On Saturday, 8 December 2012 04:50:47 CEST, Mark Constable wrote:
Whew. Please please resist the akonadi disease under all circumstances, no matter what they throw at you as to the "good" reasons why "akonadi integration" would be a good thing.
I believe that Trojitá's IMAP stack is technically superior to the IMAP resource in Akonadi -- in fact, the efficient and fast IMAP implementation is the whole reason why Trojitá is here in the first place. Switching to AKonadi would constitute a severe regression at this point, so it is not going to happen. On the other hand, the KDE project provides quite a lot of convenience classes, especially for the UI. It is intirely possible that in future versions, message boxes in plain Qt version will be augmented by likes of the KMessageWidget [1] in the KDE version. I'm also definitely open to working together where it makes sense. Especially with the (slowly) upcoming KDE 5 Frameworks which will make KDE much more modular, I'd love to unify the low-level bits like the RFC2047 decoder etc. (And in fact, other low-level bits in Trojita were copied from KDE before, like the modified UTF-7 decoder required by IMAP.) There's also an experimental SMTP library in KDE, and given the status and reliability of the current implementation, there won't be many regressions introduced by switching. The message should be clear here -- I like the KDE project and respect their work, but have no intentions to depend on kdelibs in future. If such a dependency gets brought up, it will remain optional, if only for the Windows versions. Cheers, Jan [1] https://git.reviewboard.kde.org/r/101249/s/141/
