| *** ianychoi has joined #openstack-tc | 00:19 | |
| *** adriant has quit IRC | 01:35 | |
| *** adriant has joined #openstack-tc | 01:55 | |
| *** diablo_rojo has quit IRC | 02:12 | |
| *** zhurong has joined #openstack-tc | 02:58 | |
| *** zaneb has quit IRC | 03:22 | |
| *** mugsie has quit IRC | 03:22 | |
| *** EmilienM has quit IRC | 03:22 | |
| *** mgagne has quit IRC | 03:22 | |
| *** persia has quit IRC | 03:22 | |
| *** zaneb has joined #openstack-tc | 03:38 | |
| *** mugsie has joined #openstack-tc | 03:38 | |
| *** EmilienM has joined #openstack-tc | 03:38 | |
| *** mgagne has joined #openstack-tc | 03:38 | |
| *** persia has joined #openstack-tc | 03:38 | |
| *** irclogbot_3 has quit IRC | 03:41 | |
| *** irclogbot_1 has joined #openstack-tc | 03:42 | |
| *** lbragstad has quit IRC | 04:43 | |
| *** whoami-rajat has joined #openstack-tc | 04:47 | |
| *** e0ne has joined #openstack-tc | 05:21 | |
| *** iurygregory has joined #openstack-tc | 06:58 | |
| *** tosky has joined #openstack-tc | 07:15 | |
| *** e0ne has quit IRC | 07:35 | |
| *** dtantsur|afk is now known as dtantsur | 07:44 | |
| *** jpich has joined #openstack-tc | 07:57 | |
| *** e0ne has joined #openstack-tc | 09:08 | |
| *** Luzi has joined #openstack-tc | 09:29 | |
| asettle | o/ | 09:44 |
|---|---|---|
| *** tdasilva has quit IRC | 10:04 | |
| *** jpich has quit IRC | 10:50 | |
| *** jpich has joined #openstack-tc | 10:51 | |
| *** jaosorior has joined #openstack-tc | 11:20 | |
| *** EmilienM is now known as EvilienM | 11:45 | |
| *** Luzi has quit IRC | 11:49 | |
| *** jaosorior has quit IRC | 12:09 | |
| *** ijolliffe has joined #openstack-tc | 12:34 | |
| *** tdasilva has joined #openstack-tc | 12:46 | |
| dhellmann | o/ | 12:47 |
| *** lbragstad has joined #openstack-tc | 12:59 | |
| mnaser | bonjour | 13:24 |
| mnaser | hogepodge: i know you have a lot on your plate but do you (or perhaps know someone in the community) who can help land a golang/go change similar to this? https://github.com/golang/go/commit/2ba8fc5b086942dbb23282702f61c813298867f3 | 13:25 |
| mnaser | but for opendev? | 13:25 |
| fungi | out of curiosity, what projects do we still have in go? | 13:29 |
| fungi | if none of them are proceeding then maybe it's simpler to drop that from the pti and revisit later if it's warranted | 13:30 |
| *** tdasilva has quit IRC | 13:32 | |
| evrardjp | fungi: isn't swift developing a feature using go? | 13:43 |
| evrardjp | (I don't mean hummingbird) | 13:43 |
| jroll | the patch to vcs.go isn't a blocker (details in a comment I just left there) | 13:47 |
| *** whoami-rajat has quit IRC | 13:57 | |
| fungi | evrardjp: i thought last i heard that work had stalled out a couple years back | 13:58 |
| fungi | indeed, https://opendev.org/openstack/swift/commits/branch/feature/hummingbird indicates the last change to merge was for that was in 2016 (assuming that's where the work was taking place) | 13:59 |
| *** tdasilva has joined #openstack-tc | 14:06 | |
| *** e0ne has quit IRC | 14:13 | |
| *** tdasilva has quit IRC | 14:22 | |
| *** tdasilva has joined #openstack-tc | 14:22 | |
| *** stephenfin is now known as finucannot | 14:26 | |
| evrardjp | fungi: I said I don't mean hummingbird :p | 14:30 |
| evrardjp | in other words: something else | 14:30 |
| evrardjp | you can blame my english | 14:30 |
| evrardjp | (and the fact I don't know the feature name... lots of small files?) | 14:30 |
| smcginnis | I'm not aware of any current work being done in the OpenStack community with Go, but I would kind of like to see that change made to enable it if there is anyone that wants to go that route. | 14:33 |
| smcginnis | The amount of overhead/bureaucracy (not unfounded) when it came up last time kind of ended up killing it and leaving a really bad taste in a lot of folks mouths, so it would be nice to avoid that. | 14:34 |
| fungi | i think we're at the point now where we could fairly easily readd go to the pti if there are projects wanting to use it (by mostly just reverting its removal). my bigger concern is whether we're spending a lot of effort caring for process documentation about a language nobody in our community is using | 14:41 |
| fungi | if there's at least one project using (or wanting to use) it, then i think we're justified in keeping it and updating it | 14:42 |
