Scheduling releases of 5.4.0 and 5.3.3

[Kevin Funk]

On Friday, 5 July 2019 17:01:23 CEST Friedrich W. H. Kossebau wrote:
> Hi, especially Milian, Kevin, Aleix, Sven,
> 
> seeing people adding .vscode gitignore rules to KDE projects I sense
> something is not working well with KDevelop's reception in the public. even
> in KDevelop's shell community KDE.
> 
> 
> NO RELEASES IN MONTHS? HOW DEAD IS KDEVELOP?
> 
> One thing might be that releases are not happening, showing both no activity
> to the public as well as not making bug fixes & other improvements finding
> their way instantly to users.
> 
> 
> LET THE RELEASES ROLL AGAIN: 5.3 NEXT WEEK?
> 
> So, let's improve here, and kick out some 5.3.3 release soonish: there are a
> few commits with "fix" and translation improvements worth to get to users
> (see e.g. https://cgit.kde.org/kdevelop.git/log/?h=5.3)
> What about end of next week?

Disclaimer: We had some internal discussions going on here, and just to show 
the maintainers have not jumped ship, here are some of our comments:

Friedrich suggested to get the ball rolling and release 5.3.3 soonish and 
Aleix said:
> The information in https://community.kde.org/KDevelop/Release_Process should 
> be relatively up-to-date. I used it as a personal "remember what to do" page 
> recently.

Me: Yes, that should work.

I'm not too sure whether the Windows installers are up-to-date and working, 
but I guess without Sven or me looking after them it's worthless to keep 
providing them anyway. In that case we're effectually releasing untested 
binary products. 

There's probably not much outcry if they're not being released either way.

One thing I'd try to get up and running again is at least the AppImage builds, 
but see my other mail for that.

> ... AND 5.4 NEXT MONTH?
> 
> At the same time IMHO we should schedule 5.4 release for end of July. I find
> it not motivating to review people's patches (where there thankfully are a
> few again recently and I will see to do my share the next days) when it is
> totally unsure if & when more people will gain from that review investment.

Maybe lets' check how quick we can release 5.3.3 and once you (Friedrich) feel 
comfortable completing the release process you can keep it up...? :)

And: I'm agreeing on the 'source tarball release is better than no release' 
statement. 

Cheers,
Kevin


> DOING SOURCE TARBALL RELEASE IS BETTER THAN NO RELEASE
> 
> I know current maintainers are short of time for KDevelop matters. And a
> perfect release would include things we got used to in the past, like ready-
> to-install AppImage & Windows bundle provided by KDevelop team itself.
> 
> But given no-one is around anymore who has enough time to care for those
> extra services, let's at least release the minimum, source tarballs, so the
> classic distributions can serve their users.
> Some people happy is better than no people happy, would you not agree? :)
> And once there are people again to create bundles, future releases can be
> extended again to include those.
> 
> And poking for this, of course I am available to help with some workloads,
> just tell me what to do. Okteta source tarball bug fix releases take me 15
> minutes in total, incl. announcement email ;)
> 
> Cheers
> Friedrich
> 
> 
> PS: Seems that after all there still is a small user community out there
> being attached to the joys of "kdeveloping" ;) see mentioning in
> https://isocpp.org/files/papers/CppDevSurvey-2019-04-summary.pdf
> Okay, landing after Kate only. But then some people do not big code-bases ;)
> 
> PPS: Thanks for your patience with all the unreviewed mass code maintenance
> commits to master, it just felt bad to me to have the sources of my coding
> tool be so behind the standard of the code I work on with (as well as not
> being prepared for the future). and when there was an occasion, I just ran
> wild & insane to go forward here, having done similar work before, thus
> trained to patterns.
> Of course there is some risk in all this, but next to unit tests watching I
> have been and will be dog-feeding things myself all the time, and so far
> only hit one mistake (before pushing thankfully), so quite relaxed that the
> gain of the moderate update of the code base to the world rolling is bigger
> than any potential breakage costs. And yes, flush almost over, will fill
> the normal review pipelines again after that as again well-behaving
> contributor. Until then thanks for your enforced trust in my work ;)


-- 
Kevin Funk | kfunk at kde.org | http://kfunk.org
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 195 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.kde.org/pipermail/kdevelop-devel/attachments/20190708/6f12e2dd/attachment.sig>