[maemo-developers] Extras-testing improvements

[From: Tim Teulings]

Hallo!

I (still) suggest to formulate a vision statement, that should clearly
describe the purpose of extras. This formulating should be the guiding line
to define further detailed rules and to judge if a new suggested rule is in
compliance to the existing vision. Since there seem to be two
interpretations of the purpose of extras, this is IMHO really necessary to
avoid that every discussion about a new rule or policy results in a
discussion about the vison in whole. A vision statement should define a
target/purpose and a strategy to reach  the target.

A possible vision statement for extra scould contain the following aspects:
* We want to make open source software available for the maemo platform in
the most convinient way for the user
* We want to make it as easy as possible for the developer of open source
software to publish his software to every maemo device owner.
* We do not want that the over all user experience of owner of a maemod
evice is degraded by installing and running open source software (hmmm,
tricky, in fact I want to say that *other* application or not negativly
influence by the new software and this does not imply anything on the
quality of the software itself).
* We want to aply mechanism that help the developer judge the quality of
his software, the recepted user experience of the end user while using his
sofware and want to give him clear hints how his software could be improved
to better fit the device philiosophy and end user needs.

> If we want an uncontrolled place for apps then we need the
"extras-author"
> someone suggested yesterday.  There is nothing stopping us creating that.
> But, if we do then my prediction is that Extras will be dead within 3
> months.
> If there is a place where user's can look for "latest and greatest" apps
> then
> everyone will enable it, developers will stop bothering to promote their
> apps
> to Extras, and a short time later, everyone's devices will start to die
> because the apps there are not optified, run down the battery, etc.

The question is, what would be the purpose fo the extras-author? Is there a
possible vision statement for extras and extras-authors that make them
distinguishable? For me it looks like, extras-author is just extras without
the tests. What would be the purpose? For me it sounds like a short cut for
the developer to avoid the hassle with the extras QA. But this would just
be a (not that good) workaround for the problem that current extras
processes is not optimal. Should we not spend our time and energy to
improve the extras process instead? If we have a faster walk through for
applications in extras we do not need a "hot new stuff" repository. If we
automate tests, too. If we handle updated better, too. Perhaps extras and
extras -author have the same target but only different strategies?

I think we have set up an initial version of the extras proces. Now we have
gained some experiences. Perhaps it is time (since we also have that
mameo<->MeeGo merge Thing underway) to openly discuss the process, optimize
and refine it. This would would also move us into a stronger position while
discussing future MeeGo infrastructure because we not only do have the
experience but also we have already learned something from them and are
already int he "optimize" stage :-)
 
-- 
Gruß...
    Tim