| *** dklyle has quit IRC | 14:42 | |
| *** dklyle has joined #openstack-tc | 14:42 | |
| fungi | but in the interest of spending our time where it does the most good, i've not seen evidence that this is important for us to maintain at the moment | 14:43 |
| smcginnis | That's fair. There is some maintenance overhead for us with no real consumer of that work at this point. And I think that's true that now it would be a lot easier to bring it back if needed, versus the work that needed to be done initially. | 14:43 |
| fungi | anyway, it's worth a call out to the community to see if there's point to us keeping it up (and the opendev.org entry for go resource urls is orthogonal to that in my opinion. it's something other projects in opendev would likely use outside openstack, but also the existing consumers are supported by the redirect for now) | 14:45 |
| smcginnis | I'll take a look at adding opendev support in go. It doesn't look like too much work and I've some time today I can tinker. | 14:45 |
| fungi | yeah, i think that's useful regardless | 14:45 |
| smcginnis | Especially in support of non-OpenStack users of OpenDev. | 14:45 |
| fungi | more just questioning whether maintaining the pti section for go makes sense if it's never exercised | 14:46 |
| *** lpetrut has joined #openstack-tc | 14:46 | |
| evrardjp | smcginnis: "to go that route" -> I see what you did there. | 14:50 |
| smcginnis | Hah! Not intentional, but tempted to take credit for that. :P | 14:51 |
| evrardjp | fungi: my point was, I think it is used | 14:51 |
| smcginnis | Sounds like maybe we need something on the ML asking for community feedback. If there is someone, great. If not, then I think fungi is right that we can drop it until it is needed. | 14:52 |
| evrardjp | timburke: is there current work done that uses go in the swift community? Was that the lots of small files, done by OVH? | 14:52 |
| timburke | there is a feature branch that's looking to use some go code, yes. and yes, it's the lots-of-small-files work that OVH is spearheading | 14:53 |
| timburke | what's the context? | 14:53 |
| * timburke starts to read scrollback | 14:53 | |
| evrardjp | was wondering if someone in the community used go, you just confirmed swift does | 14:53 |
| evrardjp | I am happy with that answer | 14:53 |
| hogepodge | mnaser: I can ask around. Should that patch also include opendev? | 14:54 |
| hogepodge | oh wait nevermind that's what you were asking :-P | 14:55 |
| fungi | timburke: cool, thanks for confirming! | 14:56 |
| smcginnis | hogepodge: I am looking at it now. I think I have it almost done. Just running tests and making sure I'm set up right to submit. | 14:56 |
| fungi | in that case i agree it's worth our time to keep the go pti document | 14:56 |
| *** e0ne has joined #openstack-tc | 14:57 | |
| openstackgerrit | Iury Gregory Melo Ferreira proposed openstack/governance master: Add ironic-prometheus-exporter https://review.opendev.org/665414 | 14:57 |
| mnaser | hogepodge, smcginnis: thank you both :) | 14:57 |
| timburke | glad i could help. fwiw, dev work is happening at https://review.opendev.org/#/q/project:openstack/swift+branch:feature/losf+(-age:1month+OR+is:open) -- just landed a merge from master overnight | 14:57 |
| fungi | yep, just found it at https://opendev.org/openstack/swift/commits/branch/feature/losf | 14:58 |
| fungi | my brain wants to see losf as lsof (for obvious reasons) | 14:58 |
| timburke | thanks for putting up with the maintenance of that section :-) | 14:58 |
| fungi | timburke: no worries, just want to be sure we're not spending time on things which nobody uses | 14:59 |
| timburke | yeah, mine too. or lofs | 14:59 |
| *** e0ne has quit IRC | 15:00 | |
| zaneb | I'd be surprised if someone didn't end up wanting to write k8s operators for openstack at some point too | 15:00 |
| mnaser | tc-members: horace from the OSF has spent sometime talking to the CTO of AWCloud wrt to the 100% votes and shared the link regarding reviews as well so they'll be taking the time to explain that and hopefully we get productive and useful reviews from them (cc aspiers) | 15:01 |
| mnaser | so i'm not going to follow up on this any further than this, but if there's any issues that surface again, i'd be happy to circle back | 15:01 |
| asettle | Cool :) hopefully that's helpful! | 15:01 |
