[maemo-developers] Community widgets for Fremantle

[From: quim.gil at nokia.com]

> > I agree. Maybe I'm wrong but I think Hildon updates are possible in
> > Fremantle. As long as the new widgets make sense, the patches are good and
> > they are well tested by the community and don't introduce risks or
> > regressions... Why not?
>
> Because there's a perceived *risk* of risks & regressions. How many of
> our App Manager patches shipped to Diablo users once Maemo Devices had
> moved on to Fremantle? None.
>
> Is there any reason to suspect that the situation will be different
> with Fremantle once Maemo Devices is pushing for Harmattan?

Yes, there are reasons that come to mind:

- Hildon is a library. Adding widgets doesn't affect the user experience, testing, etc of current widgets. If a new widget turns out to be conflictive it won't affect any official app since they will stick to the widgets used now for the Maemo 5 final release. Therefore the whole discussion of accepting enhancements is made more among platform developers without adding much of UI designers and product managers to the mix. If the changes don't bring any straings for localization all the better. This is very different compared to a pre-installed application.

- App Manager Diablo was moving to App Manager Fremantle both inside the Maemo team and there were clear project management instructions about where to put the resources. Instead, the Desktop team has mostly two different subteams for Hildon and the Qt based application framework for Harmattan since both projects require different timings and skills. The Hildon skills and mid term decisions are nowadays in practice more in the hands of GNOME developers working for Igalia, Lanedo, Openismus, Collabora... If they want to push Hildon forward and your patches and commitment are convincing you will need to deal with less and less Nokia internal project plans.


- It is clear that Hildon must go upstream around GNOME and GTK+ or will have clearly no sustainable future. The Maemo team needs to start handling this transfer of ownership. The schedule is not defined by the end of the Fremantle timeline but by the beginning of the Harmattan timeline. You have seen the Maemo Summit schedule and you know this time is coming already.

- The Maemo team has learned a thing or to, thanks to many factors including the App Manager Diablo contributions fiasco. We have contribution guidelines, we had Fremantle pre-releases since pre-alpha and you have here non-Nokians like Claudio or Berto able to discuss openly details about their are of responsibility and run entire garage projects and git repositories.

You will know better the technical decisions about packaging etc but at least you know my opinions as one of the persons in the Maemo team responsible of finding the best future for Hildon as a community supported project.

I hope you can keep all Hildon contributions around the Hildon trunk repository and within one real project. Hildon needs all the community interest and support concentrated in one point. Spreading energies can be too expensive. Like Claudio says, if the problem is the distribution then let's concentrate on the root of the real problem while developing code and features in the optimal way.

--
Quim Gil