IRC logs for #buildstream for Tuesday, 2018-10-23

« Monday, 2018-10-22 Index Wednesday, 2018-10-24 »
*** Prince781 has joined #buildstream00:02
*** slaf has quit IRC00:12
*** slaf has joined #buildstream00:20
*** xjuan has quit IRC01:31
*** catonano has joined #buildstream04:10
*** catonano has quit IRC05:05
*** catonano has joined #buildstream05:12
*** gitlab-br-bot has quit IRC07:20
*** Prince781 has quit IRC07:34
*** gitlab-br-bot has joined #buildstream07:37
*** alatiera_ has joined #buildstream08:30
*** jonathanmaw has joined #buildstream08:49
*** solid_black has joined #buildstream08:57
gitlab-br-botrichardmaw-codethink approved MR !888 (aevri/rm-exitstack->master: plugins/sources: refactor, rm needless ExitStack-s) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/88809:16
gitlab-br-botjuergbi opened MR !890 (juerg/test-artifactshare->master: tests/testutils/artifactshare.py: Fix has_artifact() to match core code) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/89009:22
*** lachlan has joined #buildstream09:29
gitlab-br-bottpollard approved MR !890 (juerg/test-artifactshare->master: tests/testutils/artifactshare.py: Fix has_artifact() to match core code) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/89009:32
gitlab-br-botjuergbi merged MR !890 (juerg/test-artifactshare->master: tests/testutils/artifactshare.py: Fix has_artifact() to match core code) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/89009:48
tpollardwoop09:48
*** lachlan has quit IRC09:53
juergbitpollard: there is no return after line 120 in cascache.py, so it will continue to the full extraction (the 'with' block), or am I missing something?09:55
juergbi(you could probably use a single return dest, removing the else clauses, however, modifying dest in the subdir change looks wrong)09:56
*** lachlan has joined #buildstream09:59
*** anahuelamo has joined #buildstream10:02
*** anahuelamo- has quit IRC10:02
tpollardjuergbi: it will continue to the extraction, but with dest & tree set to the values of the subdir, not the whole artifact?10:03
mablanchjmac: Thanks for the review of MR !806. I've just rebased it, if you're fine with merging it then please do, I don't have write access.10:03
tpollardjuergbi: which will only ever get set if the artifact dest exists, if it exists but we can't extract the subdir, or it has already been extracted then it behaves as before10:04
jmacI will do, thanks mablanch10:04
tpollardas I see it anyhow10:04
tpollardwhich seems to work as I expect10:06
juergbitpollard: right, that should work, I misread the code before10:06
*** lachlan has quit IRC10:08
juergbitpollard: I suspect it's overall actually slower than the delete approach, though, because the fast path gets slower (it always calls _get_subdir() which reads and parses a directory protobuf)10:08
mablanchjuergbi: I'm working on build time metrics for BuildGrid at the moment, do you think you'll be able to have a look at https://gitlab.com/BuildStream/buildbox/issues/14 anytime soon-ish?10:08
juergbimablanch: it's not planned right now but it's probably not a large amount of work. the main thing is to decide on and implement some kind of channel where we can actually output metrics10:10
*** lachlan has joined #buildstream10:12
*** lachlan has quit IRC10:15
mablanchjuergbi: Agreed. I think we already discussed the possibility of having BuildBox writing a separate file for that.10:17
mablanchjuergbi: Should we mimic bwrap and https://github.com/projectatomic/bubblewrap/pull/293: add a --json-status-fd CLI option and output a JSON structure containing exit codes and timestamps?10:18
juergbimablanch: sounds reasonable. not sure whether fd or filename makes more sense10:20
mablanchjuergbi: Ok, I'll write a proposal on the issue, we can continue discussing it there.10:21
