[Development] Meeting minutes from Qt Release Team meeting 27.02.2024

[Jani Heikkinen]

Qt 6.6 status:

  *   Qt 6.6.3 preparations started
     *   The target is to branch from ‘6.6’ to ‘6.6.3’ Mon 11th March
     *   The target is to release Qt 6.6.3 Fri 22rd March 2024
        *   Qt 6.6.3 will be the last release from Qt 6.6 series


Qt 6.7 status:

  *   Qt 6.7 API review still ongoing, see https://bugreports.qt.io/browse/QTBUG-119952
à Most probably branching from ‘6.7’ to ‘6.7.0’ will be delayed; we need to complete the review before branching

  *   Initial plan was to release Qt 6.7.0 RC Tue 5th March 2024
     *   Hoping we can freeze the RC content asap so that we could get RC out during next week
     *   Release blocker list here: https://bugreports.qt.io/issues/?filter=25675
        *   We will improve the documentation about how to create a ticket that ends up on the blocker list.
Next meeting Tue 12th March 2024 16:00 CET
br,
Jani Heikkinen
Release Manager
irc log below:
[17:00:35] <+jaheikki3> ablasche: akseli: carewolf: frkleint: mapaaso: The-Compiler: thiago: vohi: ping
[17:00:48] <Akseli> jaheikki3: pong
[17:01:11] <carewolf> pong
[17:01:56] <vohi> jaheikki3: pong
[17:02:30] <+jaheikki3> Time to start qt release team meeting. On agenda today:
[17:02:38] <+jaheikki3> Qt 6.6 status
[17:02:43] <+jaheikki3> Qt 6.7 status
[17:02:52] <+jaheikki3> Any additional item to the agenda?
[17:04:01] <vohi> Can we check if the definition of the list of blockers (as per JIRA query) makes sense?
[17:04:34] <+jaheikki3> vohi: sure, let's take that as a part of 6.7 status
[17:04:43] <vohi> 👍
[17:04:48] <+jaheikki3> But let's start from 6.6
[17:05:08] <+jaheikki3> Qt 6.6.3 preparations started.
[17:05:31] <+jaheikki3> Target is to branch from ‘6.6’ to ‘6.6.3’ at the beginning of next week
[17:05:45] <+jaheikki3> And the target is to release Qt 6.6.3 Fri 15th March 2024
[17:06:13] <+jaheikki3> And 6.6.3 will be the last release from Qt 6.6.3 series
[17:06:37] <+jaheikki3> That's actually all about 6.6 status at this time. Any comments or questions?
[17:07:02] <vohi> last week the target was to release 6.6.3 by end of March; what has changed that we need to speed that up by two weeks?
[17:07:50] <vohi> 6.6.2 is only two weeks old at this point
[17:08:35] <+jaheikki3> vohi: actually nothing has changed. Usually we have done the last release before the first new one and QT 6.7.0 is planned to happen 19th March
[17:09:53] <+jaheikki3> But we can also do the last Qt 6.6 release after Qt 6.7.0, I don't think it isn't a problem at all
[17:10:21] <+jaheikki3> And actually that would fit to my schedule a bit better as well :D
[17:12:13] <vohi> a later 6.6.3, together with (or right after) 6.7.0 makes sense tome
[17:12:15] <vohi> a later 6.6.3, together with (or right after) 6.7.0 makes sense to me
[17:13:59] <+jaheikki3> Agree. So let's do the branching from '6.6' to '6.6.3' Mon 11th March. That way we should be able to get Qt 6.6.3 out right after Qt 6.7.0
[17:14:57] <+jaheikki3> And then Qt 6.7 status
[17:15:15] <+jaheikki3> Qt 6.7 API review still ongoing, see https://bugreports.qt.io/browse/QTBUG-119952
[17:16:16] <+jaheikki3> Unfortunately it most probably means that branching from ‘6.7’ to ‘6.7.0’ will be delayed a bit; we need to complete the review (or at least have all fixes in) before branching
[17:16:55] <+jaheikki3> Initial plan was to release Qt 6.7.0 RC Tue 5th March 2024 but it might be too tight, let's see
[17:17:11] <+jaheikki3> Hoping we can freeze the RC content asap so that we could get RC out during next week
[17:17:26] <+jaheikki3> Release blocker list here: https://bugreports.qt.io/issues/?filter=25675
[17:17:50] <+jaheikki3> vohi: what was in your mind related to blocker list?
[17:18:43] <vohi> two things: a) other than the actual filter I'm not aware of any documentation that describes what it takes to make an issue a release blocker
[17:18:56] <vohi> and b) the filter itself, e.g. for 6.7 that would be https://bugreports.qt.io/issues/?filter=25675
[17:20:03] <vohi> the filter right now captures all P1 issues with "fixVersion" set to a 6.7.0* version, and all P0 issues that affect a 6.7* version. I wonder about the inconsistency.
[17:20:32] <vohi> I.e. why would a P0 issue that has fixVersion set to 6.7.0* not be a release blocker?
[17:21:18] <+jaheikki3> vohi: It is true that it's not documented so well
[17:21:22] <vohi> or why would a P1 that affects "6.7.0 RC" not be a blocker?
