[Qgis-developer] Release Today 23h59 GMT

[Tim Sutton]

Hi

On 12/29/06, Maciej Sieczka <tutey at o2.pl> wrote:
> Paolo Cavallini wrote:
> > Your position may be correct, but:
> > - the "Must fix" label does not have meaning in this context - please
> > remove it from the trac
> > - at least critical bugs should be fixed, otherwise what is the meaning
> > of "critical"?
>
> Maybe Paolo's statement is quite radical, but I second him. Either-or.
>
> > - what is the position of the release team about maintaining 0.8? I
> > imagine most developer will move to the (much more exciting) 0.9 svn,
> > and bugs in 0.8 may simply stay as such until 0.9 release

0.7 had 4 bug fix releases why would we treat 0.8 any differently?


>
> ... which will have it's own bunch of critical bugs due to new
> features, thus the average stability of QGIS might remain not good
> enough for deadline-driven usage.
>
> I'm affraid the scenario suggested by Paolo is not that unlikely. I
> realise how unconvenient it actually is to maintain 2 or more branches,
> doubling your work for having to backport changes between them. Thus
> the 0.8 branch is likely to be soon given up and all the effort moved
> into 0.9. So 0.8 might remain unstable forever.

At the moment we have some effort in bug fixing in almost no effort in
developing new features. Lets get the balance back and allow
developers to get on with their work again while at the same time
giving people who want to fix bugs can use qgis 0.8 as a base that we
can continue to stabilise. We did this in 0.7 and it worked well I
think.



>
> > (hopefully much faster than the current one, BTW!).
> > QGIS is already used in a variety of professional environments, and
> > these are delicate issues, as you can realize.
>
> A valid point. How can I ask my colleagues to use software which I know
> is not stable, and take the responsibility for delays and problems in
> their proffessional work?

As in my last post, the simple answer is 'dont'. Wait until you are
happy with the release (0.8.x or whatever) and then prompt them to
move. Same as you would do with any other software.


Moreover, once they try QGIS which crashes on
> them, they'll quit it and will not be that keen to try it again in
> future.

Dont give it to them to use until you are happy with the stability.
There is no universal rule that says you must deploy this on release.
There are many other people out there who will be grateful to have a
new release and make use of the new features in QGIS even if there are
still a few bugs. Its up to you to choose...it is 'free' software
afterall.


 The first user's approach should be successful for QGIS to take
> over on "other software". And the more successful take overs, the more
> users, developers and sponsors.

>
> Please don't get me wrong. I've been using QGIS and I'm gratefull for
> it. I hope QGIS community recognizes my (humble) contributions too. I'm
> only wishing QGIS all best, and I believe an unstable release is not
> good for QGIS.


In an ideal world I will agree that putting something unstable out
there is not good. In an ideal world I would also say that > 1 year
between software releases is not good. But we need to be pragmatic and
realise that making qgis 100% foolproof could easily take many more
months. In the meantime we lose many advantages that the open source
rapid release cycle offers.

In summary if you are not happy with the quality of 0.8 just hang for
0.8.1. Look at the maths:

svn trunk + XX months or weeks == very stable release
0.8 +  XX months or weeks = very stable release

It takes you the same time to have a stable release on your desktop
and in the second scenario people who are not interested in bug fixing
are able to get on with other work.

Regards

Tim


>
> Maciek
> _______________________________________________
> Qgis-developer mailing list
> Qgis-developer at lists.qgis.org
> http://lists.qgis.org/cgi-bin/mailman/listinfo/qgis-developer
>


-- 
-- 
Tim Sutton

Visit http://qgis.org for a great Open Source GIS
Home Page: http://linfiniti.com
Skype: timlinux
MSN: tim_bdworld at msn.com
Yahoo: tim_bdworld at yahoo.com
Jabber: timlinux
Irc: timlinux on #qgis at freenode.net