juergbita10:21
*** lachlan has joined #buildstream10:32
gitlab-br-botphildawson merged MR !888 (aevri/rm-exitstack->master: plugins/sources: refactor, rm needless ExitStack-s) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/88810:33
*** lachlan has quit IRC10:40
*** catonano has quit IRC10:51
*** catonano has joined #buildstream10:51
jmacmablanch: Sorry, !806 needs rebasing again. I did have it set to merge when the CI finished, but someone else got there first10:54
mablanchjmac: No problem. I've just rebased it again, thank you.10:55
jmacThanks mablanch10:56
*** lachlan has joined #buildstream10:56
*** lachlan has quit IRC11:03
gitlab-br-botjmacarthur merged MR !806 (mablanch/630-remote-execution-reconn->master: Handle connection losses during remote build execution) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/80611:19
*** lachlan has joined #buildstream11:21
*** lachlan has quit IRC11:31
*** catonano has quit IRC11:34
*** catonano has joined #buildstream11:35
jennishas anyone seen an error like this before when trying to use remote caching?11:46
jennisE1023 11:37:37.787300124   11904 http_proxy.cc:62]           'https' scheme not supported in proxy URI11:46
tpollardhave you set the proxy?11:54
*** lachlan has joined #buildstream11:57
*** solid_black has quit IRC11:58
coldtomjonathanmaw: i think valentind wrote the collect_integration plugin so he should probably have authorship, but MIT license seems fine11:59
jonathanmawcoldtom: okie doke11:59
jennisahh, tpollard, I thought I unset it, thanks12:03
jennisweird that it doesn't complain on another machine which also has the proxy set12:04
tpollardI think you can do something with env variables12:04
tpollardseems to ring a bell with pip, but it could be unrelated12:04
*** lachlan has quit IRC12:06
gitlab-br-botaevri opened MR !891 (aevri/include-error->master: More user-friendly reporting on include errors) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/89112:08
jonathanmawcoldtom: you submitted the git_tag plugin under your gmail, should I use that for the author, or your codethink E-mail?12:08
*** lachlan has joined #buildstream12:10
*** lachlan has quit IRC12:14
*** jonathanmaw has quit IRC12:16
*** slaf has quit IRC12:16
*** finn has quit IRC12:16
*** cs-shadow has quit IRC12:16
*** tpollard has quit IRC12:16
*** phildawson has quit IRC12:16
*** mohan43u has quit IRC12:16
*** WSalmon has quit IRC12:16
*** waltervargas[m] has quit IRC12:16
*** abderrahim[m] has quit IRC12:16
*** Demos[m] has quit IRC12:16
*** inigomartinez has quit IRC12:16
*** segfault3[m] has quit IRC12:16
*** albfan[m] has quit IRC12:16
*** alatiera has quit IRC12:16
*** persia has quit IRC12:16
*** persia has joined #buildstream12:16
*** slaf has joined #buildstream12:16
*** cs-shadow has joined #buildstream12:16
*** tpollard has joined #buildstream12:16
*** phildawson has joined #buildstream12:16
*** WSalmon has joined #buildstream12:16
*** jonathanmaw has joined #buildstream12:16
*** finn has joined #buildstream12:16
*** mohan43u has joined #buildstream12:17
coldtomjonathanmaw: the codethink one, i should probably change my git email config12:18
jonathanmawokie doke12:18
*** abderrahim[m] has joined #buildstream12:20
*** waltervargas[m] has joined #buildstream12:23
*** inigomartinez has joined #buildstream12:24
*** Demos[m] has joined #buildstream12:24
jonathanmawcoldtom: would you mind looking at https://gitlab.com/BuildStream/bst-external/merge_requests/59/diffs and checking that the copyright headers look okay to you?12:28
coldtomjonathanmaw: looks good :)12:31
* jonathanmaw mashes the merge button12:32
