Re: [Nautilus-list] High level outline of planned client testing. Tellus what's missing,please. ;)

[Ali Abdin <ALIABDIN aucegypt edu>]

On Tue, 19 Sep 2000, Seth Nickell wrote:
> Ali Abdin wrote:
> > 
> > On Mon, 18 Sep 2000, Seth Nickell wrote:
> > 
> > > I think its important to add "extended use without restarting as a
> > > desktop" or something like that. I suspect we have many nautilus
> > > problems that we do not encounter because most of us do not run nautilus
> > > continually as our desktop (e.g. we constantly restart to debug etc). I
> > > know of at least one thing that would probably cause issues if a desktop
> > > were used for a day or so, but was discovered through unrelated means
> > > :-) I bet there are others...
> > 
> > Isn't this called DogFood testing?
> > 
> > I think that all developers should use Nautilus as their 'desktop'
> > (instead of GMC) after they hit the 'bug-fixing' stage.
> 
> The problem is that this is a sort of dogfood developers don't run into.
> I use Nautilus as my main Desktop, but I restart it so constantly in the
> necessary course of my work (for example to test what I just wrote) that
> I don't end up testing it for sufficient lengths of time to find
> pernicious "extended open" bugs.

Dogfood means using the software that you wrote - whether you restart it 
or not. So when you come in to work at Eazel, and you start your computer 
and load up GNOME you get 'Nautilus' instead of GMC, and to start hacking 
you would 'navigate' (using Nautilus) to your file then double-click it, 
and it would open up your favorite editor :P That is dogfood testing.

When you want to play MP3s, use Nautilus. When you want to browse the 
web, use Nautilus. That is the kind of testing I am talking about. 

Restarting Nautilus or not is a non-issue :)