Wednesday, 2022-01-26

gmann:)00:05
spotzYeah I shufflled a few to the top and will order those that I like00:16
*** lajoskatona_ is now known as lajoskatona00:31
*** odyssey4me is now known as Guest65212:40
jungleboyjgmann: I got it.  Thank you!14:17
jungleboyjFor those that want to give input on the Z release name:  https://twitter.com/jungleboyj/status/148634491069344154114:27
*** dansmith is now known as Guest66114:38
dansmith_gmann: so, this is working: https://review.opendev.org/c/openstack/grenade/+/826101/516:09
dansmith_gmann: I assume at some point we need to reconvene a meeting of people interested in a tick-tock release plan to settle on .. a plan16:09
*** dansmith_ is now known as dansmith16:09
gmanndansmith: nice, thanks for working on that. yeah, how about scheduling it next week or 2nd week of feb?17:12
dansmithgmann: sure, sounds good17:35
dansmithmaybe we can ask for interested parties in the meeting tomorrow17:35
gmannsure. 17:43
*** dansmith is now known as Guest68819:01
ade_lee__Guest688, gmann - hey - would be great to get review/+1 on https://review.opendev.org/c/openstack/governance/+/81658719:40
ade_lee__I think we're almost there?19:40
*** Guest688 is now known as dansmith19:40
*** dansmith is now known as dansmith_19:41
*** dansmith_ is now known as dansmith19:41
dansmithade_lee__: ack, added to tab queue19:42
ade_lee__mnaser, ricolin ^^19:42
ade_lee__dansmith, thanks19:42
dansmithso does anyone know how to get gerrit to stop showing old comments?19:44
dansmithassume it's because of the upgrade yesterday, but I've just been ignoring them all day19:45
fungican you elaborate on what you mean by "old comments"?19:45
fungiwhat is it showing that it didn't previously?19:45
dansmithseems like it now shows comments from old patchsets, either where they go in the file, or stacked at the top under "original comment position not found in this patchset"19:45
fungihave an example? i rarely use the webui so an not sure quite what you're describing19:46
dansmithhttps://imgur.com/a/GrVD9F119:47
dansmithevery comment that no longer applies is stacked at the top, in a non-sensical manner as you can see19:47
dansmith"from patcheset 2"19:47
dansmiththe ones that still fit, are shown inline, even though this is patchset 5 and they're no longer valid19:47
dansmitheverything I've looked at today is like this19:47
fungiinteresting, so it's trying to thread together comments from earlier patchsets19:47
fungiclarkb: ^ do you recall seeing that called out in the release notes?19:48
fungii'm checking to see if i can find a preference toggle for it19:48
clarkb3.3 did that too and I think 3.219:49
clarkbI don't think that is new to this ugprade but to polygerrit19:49
clarkbit doesn't know where to place them so it does its "best"19:49
dansmithI've never seen this before today, fwiw19:49
clarkbI definitely saw it on 3.3 in the past19:49
dansmithit's horribly distracting 19:49
clarkbfuzzier if it was on 3.2, but I'm fairly certain this isn't a new thing I think it came with polygerrit19:50
dansmithsurely there's some way to turn it off?19:50
clarkbI don't think so? I don't know what else gerrit could do?19:50
dansmithnot show me comments from six months ago that no longer apply?19:50
dansmithI'm confused19:50
clarkboh! I see19:51
dansmiththat screenshot is patchset 5, why do I need to see typo comments from patchset 2?19:51
clarkbI thought you meant render in some other location19:51
clarkbthe reaosn for that is because they are all unresolved19:51
dansmithso we have to resolve every comment manually now?19:51
clarkbIf someone resolves those comments it shouldn't try to show them to you anymore iirc19:51
dansmiththat's a whole lotta clicking19:52
clarkbwell not each comment, I think it does them by thread. But ya19:52
fungiis that related to the "attention set" stuff from 3.3?19:52
clarkbfungi: it might be.19:52
dansmiththat's madness19:52
clarkbactually ya I think that is the underlying feature19:52
fungilike if you're not in the attention set for a comment you don't need to resolve it?19:52
clarkbbecause it needs to keep track of who's atetntion is required19:52
dansmithI don't even reply to comments that I'm just going to fix, like typos especially19:52
fungidansmith: if it helps, i'm the same way19:53
fungii rarely reply to comments if i'm just going to do whatever they ask19:53
clarkbwell gertty gets this "wrong" by resolving all of its comments and dangling the comments it responds to19:53
