kdiff3 and kde5

[Greg Troxel <gdt%lexort.com@localhost>]

"David H. Gutteridge" <david%gutteridge.ca@localhost> writes:

> On Wed, 30 Oct 2019, at 17:45:06 +0100, Kamil Rytarowski wrote:
>> Status.
>> 
>> I have upgraded what possible (around 20 packages).
>> 
>> There are pending actions for:
>> 
>>  - picard (upgrade patch shared with nia@)
>>  - bzr-explorer + qbzr/qbrz (rhialto@)
>>  - gnuradio packages (gdt@)
>>  - gns3 (markd@)
>>  - fbreader (maya@)
>>  - scidavis + qwtplot3d-qt4 (bouyer@ ?)
>>  - scribus-qt4 (nia@)
>>  - merkaartor (gdt@)
>> 
>> There is KDE4 left with its dependencies.
>> 
>> Besides KDE4, there are also:
>> 
>>  - qbrew [dead]
>>  - xxdiff [quasi-dead]
>>  - FlightCrew [39 revbumps?]
>>  - ibus-qt + kimera + scim (ryoon@ ?)
>>  - owncloud (ryoon@ ?)
>>  - bacula-qt-console + bacula-tray-monitor (markd@ ?)
>> 
>> Nice to have is to make py-qt5 aware about multiple python versions.
>> This was recently enhanced by bouyer@. This mostly affects software
>> that
>> is restricted to python2.
>
> I've updated kdiff3 from 0.9.98 (Qt4, the package was last updated in
> pkgsrc five years ago) to 1.8.1 (Qt5); it's ready to commit. kdiff3 is
> not part of the kde4 meta-package, though notionally it's a "KDE"
> package. Is the preference here to simply update in place, or to create
> a new package, so the old Qt4 version is still available? (I'd prefer
> to just update in place, as I don't personally see the need for an
> older Qt4 version anymore. But there's been discussion about retaining
> all KDE4 packages for now, so I ask.)

I'd like to hear from markd before we do anything that damages the
current state of kde4.  The number of people that have expressed
opinions is fairly small, and I think I've seen "kde4 mostly works and
is of some use" and "I tried to build it and couldn't".

If kdiff3 is not depended on my the kde4 meta-package, and if someone did
install kde4 and (assuming all worked), and then if they used kdiff3 and
that worked, then it seems good to update it.


I would like to see the goal being having fully satisfactory kde5, while
doing as little damage to kde4 as is sane*, rather than having a goal
that smells like "rm -rf qt4, because I don't think anybody should be
running it'.

* at least until the "if we delete it nobody will notice or care" usual
  test is true, and it seems like right now that's really not the case.