Re: Modulesets Reorganization

[Kenneth Nielsen <k nielsen81 gmail com>]

Hallo

I'm the Danish translation coordinator, and we were encouraged to chip
in on this discussion, so I will ;)

Most of arguments for needing a change and the proposed solution are
just fine by med. But there is one point where I very much disagree.
It concerns not including the "extra relevant features".

I don't know if they should be in the desktop module set, of if there
should be one more with "recommended application" or something. The
point is that right now someone makes a decision, about which programs
are needed for a desktop experience and mature enough for GNOME to put
its name on it and I think that it should continnue being that way,
for the following reasons:

1. It sends a signal to the users. This is core GNOME stuff, this is
where you find feature rich mature applications for your everyday use.

2. It helps us translators prioritize.

2.1.A lot of groups don't have the resouces to translate everything in
the GNOME repo and that is not likely to change. So we look at damned
lies and makes different cuts depending on how far we are in the work
and how many resources we have. Right now we (Danish team) make the
cut at (as a minimum) translation UI for everything on the release
pages, for every release, and that helps us by setting a goal.

2.2 If these applications are not in any way associated with the
release and hence thrown into teh myriade of extra applications (with
everything that is there) then we have to decide for ourselves which
applications are relavant enough, so that they should recieve out
attention first. Such decisions are bound to much more errorprone than
that way it is done now.

3. It will hurt quality

3.1 Many distros follow the GNOME release schedule, and there is a
large overlap between that applications that we now have as part of
the GNOME release and the ones the distros include. This means, that
per automation, the way we (translators) prioritize in GNOME, entails
that a large part the desktop users of these distros see, are already
translated - giving them a better experience.

3.2 If these apps are not bound to GNOME release, end hence are not
subject to GNOME release freezes, it means, first of all that all the
distroes (mentioned above) that pull at the time of their release,
will have a higher chance of pulling something "in between
applicication release" or and older version, because there is not
pressure to wrap up a release. Second of course, it means that if we
want to translator close to release, to reduce the amount of waste, we
will have way more release deadlines to keep track of, a task that is
already close to impossible as it is.

All in all, I don't mind pulling these apps out of desktop, but I
think they should go back into a "recommended apps" release set. In
this release set there wouldn't have to be any rules that we should
only have one app "per task" (referring to banshee/rhythmbox). The
selection should be made so it includes popular, mature apps that many
people will use as a part of their normal desktop use.

Saying that there should be no "official" applications and that we
want to "encurage a strong ecosystem of apps" and that the only way of
doing that it to make them all equal, is in my view a somewhat
misguided attempt at pleasing developers, very much at the expence of
the users. Many, especially new users want recommendation, even though
diversity and freedom of choise is good and they probably enjoy that
too, they want recommendations.

Kind regards Kenneth Nielsen