IRC logs for #buildstream for Monday, 2018-11-26

*** dchapman has joined #buildstream01:00
*** nimish has joined #buildstream02:35
*** alatiera has quit IRC02:48
*** abderrahim has joined #buildstream03:15
*** abderrah1 has quit IRC03:16
*** tristan has joined #buildstream04:33
*** nimish has quit IRC05:01
*** nimish has joined #buildstream05:42
*** worldofpeace has joined #buildstream06:17
*** nimish has quit IRC06:21
*** tiagogomes_ has joined #buildstream07:39
*** tiagogomes has quit IRC07:40
*** toscalix has joined #buildstream08:32
*** ChanServ sets mode: +o tristan08:45
*** tristan changes topic to "BuildStream 1.2.3 is out ! | https://gitlab.com/BuildStream/buildstream | Docs: https://docs.buildstream.build/ | IRC logs: https://irclogs.baserock.org/buildstream | Mailing List: https://mail.gnome.org/mailman/listinfo/buildstream-list | Roadmap: https://wiki.gnome.org/Projects/BuildStream/Roadmaps"08:53
gitlab-br-bottoscalix closed issue #695 (Update workspace CLI including allowing default locations) on buildstream https://gitlab.com/BuildStream/buildstream/issues/69509:13
*** abderrahim has quit IRC09:17
*** pro[m] has quit IRC09:17
*** abderrahim[m] has quit IRC09:17
*** ssssam[m] has quit IRC09:17
*** alatiera_afk has quit IRC09:17
*** kailueke[m] has quit IRC09:17
*** theawless[m] has quit IRC09:17
*** benbrown has quit IRC09:17
*** gitlab-br-bot has quit IRC09:17
*** lchlan has quit IRC09:17
*** paulsherwood has quit IRC09:17
*** jmac has quit IRC09:17
*** dineshdb[m] has quit IRC09:17
*** waltervargas[m] has quit IRC09:17
*** doras[m] has quit IRC09:17
*** inigomartinez has quit IRC09:17
*** connorshea[m] has quit IRC09:17
*** oknf[m] has quit IRC09:17
*** segfault3[m] has quit IRC09:17
*** jjardon[m] has quit IRC09:17
*** asingh_[m] has quit IRC09:17
*** tlater[m] has quit IRC09:17
*** krichter[m] has quit IRC09:17
*** cgmcintyre[m] has quit IRC09:17
*** rafaelff[m] has quit IRC09:17
*** mattiasb has quit IRC09:17
*** albfan[m] has quit IRC09:17
*** m_22[m] has quit IRC09:17
*** juergbi has quit IRC09:17
*** jjardon has quit IRC09:17
*** hergertme has quit IRC09:17
*** ironfoot has quit IRC09:17
*** tintou has quit IRC09:17
*** jjardon_ has joined #buildstream09:17
*** lchlan has joined #buildstream09:17
*** tintou has joined #buildstream09:18
*** ironfoot has joined #buildstream09:19
*** benbrown has joined #buildstream09:19
*** abderrahim has joined #buildstream09:20
*** jmac has joined #buildstream09:20
*** juergbi has joined #buildstream09:24
*** paulsherwood has joined #buildstream09:27
*** gitlab-br-bot has joined #buildstream09:31
*** tintou has quit IRC09:41
*** tintou has joined #buildstream09:42
*** inigomartinez has joined #buildstream09:52
*** asingh_[m] has joined #buildstream09:54
*** alatiera_afk has joined #buildstream10:06
*** raoul has joined #buildstream10:11
*** mattiasb has joined #buildstream10:12
*** jonathanmaw has joined #buildstream10:22
*** lachlan has joined #buildstream10:30
*** lachlan has quit IRC10:30
*** lachlan has joined #buildstream10:30
*** albfan[m] has joined #buildstream10:45
valentindjjardon_, I would like to understand how you can reproduce #645 after the fix.10:47
*** jjardon_ is now known as jjardon10:47
valentindThere is nothing else in Freedesktop SDK strip-command that I see could help.10:47
valentindUnless if you have symbolic links.10:48
*** jjardon has quit IRC10:48
valentindWell, it would not help either. But I can see how it could fail.10:48
*** jjardon has joined #buildstream10:48
*** ChanServ sets mode: +o jjardon10:48
jjardonvalentind: sure, I will try to prepare a minimal project where I can see the issue10:48
valentindjjardon, ok thank you.10:48
jjardonMondays are a bit busy so probably it will have to wait until tomorrow10:49
valentindUnfortunately, it depends on file ordering.10:49
valentindSo it is hard to reproduce.10:49
jjardonvalentind: yeah, as a reference, this is a project that uses fsdk as a junction, then put some things on top; I will try to create the minimum amount of changes to see the bug10:50
