Toulouse 2015 Meeting Notes

Daniel fixed something (kdepimlibs unittests in master, except 5)
He also fixed a bug in akonadi server
Christian discussed upstreaming zanshin work of kolab with Kevin
All agreed to have the next sprint in Randa (kevin won't be joining)
Talking about ABI/API level of kdepimlibs
close ~850 bugs (kmail1)

The Future of KDE PIM

we will do Qt5 release of the entire KDE PIM stack ASAP
most things should already be working (TM), we just need to make sure there are no regressions caused by behavior changes in Qt 5
after the release happens, Christian will start working on Akonadi Next
Dan (and others?) will work on removing Akonadi code from applications and moving it down to a conversion layer (basically what Zanshin does)
this will make it easier to do the switch to Akonadi Next later on
we can also keep adding new features and of course providing bugfixes
at some point we decide that Akonadi Next is ready and KDE PIM applications are ready and we simply do the switch
this will make everyone happy: devs because they can work on what they like and users because their favorite PIM suite will be well maintained, and not in limbo for several years
before Akonadi Next would be ready

To make the KDE PIM Qt 5 release happen, we want to move some kdepimlibs modules to Frameworks ASAP. The remaining modules (including Akonadi libraries) will be released as part of KDE Applications.

Move to Frameworks (Ready right now)

gpgme
kcontacts
kldap
kmbox
syndication

Make into Frameworks (Requires removing KDELIbs4Support)

KMime (depends only on KCalendarSystem, should be super easy to remove)
KImap (depends on KMime)
KCalCore (KDateTime -> QDateTime)

The ordering means priority + how easy it is.

We will want to change how some frameworks work in future (for instance KImap), but for that we simply release "KImap2"