*** WSalmon_ has joined #buildstream12:45
*** WSalmon has quit IRC12:45
*** segfault3[m] has joined #buildstream12:47
*** johnward has quit IRC12:55
*** albfan[m] has joined #buildstream12:55
*** johnward has joined #buildstream12:55
*** phildawson has quit IRC12:55
*** phildawson has joined #buildstream12:56
*** laurence has quit IRC12:56
*** gitlab-br-bot has quit IRC12:56
*** gitlab-br-bot has joined #buildstream12:56
*** dabukalam has quit IRC12:56
*** dabukalam has joined #buildstream12:56
*** alatiera has joined #buildstream12:58
*** laurence has joined #buildstream13:00
*** lachlan has joined #buildstream13:18
*** lachlan has quit IRC13:25
tiagogomeshmm, when I checkout an artifact there are configuration files with executable permission. I am pretty sure they didn't have it when the artifact has staged13:37
*** catonano has quit IRC13:37
*** catonano has joined #buildstream13:37
*** lachlan has joined #buildstream13:40
*** finn_ has joined #buildstream13:56
*** finn has quit IRC13:57
*** anahuelamo has quit IRC14:00
*** anahuelamo has joined #buildstream14:01
gitlab-br-bottpollard closed MR !786 (tpollard/494->master: Don't pull artifact buildtrees by default) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/78614:06
tpollardI did not close it....14:07
jonathanmaw:/ looks like pexpect isn't useful for interactive testing, either, since it would require buildstream to be installed so we can invoke it via the command-line14:08
* jonathanmaw will look into calling cli.run in a subprocess and whether that'd work.14:09
tpollardis there anyway to unclose an MR?, I'm not sure what happened to the branch / mr78614:11
*** catonano has quit IRC14:11
gitlab-br-bottpollard reopened MR !786 (tpollard/494->master: Don't pull artifact buildtrees by default) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/78614:11
tpollardmagical button appeared14:12
tpollardis there an issue with the ci server, https://gitlab.com/BuildStream/buildstream/merge_requests/786/ is getting 'Could not connect to the CI server. Please check your settings and try again' for me at least14:24
* tpollard believes gitlab is trolling 14:35
*** anahuelamo has quit IRC14:48
*** anahuelamo has joined #buildstream14:52
*** bochecha has joined #buildstream15:02
*** anahuelamo has quit IRC15:02
*** anahuelamo has joined #buildstream15:06
gitlab-br-botdanielsilverstone-ct opened issue #722 (Parameterisation not used) on buildstream https://gitlab.com/BuildStream/buildstream/issues/72215:07
*** anahuelamo has quit IRC15:18
*** finn_ has quit IRC15:18
*** laurence has quit IRC15:18
*** slaf has quit IRC15:18
*** thinkl33t has quit IRC15:18
*** mablanch has quit IRC15:18
*** Nexus has quit IRC15:18
*** milloni has quit IRC15:18
*** valentind has quit IRC15:18
*** aiden has quit IRC15:18
*** flatmush has quit IRC15:18
*** phildawson has quit IRC15:18
*** albfan[m] has quit IRC15:18
*** segfault3[m] has quit IRC15:18
*** WSalmon_ has quit IRC15:18
*** waltervargas[m] has quit IRC15:18
*** tpollard has quit IRC15:18
*** cs-shadow has quit IRC15:18
*** persia has quit IRC15:18
*** flatmush has joined #buildstream15:19
*** thinkl33t has joined #buildstream15:19
*** anahuelamo has joined #buildstream15:19
*** finn_ has joined #buildstream15:19
*** laurence has joined #buildstream15:19
*** phildawson has joined #buildstream15:19
*** albfan[m] has joined #buildstream15:19
*** segfault3[m] has joined #buildstream15:19
*** WSalmon_ has joined #buildstream15:19
*** waltervargas[m] has joined #buildstream15:19
*** tpollard has joined #buildstream15:19
*** cs-shadow has joined #buildstream15:19
*** aiden has joined #buildstream15:19
*** Nexus has joined #buildstream15:19
*** mablanch has joined #buildstream15:19
*** persia has joined #buildstream15:19
*** slaf has joined #buildstream15:22
*** valentind has joined #buildstream15:23
*** milloni has joined #buildstream15:24
*** lachlan has quit IRC15:25
juergbitpollard: doesn't this cause problems the other way round, i.e., if the extract dir doesn't exist yet, it will now be created but only containing buildtrees? and next regular extract() will happily skip it even though it only contains that one subdir?15:29
