[Qgis-developer] Re: composer redesign branch

[Martin Dobias]

On Feb 13, 2008 12:04 PM, Tim Sutton <tim at linfiniti.com> wrote:
> Hi
>
> 2008/2/13, Martin Dobias <wonder.sk at gmail.com>:
> > I completely agree with RFC process and will respect it. The problem I
> > see with RFCs is that they get nearly no feedback. Geometry handling
> > RFC which I've created several months ago got comments only from
> > Marco, voting has been postponed and then forgotten :-/ I'm wondering
> > what's the reason - is it written unclear? Or is that just lack of
> > time of others?
>
> Yes this is a similar problem to many other areas of the project (e.g.
> so far we got only me writing unit tests for new features going into
> core even though it was agreed that any new code going into core would
> be accompanied by tests, e.g. so far the idea of committing code with
> no warnings didnt take etc). However lets not give up yet - sometimes
> it takes a while for new ways of working to 'settle in'...

Ok, I understand. I just thought that it's better to participate a bit
during design stage than later be surprised of the changes :-)
Regarding the other areas... unit tests are also important to keep
some level of quality, to ensure that tests get written why not deny
merging of new branches if they're not accompanied by tests? Finally,
I don't see warnings in code that important - since you can't assure
that new code will compile without warnings on all platformas,
moreover fixing a warning is usually one minute issue. However,
remembering our conversation on IRC some time ago, we probably look at
it from different sides :-)

Bye
Martin