| aspiers | s/aspiers/asettle/ or am I missing something here? | 15:02 |
| asettle | This is how mnaser tells me he actually wants you as vice-chair :p | 15:02 |
| mnaser | aspiers: i think it was you who broguht up the tack on +1s ? | 15:02 |
| aspiers | Ohhh, that topic | 15:02 |
| asettle | Ohhh, yep not me | 15:03 |
| aspiers | Sorry, already got swapped out of my RAM :) | 15:03 |
| asettle | mate | 15:03 |
| aspiers | mnaser: Awesome, thanks a lot! | 15:03 |
| mnaser | just wanted to make sure you got the update on that :) | 15:03 |
| aspiers | Really appreciate it | 15:03 |
| aspiers | :) | 15:03 |
| *** e0ne has joined #openstack-tc | 15:15 | |
| *** e0ne has quit IRC | 15:21 | |
| smcginnis | mnaser, hogepodge: https://go-review.googlesource.com/c/go/+/182418 | 15:21 |
| jroll | nice, thanks for that smcginnis | 15:22 |
| jroll | wow, that's a very different gerrit | 15:22 |
| mnaser | gerrit 3.0.0 :> | 15:22 |
| mnaser | it's perdy | 15:22 |
| mnaser | but thanks so much smcginnis | 15:22 |
| smcginnis | jroll: Yeah, the latest and unstyled looks so different. | 15:22 |
| smcginnis | No problem | 15:22 |
| smcginnis | Bah, I think I see something I missed already. :D | 15:23 |
| jroll | smcginnis: since I can't be bothered to get an account to comment, why not add additional tests instead of changing the git.o.o tests? | 15:23 |
| *** e0ne has joined #openstack-tc | 15:24 | |
| hogepodge | smcginnis: I don't have an account to comment on the patch, but I would suggest we say "the new expanded OpenStack Foundation infra service" to help clarify that the scope of opendev is growing beyond just the OpenStack project itself. | 15:25 |
| smcginnis | jroll: I did. Lines 104-111 are the existing ones, then I added 112-118. The ones below are not really specific, but thought I would update them to use the newer URLs anyway. | 15:25 |
| fungi | https://opendev.org/ has plenty of prose to steal if needed | 15:25 |
| smcginnis | hogepodge: OK, I'll update that. | 15:26 |
| *** tdasilva has quit IRC | 15:26 | |
| smcginnis | And add a link for reference to opendev.org. | 15:26 |
| *** e0ne has quit IRC | 15:27 | |
| jroll | smcginnis: yeah, you left one test, but now the "with .git" and "subdirectory" tests are removed for git.o.o. (and "with .git" are removed altogether) | 15:27 |
| smcginnis | jroll: I didn't think keeping git.o.o was of any use since they are now just redirects anyway. | 15:28 |
| jroll | smcginnis: eh, it tests that referencing it still does the right thing | 15:28 |
| jroll | I'm fine either way, just a suggestion | 15:29 |
| fungi | i *think* (but am not certain) that if git.openstack.org doesn't continue to be supported in go's vcs list, then existing references using git.openstack.org will break | 15:29 |
| *** e0ne has joined #openstack-tc | 15:29 | |
| fungi | so presumably continuing to test it at least as thoroughly as it was tested in the past is still warranted | 15:30 |
| jroll | existing references which use the shortcut will break, yes (e.g. git.o.o/openstack/swift will break, https://git.o.o/openstack/swift.git will not) | 15:30 |
| jroll | but I agree, continuing to test is a good thing | 15:31 |
| clarkb | why not just use a proper url? | 15:32 |
| smcginnis | OK, just pushed an update. | 15:32 |
| fungi | because that's not what go used to recommend | 15:32 |
| fungi | clarkb: ^ | 15:32 |
| fungi | so those references are out there in the wild already | 15:32 |
| clarkb | ya and thatd why docker cant ipv6 might be good to push back on such things more :) | 15:33 |
| clarkb | fungi: yes but there arent opendev shortcuts yet | 15:33 |
| *** e0ne has quit IRC | 15:33 | |
| clarkb | can put git.o.o out to pasture and use urls for opendev | 15:33 |
| fungi | that's an excellent point | 15:33 |
| smcginnis | clarkb: I'm not sure I'm following what you mean by "proper url" | 15:34 |
| fungi | i thought you were commenting on the removal of the .git suffix testing for the old non-url legacy references | 15:34 |
| clarkb | https://opendev.org/openstack/swift | 15:34 |
| fungi | but you're suggesting 182418 isn't really needed at all | 15:34 |
| clarkb | yes | 15:34 |
| fungi | because those incomplete domain-only references are deprecated anyway | 15:34 |
| smcginnis | If pasting that URL was supposed to illustrate better, it didn't. :) | 15:34 |