jjardonvalentind: BTW, in the issue I was suggesting to projects using buildstream to use the freedesktop-sdk rules, no to use them in buildstream itself. Only in case someone else hit the issue so at least they can workaround for now10:51
valentindjjardon, we can do that.10:51
*** BradR has joined #buildstream10:56
*** segfault3[m] has joined #buildstream10:56
*** lachlan has quit IRC10:57
*** lachlan has joined #buildstream11:02
*** krichter[m] has joined #buildstream11:13
*** connorshea[m] has joined #buildstream11:14
*** lachlan has quit IRC11:15
*** tlater[m] has joined #buildstream11:18
*** lachlan has joined #buildstream11:22
*** oknf[m] has joined #buildstream11:24
*** alatiera has joined #buildstream11:24
*** abderrahim[m] has joined #buildstream11:35
*** doras[m] has joined #buildstream11:37
*** waltervargas[m] has joined #buildstream11:43
*** lachlan has quit IRC11:51
*** cgmcintyre[m] has joined #buildstream11:52
*** m_22[m] has joined #buildstream11:52
jmacjuergbi: when you get a chance, could you have another look at !946 and see if my repsonse to your review is satisfactory?12:03
juergbisure12:03
*** lachlan has joined #buildstream12:05
*** lachlan has quit IRC12:10
*** jjardon[m] has joined #buildstream12:16
*** lachlan has joined #buildstream12:21
*** rafaelff[m] has joined #buildstream12:24
*** dineshdb[m] has joined #buildstream12:29
*** kailueke[m] has joined #buildstream12:31
*** pro[m] has joined #buildstream12:35
*** ssssam[m] has joined #buildstream12:41
jmacI'm getting the "fatal: reference is not a tree: 0d4d709a4b49c041654d447ebc838d63c70d074612:45
jmac... error again12:45
jmachttps://gitlab.com/BuildStream/buildstream/-/jobs/12547087212:45
jmacAnyone else seen that? It feels familiar12:45
*** lachlan has quit IRC12:47
*** theawless[m] has joined #buildstream12:50
*** dbuch has joined #buildstream12:51
juergbijmac: pipeline from before last force push?12:52
juergbii.e., commit no longer fetchable by git?12:52
*** pepesza has joined #buildstream12:54
*** lachlan has joined #buildstream12:54
toscalixadded the free tools and editors devroom CfP deadline to the buildstream calendar12:58
*** lachlan has quit IRC12:58
toscalixalso the distro one12:59
toscalixthe lightning talks was already there13:00
tristanvalentind, Commented on the workspace/build tree mind boggler as requested: https://gitlab.com/BuildStream/buildstream/issues/311#note_12006651213:05
tristannot an easy discussion :)13:05
*** lachlan has joined #buildstream13:05
*** lachlan has quit IRC13:10
valentindtristan, thank you.13:14
Nexusjuergbi: What are your thoughts on refactoring pugin.warn() to make it accessable to loader.py?13:31
jmacjuergbi: Yes, you're right, I'm looking at an old ref13:37
jmacThanks!13:37
*** nimish has joined #buildstream14:02
tristanNexus, A Plugin instance cannot exist while the loader is loading, but the Context object is available way in advance and all messages are proxied through it. A sane approach would be to have the Context object support the warn functionality, and have both Plugin.warn() and the Loader call through there14:05
skullmantristan: whether a warning is fatal depends on the project though, and the Context looks to only be concerned with the top-level project14:06
tristanProbably it needs to be told what project is the source of the warning too, since the project is what declares which warnings are fatal (and both the Plugin and the Loader will have different ways of determining the associated project)14:06
tristanskullman, jinx :)14:06
skullmantristan: given each Project has a loader, it sounds like maybe it would be better to pass the policy of which warnings are fatal down, rather than having the Loader have a reference back to the project.14:07
jjardonHi, can I have reviews of https://gitlab.com/BuildStream/buildstream/merge_requests/972 , please?14:08
tristanskullman, Hmmm interesting, currently there is already a two way reference14:10
tristanI.e. every loader has a reference to it's project14:10
skullmanhuh, I failed to spot that14:10