*** catonano has joined #buildstream15:30
*** jonathanmaw has quit IRC15:51
*** jonathanmaw has joined #buildstream15:51
juergbiah, the subdir pull is never triggered if there is no extract dir yet15:52
juergbithere should at least be an assert or similar in pull() for this and should also be in the API contract/documentation for pull() as this is not clear at all from the point of view of the artifact cache15:53
gitlab-br-botdanielsilverstone-ct opened MR !892 (danielsilverstone-ct/fix-lint-issues->master: WIP: Fix lint issues) on buildstream https://gitlab.com/BuildStream/buildstream/merge_requests/89215:56
KinnisonPhew15:57
* Kinnison has been going through lint issues with a newer pylint than is on the CI15:57
KinnisonSo far I've made 19 commits and cleared down issues in 11 files.15:57
* Kinnison will have to continue this tomorrow15:57
*** finn_ has quit IRC16:09
*** finn has joined #buildstream16:11
*** abderrahim2 has joined #buildstream16:14
tpollardjuergbi: yes, it *should* never trigger but we probably can't guarantee it. I think adding an assert before the assumption is the correct thing to do16:15
*** abderrahim1 has quit IRC16:15
tpollardor I move some of that decision into extract()16:16
tpollardI've also modified the tests to check this behaviour over the extract dir, as it wasn't included before16:16
tpollard(checks before and after if the buildtree dir is populated in the extract dir)16:17
*** catonano has quit IRC16:25
*** Prince781 has joined #buildstream16:33
*** lachlan has joined #buildstream16:38
*** xjuan has joined #buildstream16:40
*** jonathanmaw has quit IRC17:00
*** aiden has quit IRC17:11
*** laurence has quit IRC17:11
*** johnward has quit IRC17:12
*** aiden has joined #buildstream17:13
*** johnward has joined #buildstream17:14
*** laurence has joined #buildstream17:14
*** Nexus has quit IRC17:22
*** paulsherwood has quit IRC17:22
*** jmac has quit IRC17:22
*** aiden has quit IRC17:22
*** Nexus has joined #buildstream17:23
*** paulsherwood has joined #buildstream17:23
*** adds68 has quit IRC17:23
*** jmac has joined #buildstream17:23
*** johnward has quit IRC17:23
*** laurence has quit IRC17:23
*** ikerperez has quit IRC17:23
*** laurence has joined #buildstream17:23
*** coldtom has quit IRC17:23
*** johnward has joined #buildstream17:23
*** jennis has quit IRC17:23
*** aiden has joined #buildstream17:26
*** finn has quit IRC17:27
gitlab-br-botlukas.bulwahn opened issue #723 (Use .netrc information to fetch archive files) on buildstream https://gitlab.com/BuildStream/buildstream/issues/72317:47
*** lachlan has quit IRC17:51
*** bethw has quit IRC18:00
*** milloni has quit IRC18:00
*** mablanch has quit IRC18:00
*** valentind has quit IRC18:00
*** laurence has quit IRC18:00
*** johnward has quit IRC18:00
*** aiden has quit IRC18:00
*** aiden has joined #buildstream18:00
*** milloni has joined #buildstream18:02
*** johnward has joined #buildstream18:03
*** mablanch has joined #buildstream18:03
*** xjuan has quit IRC18:03
*** valentind has joined #buildstream18:04
*** bethw has joined #buildstream18:04
*** laurence has joined #buildstream18:05
*** paulsherwood has quit IRC18:10
*** jmac has quit IRC18:10
*** Nexus has quit IRC18:10
*** benbrown has quit IRC18:10
*** xjuan has joined #buildstream18:20
*** raoul has quit IRC18:31
*** benbrown has joined #buildstream18:39
*** catonano has joined #buildstream18:47
*** catonano has quit IRC19:06
*** Prince781 has quit IRC23:28
*** alatiera_ has quit IRC23:32
*** Prince781 has joined #buildstream23:48
« Monday, 2018-10-22 Index Wednesday, 2018-10-24 »

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!