Qt 3.2 requirement

[Stephan Kulow]

On Tuesday 29 July 2003 18:48, Rob Kaper wrote:
> On Tue, Jul 29, 2003 at 06:17:28PM +0200, Guillaume Laurent wrote:
> > "decent layering" is a nice theory. "Fixing the source of the problem and not 
> > symptoms" is another. None withstand the test of reality.
> 
> That's only true when we don't practice those theories.
Oh! How long did it take you to come up with a phrase that useless?

I'd really welcome if people could come up with arguments instead of phrases.
Marc at least tried, even though his arguments didn't convince me so far
we're doing something wrong (But I better continue reading - I see there are
some mails left to read :)

I'm actually not questioning pracmatism is a good thing to do. We simply do
not have time for things nice to have. So the only balance I see in question is:
how many bug fixes we loose when developer A puts in a feature/bug fix that
requires developer B to compile instead of code. 
But if this "support Qt 3.1" burden creates more work to those doing the fixes,
then I prefer removing the burden. Every line added with #ifdefs _is_ work,
so it needs to be balanced. 

So far supporting Qt 3.1 isn't a huge problem, but with every developer switching 
to 3.2 it becomes a bigger problem as 3.1 becomes more untested and untested 
software is _bad_ software (another phrase to add to your diary, Rob) - so I for 
myself prefer starting a compile job overnight and from then beeing sure I test 
what will be KDE 3.2 instead of saving the CPU cycles for the screensaver. That 
many others do not share that, I'm aware of though.

Greetings, Stephan