[17:21:52] <vohi> so I think we can perhaps make it a bit more intuitive and less likely for something that should be a blocker to fall through the cracks just because someone doesn't pick the exact right combination of field values
[17:23:34] <+jaheikki3> But not sure what you mean, P0 issue (affecting to some 6.7 release) will be in 6.7.0 blocker list, all P0 issues affecting to that release are in the list
[17:25:05] <+jaheikki3> But P1 ones aren by default in release blocker list but if we want to add those in the blocker list then we need to add Fix Version(s): 6.7.0 or 6.7.0 RC
[17:25:27] <+jaheikki3> That way we have handled release blockers in the past as well
[17:26:31] <+jaheikki3> Actually a few releases earlier we add any open bug having Fix Version set in the blocker list but then we agreed to add only P1 ones + of course all p0
[17:27:14] <+jaheikki3> vohi: ANd actually that is also stated in https://bugreports.qt.io/secure/ShowConstantsHelp.jspa?decorator=popup#PriorityLevels
[17:28:16] <vohi> I can understand that we might not want all P1s to be blocker for the release; and setting the "fixversion" explicitly to one of the upcoming releases to make it a blocker is perhaps ok; it's mostly hard to find that information.
[17:28:47] <vohi> but I do wonder if P1s reported as a new bug in a 6.7 beta (ie only affecting 6.7* but not 6.6* or older) shouldn't be a release blocker as well
[17:29:03] <+jaheikki3> I don't agree that; that process has been there since 5.0 or 5.1 releases
[17:29:07] <vohi> or at least discussed like a release blocker in some form, so that we don't end up shipping 6.7.0 with known brown-bag issues
[17:29:23] <vohi> many contributors haven't been there since 5.0 :)
[17:30:27] <+jaheikki3> vohi: That's why I try to communicate the blocker list in every status report etc to make contributors possible to see if something is missing from the list
[17:31:36] <+jaheikki3> But that's really an issue: Almost every release we notice some issue too late to be a blocker for the release
[17:33:14] <vohi> sure, sometimes bugs are reported when the release is already packaged an on the servers; we can still discuss whether the bug is bad enough to postpone, if it shows up on the blocker list
[17:33:48] <+jaheikki3> In my opinion contributors are quite well knowing the way they can raise a bug as a blocker for the release and new issues are finding their way in the list pretty well but the problem is a bit older bugs which are staying in the Jira and someone notice those more critical than initially very late in the process
[17:34:38] <vohi> if you think the filter is ok as it is (ie. a P1 reported as a new issue in a 6.7 beta should not be a blocker), then it's just a matter of having clearer documentation about how to create a ticket that will block the release
[17:35:00] <vohi> I can add a page to the wiki
[17:35:11] <vohi> (or add something to https://wiki.qt.io/JIRA-Priorities, perhaps)
[17:36:21] <+jaheikki3> vohi: Ok, thanks. I agree more/better documentation is welcome and I can help also with that
[17:37:21] <+jaheikki3> Would it be best to update https://bugreports.qt.io/secure/ShowConstantsHelp.jspa?decorator=popup#PriorityLevels
[17:37:35] <+jaheikki3> It should be easy to find for everyone
[17:38:08] <+jaheikki3> (I didn't know we have https://wiki.qt.io/JIRA-Priorities :D)
[17:38:41] <Akseli> The authoritative resource is on Jira
[17:39:54] <vohi> any place will do, as long as the other places link to it, I suppose
[17:40:22] <vohi> the difference being that only very few people (not including myself) know how to modify the Jira help page :)
[17:40:43] <+jaheikki3> That's true. I don'
[17:40:49] <+jaheikki3> t know that either :D
[17:41:00] <+jaheikki3> But I know ablasche knows
[17:41:59] <+jaheikki3> I will also increase visibility of blocker list(s) by adding links to it in release wiki pages
[17:44:58] <+jaheikki3> I think it was all at this time. Let's end this meeting now & have new meeting Tue 12th March at this same; I can't join Tue 5th
[17:45:16] <+jaheikki3> thanks for your participation, bye!
[17:45:20] <frkleint> is there a change of a new 6.7 release in the online installer? beta3 is qite dated
[17:46:52] <+jaheikki3> frkleint: Let's see, at the moment there is a regression preventing us to update packages. It has to be fixed latest before RC
[17:47:00] <frkleint> uah, ok ,thanks
[17:47:11] <frkleint> bye

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.qt-project.org/pipermail/development/attachments/20240228/52a79bfd/attachment-0001.htm>