| clarkb | smcginnis: dont use the shortcut that requires your code change for opendev.org. Instead people should use proper urls like the one above | 15:35 |
| fungi | smcginnis: the go community now recommends dependencies be specified via a full url, so that map of domains to vcs urls is not needed for new domains | 15:35 |
| fungi | it might be relevant to update the git.openstack.org entry in their map to go to https://opendev.org though | 15:36 |
| smcginnis | So tacking on https:// instead of starting with opendev? I'm not seeing that happen much in practice, even if that is the recomendation. | 15:36 |
| clarkb | smcginnis: well we can sort of force the issue by not having ashortcut | 15:37 |
| jroll | I'm still not even sure the https:// bit is required | 15:37 |
| fungi | probably because 99.999% of those references all start with github.com so nobody has any incentive to use the new form | 15:37 |
| clarkb | and considering how broken that ecosystem is with urls Im of the opinion we should | 15:37 |
| clarkb | jroll ah it may assume https ? | 15:38 |
| jroll | yeah | 15:38 |
| jroll | looks like it'll try all possible schemes until it works | 15:38 |
| jroll | https://github.com/golang/go/blob/master/src/cmd/go/internal/get/vcs.go#L167 | 15:38 |
| fungi | in that case even the old git.openstack.org entry may be redundant? | 15:38 |
| jroll | (with some sort of "secure" setting) | 15:38 |
| fungi | if it weren't in the map it would still get turned into a viable irl | 15:38 |
| fungi | s/irl/url/ | 15:38 |
| mnaser | i feel like smcginnis ended up signing themselves up for more work :) | 15:39 |
| smcginnis | ;) | 15:39 |
| fungi | mnaser: or for no work | 15:39 |
| mnaser | i think if someone feels like fixing the go ecosystem and changing things, they can push up a patch, heh | 15:39 |
| jroll | fungi: still trying to figure out if it defaults to .git or if a .git suffix is required for anything not in the map | 15:39 |
| mnaser | if not then i think the patch is fine, we can wait for a review i guess :> | 15:39 |
| clarkb | its less work | 15:39 |
| fungi | jroll: .git suffixes work in both cases anyway | 15:39 |
| jroll | fungi: but because we have the map, not all usage in the wild may have the .git suffix | 15:40 |
| clarkb | you dont change anyrhing and people that want to include code from opendev just do so using a url possibly without the protocol prefix | 15:40 |
| smcginnis | I will just wait for the go maintainers to comment on whether this is needed or not. If they don't think so, and they agree it can just all be removed, I can update the patch to drop both forms. | 15:40 |
| jroll | fungi: that is to say, the git.o.o entry may be required | 15:40 |
| fungi | jroll: but the .git suffix is also optional in both cases | 15:40 |
| jroll | fungi: on the go side? it looks that way, but trying to validate in code | 15:40 |
| fungi | jroll: or are you saying that if references don't include .git then the legacy map is required so that the vcs type field can be inferred? | 15:41 |
| jroll | fungi: right, that's what I'm trying to verify | 15:41 |
| fungi | ahh | 15:41 |
| jroll | looking at an internal repository, it seems like it will figure out that it is a git repo (because our internal git server is obviously not in the mapping) | 15:41 |
| fungi | anyway, might be useful for someone to confirm that opendev.org references for go deps don't already "just work" before assuming we need an entry in a legacy vcs url mapping list | 15:42 |
| jroll | I believe it will try to detect the VCS system | 15:42 |
| clarkb | fungi: ++ | 15:42 |
| clarkb | it wasmy understanding that this should just work if people use a valid url | 15:43 |
| fungi | whether the existing git.openstack.org entry is redundant or not is a side concern | 15:43 |
| smcginnis | They have a nice "welcome new contributor" message. I wonder if we should expand on ours. | 15:43 |
| * jroll installs go locally while he eats lunch | 15:43 | |
| fungi | smcginnis: their welcome message looks an awful lot like what i remember ours saying too, but i won't argue that ours could probably still do with some updating. pretty sure the prose which is there is still the same as fifieldt drafted 5+ years ago | 15:44 |