tristanskullman, it's definitely a bit of a mish mash around there14:10
tristanwhich is interestingly necessary because of cross project includes and multi stage loading, which is... a bit of a mind bender14:11
tristananyway... that is certainly an area which could do with a nice monumental refactor, but the state of things don't look *that* horrible as it stands14:13
skullmanYeah, I'm currently looking at what it would take to add the minimal amount of metadata to artifacts to be able to load elements out of the artifact cache. I've not been able to get my head around it, so might have to suggest duplicating code-paths for staging.14:13
tristanAh interesting... not sure what you mean by "... for staging", hopefully it only means duplicating code paths for loading, and Element.stage_*() operations can be run normally from the same data model ?14:15
tristanWould not be nice if a data model loaded from artifact metadata differs from a data model loaded from the source project files14:15
*** lachlan has joined #buildstream14:18
tristanjjardon, I was unable to figure out how to lift the +r thing on this channel, which tends to annoy people14:19
skullmanBy "for staging" I meant that's the only operation I need them to do, but I'm not so confident that it only needs changes to it being loaded, since the last time I looked at whether I could make something quack like an Element for restricted circumstances, I'd have needed to reimplement most of it.14:19
jjardontristan: It was flooded with SPAM when I removed it14:19
jjardonliterally the same day14:19
tristanjjardon, Think we could do the following... (A) Lift the join requirement for entering the channel for now... (B) Give more people op status so many people can do it when needed ?14:20
tristanjjardon, Oh14:20
tristanDidnt realize that14:20
jjardontristan: anyone can enter the channel14:20
tristanI see that #gnome-hackers was +r today too, seems the spam sometimes hits14:20
tristanOh right14:20
tristanit's voice14:20
*** lachlan has quit IRC14:20
jjardonyeah, only registered nicks can talk14:20
*** lachlan has joined #buildstream14:20
tristanstill... it would be nice if we had a situation where we were less strict but had many people who can +r/-r as needed14:21
jjardonI do not think ask for a nick registration is too much?14:21
tristananyway, I don't mind so much, except that I have to act all "op status" whenever I want to say anything haha14:21
jjardonBut sure feel free to remove the channel mode14:21
tristanI mean, I would not remove the channel mode before giving op status to like... all of the full time contributors (who are already a lot)14:22
tristanthat way anyone could do it whenever we notice spam14:22
tristanjjardon, I didnt figure out how to do it myself though... I msged chanserv for help and still couldnt figure it out hehe14:23
*** lachlan has quit IRC14:23
jjardontristan: /mode #buildstream -M14:25
tristanjjardon, Ah, when I'm signed in an opped, I got it thanks !14:27
tristananyway not really a huge deal... just a thought before closing time... which is in 3... 2... 1...14:28
tristanSee ya :)14:28
jjardontristan: np :) o/14:28
*** jjardon sets mode: -M 14:28
tristanright, lets give op status to all the regulars here soon in case spam does hit :)14:29
*** jjardon sets mode: +o juergbi14:29
*** jjardon sets mode: +o valentind14:29
*** lachlan has joined #buildstream14:29
tristanjjardon, Is that enough ?14:29
tristanjjardon, I mean.. don't we have to tell ChanServ or smth ?14:30
jjardonif they want to be op automatically when login in, yes14:30
paulsherwooddidn't we already establish that we couldn't defend against the spam without +r? not sure what this new step is hoping to achieve?14:31
*** raoul_ has joined #buildstream14:31
tristanjjardon, Thats what I want... I mean... a situation where anyone we know is a regular has the ability to +M if spam does hit and you and I are not around14:31
tristanpaulsherwood, Well, yeah basically, there is that or the mode which doesnt let you in the channel14:32
