Tracking issue for the 2.5 release.
semantic-mediawiki.orgcomposer updateLocalSettings.phpgit shortlog -sn 2.4.0...2.5.0))//docs/semantic-mediawiki.orgcomposer update LocalSettings.phpMaybe its time to start making preparations since we are ~540 commits into 2.5.0-alpha (2.4.0 was released on Jul 9) and avoid some frustration and time constraints as I recall from the last release process.
I don't have anything big left on my task list with the following features being available:
+lang=...@@@ annotation syntax supportI'm not yet ready to call for a code-freeze but I'd like some people to use the 2.5.0-alpha version (its pretty stable otherwise our tests would tell us something else and the sandbox seems fairly stable as well) to get some early feedback as a lot of extra refactoring was necessary to support those listed features.
Tests went from 2.4 ~85% coverage (Tests: 3685, Assertions: 12014) to preliminary Tests: 4300, Assertions: 13891 (~87% coverage)
I still think it'd be nice to just create a release whenever a new feature is available. Put it in the release notes, create a tag, done. No need for code freeze or a long process of trying to figure what happened half a year ago. It is what I do with Maps, and it is a lot more fun for sure.
--
All good from my side
This builds up pressure on me to do at least a minimal documentation. I will try to do my very best.
The dates where are wrong now: https://github.com/SemanticMediaWiki/SemanticMediaWiki/blob/master/docs/COMPATIBILITY.md
Tests went from 2.4 ~85% coverage (Tests: 3685, Assertions: 12014) to preliminary Tests: 4300, Assertions: 13891 (~87% coverage)
I'm officially done with 2.5, now that master contains +850 commits for the 2.5 release cycle (compared to 2.4.x) and amounts a total of 4652 tests (Assertions: 14830) it is time to close the release.
Updated the initial comment with the list-o-doom
@mwjames @JeroenDeDauw What about doing -rc0 today pinging testers and release next week according to the schedule?
rc0? One starts with rc1 AFAIK
Fine from my side
Thank you @mwjames for making the release notes so fluff right off the bat!

rc0? One starts with rc1 AFAIK
Yeah, it is rc.1 according to spec-item-9ff. I got off track since MW uses rc.0 O_o
@SemanticMediaWiki/testers It will be nice if you could test and provide feedback. Thanks a lot.
Yeah, it is rc.1 according to spec-item-9ff.
Bleh, that's not what we've been using... does not really matter I suppose
Bleh, that's not what we've been using... does not really matter I suppose
Oh, I guess semver is more complicated that one may expect. No, does not really matter as long as we get the regular releases right. I myself was more unhappy that the mailinglist munched all linebreaks. I was in grump as I saw this.
@mwjames Note that sandbox still runs master thus updating beyond rc.1
@mwjames @JeroenDeDauw I will add some more docu to the wiki today. Since we got translation updates today we could release tomorrow. Any objections?

Wow things are moving soooooo faaaaaaaaaaaaaaaaaaaaaaast
All OK from my side. Want me to do anything?
Want me to do anything?
You could somehow retrieve the contributors. If you want https://github.com/SemanticMediaWiki/SemanticMediaWiki/issues/2318 in you could act on this one, too. + merging your pull request of today. I guess the rest may be up to me.
I'll have a look at the stuff tomorrow
The array syntax stuff provides no benefit to the user. Since there is some chance it breaks something, I'd be included to merge this AFTER the release. I suppose that does make backporting harder, so it is a bit of a tradeoff.
git shortlog -sn 2.4.0...master
688 mwjames
54 Karsten Hoffmeyer
38 Jeroen De Dauw
37 Niklas Laxstr枚m
14 translatewiki.net
13 jeroendedauw
5 Kghbln
5 Maciej Brencz
4 Siebrand Mazeland
3 jongfeli
2 Alex Winkler
2 Toni Hermoso Pulido
1 Amire80
1 Felipe de Jong
1 Jaider Andrade Ferreira
1 James Forrester
1 Justin Du
1 Seb35
1 Sophivorus
1 Stephan Gambke
1 Virginia Cepeda
1 s7eph4n
Ok, let's get 2.5.0 out to the world.
Given that we are now a fortnight into having 2.5.0 out and seeing the issue reports we hardly had any issues so far. This is a leap forward in comparison with previous releases. So thank you @mwjames for such robust programming flanked by automated testing.