> > Remember when Windows 2000 was released? It was one of Microsoft's most
> > stable and reliable OS releases ever, despite a list of over 60000
> > known issues.
>
> That doesn't say much good about WIN2000!
Here's a scenario. There's a bug open "Implement support for <LINK>".
We're about to fix it, but there are a few extra things we'd like to do
after it's fixed. So, when it's checked in, another few bugs will be
opened - stuff like "Implement HTTP LINK headers" and "Make link toolbar
support <A> element (or not ;-)" and "Get proper icons for links toolbar".
So, this checkin makes the number of "bugs" in Bugzilla go up. Does it
mean Mozilla is more buggy? That's not a question with an obvious answer.
Also, the number of bugs currently being submitted to Bugzilla is greater
than the number being fixed - so the number of open bugs is rising
inexorably. Does this mean that, as 100 developers work on it every day,
they are making the software buggier and of lower quality? No.
Open bug counts are not a reliable metric of anything except the size of
the Mozilla testing community.
Gerv