tristanpaulsherwood, it's just sometimes an annoyance to people, and seems to only be periodically necessary from time to time14:32
*** raoul has quit IRC14:32
*** lachlan has quit IRC14:32
tristanI dont really feel super strongly about it anyway14:32
paulsherwoodi felt strongly, but when we reopened last time the spammers were back within hours14:33
paulsherwoodthis is why we can't have nice things, etc14:33
tristanjust thought we could lift it when spam is not bothering us, and have it so anyone had the ability to enable/disable14:33
tristanLike, if you had op status automatically, you could always sign in and +r the place if there was spam14:34
tristanand anyone who is here on a day to day basis could too14:34
paulsherwoodthat's just more work, and more spam in the logs14:34
tristan(not like its a big deal to let everyone we know have /op powers)14:34
tristanYeah... if it happens a lot, I thought it happened a couple times a month with a lot of flooding14:35
jjardonpaulsherwood: the channel was always being open14:35
jjardonwe protect with +M, which means only people with registered nicks can talk14:35
paulsherwoodack14:36
jjardonI think is not as bad as +r where you can not enter to the channel at all14:36
jjardonbut yeah still can be annoying to someone14:36
*** phildawson82 is now known as phildawson14:36
tristanOh yeah, I keep mixing up +M and +r sorry14:36
* paulsherwood too14:36
jjardonI think +M is a good compromise14:37
paulsherwood+114:37
tristan:)14:37
paulsherwoodso no need for ops to everyone14:37
tristanSure :)14:37
jjardonok, so should I put +M back? :)14:39
tristanjjardon, Yeah go ahead :)14:39
*** jjardon sets mode: +M 14:39
tristanheh14:39
* jjardon is actually surprised is not possible to set +M in freenode14:39
alatierawhat does +M do?14:41
jjardonalatiera: 14:35 <@jjardon> we protect with +M, which means only people with registered nicks can talk14:43
*** tristan has quit IRC14:44
alatierajjardon, thanks!14:45
*** lachlan has joined #buildstream14:46
*** nimish has quit IRC14:47
*** nimish has joined #buildstream14:47
*** nimish has quit IRC15:02
*** nimish has joined #buildstream15:02
*** raoul_ has quit IRC15:08
*** raoul_ has joined #buildstream15:09
*** lachlan_ has joined #buildstream15:10
*** tristan has joined #buildstream15:15
*** lachlan_ has quit IRC15:16
*** lachlan_ has joined #buildstream15:32
*** nimish has quit IRC15:32
*** nimish has joined #buildstream15:33
*** finn has joined #buildstream15:40
*** mohan43u has quit IRC15:53
*** nimish has quit IRC16:08
*** nimish has joined #buildstream16:08
*** mohan43u has joined #buildstream16:28
*** solid_black has quit IRC16:31
*** lachlan_ has quit IRC16:36
*** lachlan has quit IRC16:36
*** lachlan has joined #buildstream16:53
*** lachlan_ has joined #buildstream16:53
*** toscalix has quit IRC16:54
*** lachlan_ has quit IRC17:05
*** lachlan_ has joined #buildstream17:06
*** lachlan_ has quit IRC17:11
*** lachlan has quit IRC17:11
*** nimish has quit IRC17:33
*** nimish has joined #buildstream17:34
*** lachlan has joined #buildstream17:54
*** lachlan_ has joined #buildstream17:54
*** finn has quit IRC17:56
*** lachlan_ has quit IRC17:58
*** lachlan has quit IRC17:58
*** ctgriffiths has quit IRC17:58
*** ctgriffiths has joined #buildstream17:59
*** lachlan_ has joined #buildstream18:09
*** lachlan has joined #buildstream18:09
*** nimish has quit IRC18:14
*** nimish has joined #buildstream18:14
*** nimish has quit IRC18:32
*** finn has joined #buildstream18:59
*** lachlan_ has quit IRC19:04
*** lachlan has quit IRC19:05
*** xjuan has joined #buildstream19:32
*** lachlan has joined #buildstream19:39
*** finn has quit IRC19:49
*** Shrooms has joined #buildstream19:54
*** tristan has quit IRC20:32
*** lachlan has quit IRC20:46
*** jonathanmaw has quit IRC20:53
*** xjuan has quit IRC20:53
*** toscalix has joined #buildstream21:20
*** toscalix has quit IRC21:21
*** alatiera has quit IRC21:36
*** ARMEWT has joined #buildstream22:00
*** cs-shadow has joined #buildstream22:17
*** ARMEWT has quit IRC22:29

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