clarkbBut ya I think the idea is to feed info to attention sets so that people know who needs to go next on a change19:53
dansmithfungi: I mean seriously, that's like a ton of clicking on each PS revision now, and either (a) you have to chase people to resolve their comments so you can read the thing, or someone else marks them resolved who didn't actually resolve them19:54
dansmitheither way, we'll wear out our mouse buttons19:54
fungiit certainly does seem inconvenient19:55
clarkbhttps://gerrit-review.googlesource.com/Documentation/user-attention-set.html are the docs on it19:56
fungiany ideas on how we can make it less of a treadmill?19:56
clarkbit mentions unresolved comments for merged and abandoned changes19:56
dansmith"but you can disable it by setting enableAttentionSet to false."19:56
clarkbI mean I haven't had any problems with it... I dunno what I'm doing differently19:56
dansmiththat disables all of the attention stuff I assume, but tbh that has been noise since it was introduced, IMHO19:57
dansmithI don't know anyone specifically using it19:57
clarkband I work with a lot of people that use gertty which gets it completely wrong :)19:57
clarkbdansmith: my biggest concern with globally disabling it is gerrit has a habit of introducing these features then force enabling them a few releases later19:57
fungiyeah, i expect gertty's commenting functionality needs a fresh coat of paint since the switch to patchset level comments and all that19:58
clarkbit is better for us to give them feedback through use and improve it. Than ignore it until we can't anymore then be stuck with something more broken19:58
dansmithyeah, I'm sure that's an eventuality, but still..19:58
dansmithI mean, I wish there was at least a user pref to "hide unresolved comments from a billion years ago"19:58
clarkbya or even if we made it so the default state of a comment was resolved or somet third state of "don't put this in the binary resolved toggling treadmill"19:59
clarkbI can write an email to their mailing list providing this input to see if they have any good ideas for improving it20:00
dansmithcool, seems like there should be a middle ground between "treadmill" and "global nuke of all attention set features"20:00
dansmithclarkb: thank you for doing that on our behalf20:01
clarkbbasically Most of our users don't use the unresolved resolved togging of comments through their back and forth interactions as they have many years of experience with older gerrit that didn't have this functionality and it leads to a lot of mouse clicking. One problem with this is that unresolved comments from ancient patchsets continue to show up on new patchsets even though20:02
clarkbthey are no longer revelant and the original comment position cannot be found. Is there some way we can make the default state of a comments "resolved" or some new state that isn't expected to be explicitly toggled later? Maybe the necessary changes can be made on a per user basis?20:02
clarkbdansmith: any concern with me sharing your imgur link?20:02
fungifwiw, i like that gertty automatically resolves any comments when a new patchset is pushed, though i agree the threading could use work20:04
clarkbfungi: I don't like it because it orphans the comments you are replying to20:04
clarkbfungi: I haven't been able to figure out how to resolve a comment if it has been replied to by gertty20:05
clarkbBut its not the end of the world I basically just accept the comment threads will have two colors mixed throughout20:05
fungiyeah, ideally pushing a new patchset would automatically resolve any unresolved comments from previous revisions20:06
fungithat was effectively the behavior before 3.3, i think20:06
fungisince comments from earlier patchsets weren't displayed20:07
dansmithclarkb: no, but let me get you another20:15
dansmithbecause the things at the top are annoying, but the real annoying ones are the 6mo old ones in the middle20:16
clarkbgot it20:16
dansmithhttps://imgur.com/a/KtQOnYu20:17
dansmithpatchset 16, showing unresolved comments from patchset 5, from last summer20:17
fungithat really does help make the case ;)20:23
dansmith...yeah...20:31
clarkbfungi: dansmith: https://groups.google.com/g/repo-discuss/c/evFgN3tTwxE/m/t9B5UOcSAQAJ hopefully I captured that properly22:31
dansmithclarkb: sure, looks great, thanks for doing that22:42

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!