the corrected version number seems safer. Regards, Adam
Bug#950272: buster-pu: package mariadb-10.3 10.3.22-0+deb10u1
Thanks for pointing out this mistake. Unfortunately it was already uploaded. There are frequently releases to unstable/testing so I don't expect this to cause any bigger problems. Past and future changelog entries will not have this mistake. https://salsa.debian.org/mariadb-team/mariadb-10.3/blob/buster/debian/changelog
Bug#950272: buster-pu: package mariadb-10.3 10.3.22-0+deb10u1
Hi, [Disclaimer, not part of the release team but I noticed the following] On Thu, Jan 30, 2020 at 10:06:52PM +0200, Otto Kekäläinen wrote: > Package: release.debian.org > Severity: serious > Tags: buster, moreinfo > User: release.debian@packages.debian.org > Usertags: pu > > Changelog: > > mariadb-10.3 (1:10.3.22-1+deb10u1) buster; urgency=high I think this should have been 1:10.3.22-*0*+deb10u1 to have an import of 10.3.22 on top of the buster packaging and guarantee that the version in buster is smaller than the version in unstable. We right now have: mariadb-10.3 | 1:10.3.22-1 | testing| source mariadb-10.3 | 1:10.3.22-1 | unstable | source and mariadb-10.3 | 1:10.3.22-1+deb10u1 | stable-new | source Regards, Salvatore