|
01:29
MasterDuke joined
03:54
MasterDuke left
06:41
Tux__ left
06:51
|Tux| joined,
Tux__ joined
07:49
Tux__ left,
camelia left
07:50
rypervenche left
07:52
rypervenche joined
08:44
camelia joined
10:59
|Tux| left
11:04
|Tux| joined
11:19
finanalyst joined
11:58
finanalyst left
|
|||
| [Coke] | up to 81.1% :P | 13:19 | |
| 49 hours, 82.6% | 15:00 | ||
| Sorry, I realize no one wants this level of detail. :) | 15:01 | ||
| timo | salsa.debian.org/perl6-team/moarvm...requests/4 i have this merge request that is clean except for a thing that is allowed to fail. i wonder what i need to do before i should bring it up to Dominique | 15:02 | |
| huh it says that lintian is unhappy about the changelog format but i thought i fixed it tho | 15:04 | ||
| ah, got it | 15:05 | ||
| [Coke] | I made it to salsa.debian.org/timotimo-guest/mo...est_report but not seeing where the actual errors are. | 15:07 | |
| timo | you click on "lintian", then for the moarvm and moarvm-dev categories there's a "view details" button on the right | 15:09 | |
| but the "failed jobs" tab near the top also has a snippet of build log output | |||
| oh but that's unrelated | |||
| [Coke] | pre-existing condition? | 15:11 | |
| oh, no, it *was* you. :) | 15:12 | ||
| timo | the moar package wasn't yet carefully set up to make cross-compilation possible, that's what the "allowed to fail" job is about | ||
| the lintian errors are from my hand-writing the changelog entry with its date missing the seconds | 15:13 | ||
| hm, no nmu in changelog, and source nmu has wrong version | 15:16 | ||
| that's for when a non-maintainer makes a change, which i guess i am until i am put into the right file, and it's supposed to get a -nmu1 in that case | 15:17 | ||
| [Coke] | glad you were able to hit the ground running on this today! | 15:20 | |
| timo | i'm not exactly sure if we're actually compiling with the flags that caused the error from the original bug | 15:28 | |
| To build with GCC 14, either set CC=gcc-14 CXX=g++-14 explicitly, | 15:30 | ||
| or install the gcc, g++, gfortran, ... packages from experimental. | |||
| apt-get -t=experimental install g++ | |||
| ^- i guess i have to do that | |||
| ok i think i'm confused | 15:53 | ||
| [Coke] | you got the tests to pass, anyway. | 15:58 | |
| ah. some of the processes that are on for like 25 hours at this point appear to be hung tests on this blin run. | 16:00 | ||
| Wonder if that's why it slowed down. Guessing we should set a generous timeout "if your tests don't complete in 30m, mark it and move on." | 16:01 | ||
| timo | ok, so i think i was applying the fix to the wrong version | 16:02 | |
| the more i look at what i did the wronger it looks | 16:07 | ||
| [Coke] | heh | 16:09 | |
| it seemed so reasonable. :) | |||
| timo | maybe i'm actually doing it half-correct. i just also need to start a new branch on top of the debian/2022.12+dfsg-1 tag where the patch also needs to go | 16:11 | |
| and i assume i have to transplant the change log entries from timeline-wise newer versions into the changes file, then add a new changelog entry for a lower version at the top | 16:13 | ||
| jdv | i only got some hung a few times in the dozens of runs ive down | 16:23 | |
| [Coke] | I just killed anything that was 20+ hours old. | 16:24 | |
| jdv | and i think it was because a large change in rakudo caused essentially a mem leak or something with a dist... | ||
| good times | |||
| timo | i forgot to put the patch in the patches/series file haha | 16:28 | |
| i guess i made that much more of an ordeal than it had to be | 17:03 | ||
|
17:44
sena_kun joined
|
|||
| [Coke] | so is the PR ready for his review at this point (saw the email but wasn't sure if that included everything) | 18:10 | |
| timo | i believe that it is ready | 18:24 | |
| in order for getting moar off the track to expulsion, the branch mentioned in a comment in the merge request also needs to have someone go in and turn it into a debian package and upload it, is my understanding | 18:26 | ||
| i think we can't just willy-nilly put a much newer version in an older debian release | |||
| that's why i'm patching on top of previous releases | |||
| getting latest upstream versions of moar, nqp, and rakudo out is a step later on the roadmap. these new versions have to go through maybe experimental before unstable, but definitely from unstable through the natural process of aging into stable and oldstable | 18:27 | ||
| and testing is in between there also | |||
| maybe i should verify with Dominique that this is actually accurate | 18:29 | ||
| but for now, a nap | |||
| [Tux] | Today I upped a openSUSE 15.4 to 15.5. The most smooth upgrade ever! BUT… | 18:40 | |
| Starting applications that used "advanced" unicode cause all kind of weird crashes! | 18:41 | ||
| I had to manually install libXft-newest from some non-standard repo to make the crashes go away | 18:42 | ||
| why is 15.5's libXft so "old"/incompatible with modern unicode? | |||
| I disable *a lot* of repo's off-standard this time in my upgrade to stay as close to standard as possibel | 18:43 | ||
| (another system is now updating from 15.4 to 15.6, I'm afraid to hit the same issue) | 18:44 | ||
| just 6053 packages to go before th next command | 18:45 | ||
|
19:29
finanalyst joined
21:05
kjp left
22:23
finanalyst left
22:34
sena_kun left
23:06
MasterDuke joined
23:13
kjp joined
|
|||