4.6 roadmap of sorts

[Stephan Arts]

On Sun, Apr 6, 2008 at 10:33 AM, Olivier Fourdan <fourdan at gmail.com> wrote:
> Hi all.
>
> On Fri, Apr 4, 2008 at 11:57 PM, Brian J. Tarricone <bjt23 at cornell.edu> wrote:
> >  Moritz was kind enough to set up a sort of roadmap page to help gather
> >  in one place the things that need to be done before 4.6 can be released.
> >   You can see it here:
> >  http://wiki.xfce.org/roadmap_to_46
>
> That's really great and much, much needed. I would like to take this
> opportunity to discuss the way we are  I am not sure it addresses one
> of the main concern I have with our development scheme.
>
> Don't shoot yet, let me explain :) we are a very small group of
> developers with very little time, so placing the bar too high may
> cause severe delays in the releases cycle and may discourage some of
> us.
>
> Wouldn't it better to choose a time based release cycle, a date we all
> agree so that we can then decide what can goes in and what cannot,
> depending on each developers availability and agenda.
>
> If something cannot make it in tome for 4.6, no big deal, it will
> hopefully make it in 4.8 but at least we shall have 4.6 out before
> everybody forgets about Xfce. But since we all know and agree on the
> date, we can arrange to make 4.6 stable even if this is not as feature
> rich as it was initially planned,
>
> What do you guys think? Please note that this is not incompatible with
> current roadmap, it is just that I would like to see some kind of date
> in this roadmap otherwise we'll end up doing the same as with 4.4, be
> more than 1 year beyond schedule (even if schedule do not mean much
> for us, it's still important I think).

I can not agree more :-)

I have some ideas about how the configuration-dialogs can be improved.
But trying to write those for 4.6 would probably push back the release
further. Therefore I agree with Brian that this is best to be
considered for 4.8. As an alternative I will import my code to the
goodies repositories (xfce4-settings-manager or something), you can
then follow its development, and comment at it. And we can discuss if
it is something for 4.8 or not.

A fixed release-date (like once a year, I doubt 6-months is doable for
us), will increase the exposure of xfce. Not only that, but it makes
it easier for us to hold off features for future versions. When you
take a look at the changes between 4.2 > 4.4 and 4.4 > 4.6, after each
release we have rewritten at least one component from scratch. In
general these changes are all improvements. But such huge changes come
with the price we are all to familiar with. The release is postponed.

When we pick a release-date, say: "November '08", then it we can pick
dates for feature-freezes, string-freezes and beta releases. I think
it makes developing xfce more fun, because it makes us see the fruits
of our labour sooner then is the case now. The 4.4 release was
'delayed' for over a year (as far as we can speak of any delay, since
we haven't missed a deadline ^_~). And we are currently developing 4.6
for over a year, with no release in sight (yet).

I'd say, pick a date and work towards that. And as long as there is no
'blocking' bug open at that time, push the release.

>
> Cheers,
> Olivier.


Regards,
Stephan