Deprecating Old Amarok Versions

[Luigi Toscano]

Boudewijn Rempt ha scritto:
> On woensdag 2 oktober 2019 18:12:46 CEST Luigi Toscano wrote:
> 
>> Regarding this new, different topic that you just brought up for discussion,
>> and waiting for some words from the remaining amarok contributors, if I had to
>> choose I'd advise to not deprecate the current amarok repository (whose master
>> is KF5-based) until we switch to Qt6.
> 
> Nobody was talking about the repository; the proposal was to remove the tarballs of a qt4-based amarok.
> 
I disagree, or at least there is a terminology problem.

https://mail.kde.org/pipermail/release-team/2019-October/011546.html
This answer talks about deprecating applications. All Qt4 stuff are already
deprecated, and (with one sad exception) stopped shipping them. We don't
extract their translations anymore (with one exception, hopefully not for long).

So there is really no further action item that needs to be taken, unless
"deprecating" is interpreted in the way of "moving to unmaintained", hence my
answer.

Regarding the tarballs, I'm not sure we can remove them completely. You may
argue that their content can be reconstructed from the git tags, but let's
play it on the safe side and keep them somewhere, also for historical reference.

Should we move them to a separate place? Sure, why not. Just remember that
this is going to be more work for the sysadmins, so it's up to them to decide
when (also because some applications have mixed Qt4 and Qt5 tarballs in their
/stable/ or /unstable/ directories). As I wrote, I think it should be more
consistent than just moving one specific tarball, and that in general this
move is not really going to provide us any relevant gain.

-- 
Luigi