My take on things from an organisational perspective is quite simple. As other have said, we need to decide what we want to happen.

For me, that means I want development to continue on the Maemo platform, but under the acknowledgement that no new hardware is going to be produced compatible with the closed components, and that directs the approach that needs to be taken. In order to perpetuate Maemo's legacy we need to ensure it's code can be easily used elsewhere, and that means having somewhere to focus closed-source-replacement and portability efforts. I can appreciate why stkeeps decided the Meego CE was the way to achieve this goal...but I can't help but feel it throws out too much of the baby with the bathwater. I feel we need to become the similar to the Meego infrastructure, just obviously with a greater focus on Maemo.

To me, that means we need development infrastructure: mailing lists, forums, the wiki, a bugtracker, source control, irc logging, and of course, the autobuilder. Everything else is just fluff. Once we have those components, we can integrate them far better than is currently the case.

Cordia is a great project, but I'm always wary of forks as they usually just fragment an already small developer base. I want the bugs they've fixed and the code improvements they're making included in the software running on my Maemo device. We can bring Maemo to Meego if you will, but we shouldn't leave Maemo behind. Likewise, all the love going into the CSSU should serve to help the Cordia project.

I'm struggling to think of examples of this for illustration, but I guess the best I can manage is the awesome cooperation between the ReactOS and WINE teams (or perhaps the relationship between the various WINE forks).

I guess that' brain dump will do for now.

...oh, one additional side note is Qt's Maemo support (overall, i.e. Creator, not just the libs). It's treated very much as a bolted-on feature...do we know how long that will hang around before it gets "cleaned away"?