[maemo-developers] Looking for help from a GTK guy with 2 projects

[From: Michael Hasselmann]

On Sun, 2011-11-13 at 17:53 +0800, Jonathan Wilson wrote:
> > Have you considered Maliit?
> > http://wiki.maliit.org/Documentation/Installing#Maemo_5_.28Fremantle.29
> That looks nice but:
> 1.Its clearly stated that its use on Maemo5 is "not ready for prime time"
> 2.Its written in QT (and not only that, it uses QT version labeled 
> "experimental") and takes over a lot more of the input system than I would 
> like (including not using Maemo keyboard layouts which is something I want 
> to keep doing)
> and 3.There is no way Maliit or anything based on it would be accepted into 
> the CSSU because of #1 and #2. (and the CSSU is the eventual target for any 
> keyboard replacement.

Huh? You sound a bit confused, my friend. Since when would being
Qt-based be a reason against inclusion in CSSU? These days, QML + PySide
is all the rage, and Qt has been in base Maemo5 image for ages now.

Maliit is not ready for prime-time for Fremantle because no one is
actively working on it. What I did back then was a proof of concept,
nothing more.

We do have GTK+ support and we even took in some matchbox-specific
patches. But you need to scratch your own itch. You cannot expect others
to do it for you. That's just how open-source works. So yes, it requires
someone to work on it, and while it won't be me, I'll gladly accept
patches. In an open-source community, your first argument simply doesn't
hold.

You also seem to ignore that it wouldn't only be the MeeGo Keyboard from
Harmattan that you could get (together with all the language layouts
that have gone through extensive usability tests). Maliit is a fully
pluggable framework. Hildon IM was that, too, which is no surprise, as
Maliit is the legitimate successor of Hildon IM – in fact, it's been
designed by the same architect. The architecture allows the possibility
to run other Maliit plugins, Swype being one of them (on the N9), see
http://wiki.maliit.org/Plugins

Sorry to sound harsh, but your arguments appear to belong in the
category of rationalize-after-the-fact (of having decided to go with
reverse engineering hildon-im-vkbrenderer). I know that you're looking
for a cool hack, and I wish you good luck with that. But why not invest
your time in something with wider benefits and contribute to an active
project?

regards,
Michael