| smcginnis | I'll admit, it's been a very long time since I paid any attention to ours, so maybe it is more similar than I remember. | 15:46 |
| *** jaypipes has quit IRC | 15:51 | |
| fungi | smcginnis: it's configured here... https://opendev.org/opendev/jeepyb/src/branch/master/jeepyb/cmd/welcome_message.py#L72-L96 | 15:55 |
| jroll | smcginnis: fungi: looks like clarkb was right, we probably don't need it: http://paste.openstack.org/show/752934/ | 15:56 |
| *** jamesmcarthur has joined #openstack-tc | 15:56 | |
| fungi | smcginnis: that could likely be improved on to make it configurable per project | 15:56 |
| smcginnis | jroll: Did it actually pull that down, or did it see "imported but not used" and skipped it. | 16:05 |
| smcginnis | fungi: That's an interesting idea. | 16:05 |
| fungi | it gets the commit info passed into the containing function so has the necessary context | 16:06 |
| fungi | and seems to already be plumbed for using multiple message files (though i think we need to alter the logic for selecting the message file, probably based on a lookup from the gerrit/projects.yaml file) | 16:10 |
| fungi | right now the alternative message implementation assumes you're passing --message-file from the gerrit hook script, so configured static site-wide | 16:11 |
| evrardjp | zaneb: really surprised? :p | 16:12 |
| fungi | smcginnis: and the logic for picking a non-default message for specific projects would be harder to implement in the hook script (which is shell), so better done in that python executable | 16:12 |
| zaneb | evrardjp: I mean, tripleo have discussed it and AIUI decided not to try yet. but I feel like it's inevitable that somebody will eventually | 16:23 |
| *** dtantsur is now known as dtantsur|afk | 16:26 | |
| *** jpich has quit IRC | 16:29 | |
| *** iurygregory has quit IRC | 16:53 | |
| jroll | smcginnis: the "go get" command pulled it down | 17:01 |
| jroll | as opposed to http://paste.openstack.org/show/752942/ | 17:01 |
| jroll | or an invalid package http://paste.openstack.org/show/752943/ | 17:02 |
| *** lbragstad is now known as lbragstad_runnin | 17:04 | |
| *** iurygregory has joined #openstack-tc | 17:15 | |
| *** iurygregory has quit IRC | 17:23 | |
| smcginnis | jroll: OK, cool. Maybe really not needed then. | 17:26 |
| *** lpetrut has quit IRC | 17:30 | |
| *** iurygregory has joined #openstack-tc | 17:36 | |
| *** ijolliffe has quit IRC | 17:37 | |
| *** iurygregory has quit IRC | 17:38 | |
| *** e0ne has joined #openstack-tc | 17:41 | |
| *** lbragstad_runnin is now known as lbragstad | 17:43 | |
| *** iurygregory has joined #openstack-tc | 17:44 | |
| *** e0ne has quit IRC | 17:46 | |
| fungi | so instead we can just update the reference in the pti and call it a day? | 18:04 |
| jroll | seems like it :) | 18:07 |
| fungi | and in all likelihood the existing git.openstack.org entry is also redundant given the behavior of modern go toolchains? | 18:14 |
| *** jamesmcarthur_ has joined #openstack-tc | 18:17 | |
| jroll | seems like they all are | 18:18 |
| jroll | I believe this map is just some shortcuts, so that go doesn't have to try to sniff the VCS to use or the protocol to connect with | 18:19 |
| *** jamesmcarthur has quit IRC | 18:21 | |
| *** jamesmcarthur_ has quit IRC | 18:24 | |
| *** e0ne has joined #openstack-tc | 18:42 | |
| *** Sundar has joined #openstack-tc | 18:56 | |
| *** dims has quit IRC | 18:57 | |
| *** jamesmcarthur has joined #openstack-tc | 19:02 | |
| *** dims has joined #openstack-tc | 19:05 | |
| *** iurygregory has quit IRC | 19:07 | |
| *** e0ne has quit IRC | 19:16 | |
| *** e0ne has joined #openstack-tc | 19:16 | |
| *** e0ne has quit IRC | 19:21 | |
| *** e0ne has joined #openstack-tc | 19:23 | |
| *** e0ne has quit IRC | 19:33 | |
| *** diablo_rojo has joined #openstack-tc | 19:33 | |
| *** jamesmcarthur has quit IRC | 19:41 | |
| *** jamesmcarthur has joined #openstack-tc | 19:42 | |
| *** jamesmcarthur has quit IRC | 19:44 | |
| *** jamesmcarthur has joined #openstack-tc | 19:44 | |
| *** jamesmcarthur has quit IRC | 20:00 | |
| *** jamesmcarthur has joined #openstack-tc | 20:04 | |
| *** jamesmcarthur has quit IRC | 20:05 | |
| *** Sundar has quit IRC | 21:05 | |
| *** EvilienM is now known as EmilienM | 22:03 | |
| *** diablo_rojo has quit IRC | 22:32 | |
| *** tosky has quit IRC | 22:48 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!