tdaitx | doko: I forgot to ask you if you had any expectation on rebasing this GTK change on top of the disco version (11.0.3) | 01:06 |
---|---|---|
tdaitx | for now I'm uploading a fixed version for bionic/cosmic on top of the versions in proposed (11.0.2) | 01:06 |
tdaitx | from what I have seen 11.0.3 is still in RDP2, so not yet released... an email from early March said it was going to be released "in 6 weeks" which is about middle April | 01:06 |
tdaitx | oh my, disco's lintian is taking ages to run on big projects like scilab and openjdk... | 01:20 |
tdaitx | what a pain | 01:20 |
tdaitx | and it is up! | 01:24 |
tdaitx | openjdk-lts 11.0.2 uploaded to stage5 for bionic and cosmic (will check tests tomorrow when build is done, I need to look for regressions in the jdk testsuite) | 01:24 |
tdaitx | openjdk-lts 11.0.3 uploaded to disco | 01:24 |
tdaitx | disco needs approval due to the ff | 01:25 |
doko | tdaitx: are you really sure that no gtk is needed for the tests during the build? | 06:41 |
doko | $ fgrep -ri gtk test/|wc -l | 07:00 |
doko | 165 | 07:00 |
doko | keeping the b-d | 07:04 |
doko | tdaitx: why stage5 instead of stage3? (no, don't re-upload) | 10:06 |
tdaitx | doko: yeah, it does not need gtk dev for build-depends, I'm sure | 13:33 |
tdaitx | the difference between the previous bionic/cosmic builds (gtk3, gtk-dev) and the new one (gtk2|gtk3 and no gtk dev) are mostly rmi tests... the gtk failures you mention were already there before | 13:33 |
tdaitx | I just checked the buildlogs and compared the tests, all seems fine | 13:35 |
tdaitx | and while I didn't mention it before, I did a build without tests before and ran jabref and the jfilechooser demo to check if GTK2 was running and how it looked, all was fine | 13:36 |
tdaitx | also tested gtk3, which is possible to enable through -Djdk.gtk.version=3 | 13:37 |
tdaitx | yeah, it was looking better with upstream patches but still broken for padding and at least 2 swing components | 13:37 |
tdaitx | and I uploaded to stage5 to avoid overwriting our packages in stage3, just in case you catched any issue that I might have missed | 13:39 |
doko | well, we have the copy in -proposed | 13:41 |
doko | but now copied to bionic and cosmic | 13:42 |
tdaitx | thanks! | 13:42 |
tdaitx | but seriously, feel free to remove the gtk deb from d-beps | 13:42 |
tdaitx | gtk-dev | 13:42 |
tdaitx | openjdk only cares about runtime, they declare everything they need in their own headers | 13:43 |
doko | what about the tests, will those be skipped? | 13:44 |
tdaitx | no, they will use the library that is available, the tests don't need the headers | 13:46 |
doko | tdaitx: what about the unattended-upgrades test? | 13:47 |
tdaitx | I need to get back to it | 13:47 |
tdaitx | I was going to take a look at openjfx before, it might also need to get the gtk preference reverted | 13:48 |
doko | ok | 13:51 |
doko | tdaitx: can the disco delta for ca-certificates-java be dropped? | 14:06 |
tdaitx | doko: ok, I tracked down this and I think we should be able to propose the --release flag to debian now | 14:47 |
tdaitx | as this is basically the only pending delta we have | 14:47 |
tdaitx | the explicit b-dep we had on openjdk-11 can be replaced with the default-jre-headless that debian is using | 14:50 |
doko | crap I copied the lts bionic package too early | 14:59 |
tdaitx | doko: want me to upload openjdk-lts with a bumped version for bionic? | 15:30 |
tdaitx | the worst I have to go is waiting for lintian to run through | 15:30 |
doko | tdaitx: not yet. waiting for the britney run | 15:30 |
tdaitx | doko: btw, swing.properties is still being copied to lib/ but code now expects it in conf/ | 15:35 |
tdaitx | uploaded new version into the test ppa: bumped version and updated the swing.properties path from lib/ to conf/ | 16:45 |
tdaitx | btw, pdfsam (which uses openjfx) is loading gtk2 even without changes to openjfx, so we might not need that revert | 16:47 |
doko | tdaitx: we can do that later with the next security release. the copy in -proposed is fine, and autopkg tests are running | 16:55 |
doko | just checked that it's not a regression | 17:22 |
tdaitx | ah, ok, I thought we would have to upload it again after cjwatson said, so I just did just that into the test ppa just in case | 18:51 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!