Re: 3.3-branch QA assessment

[Richard Earnshaw <rearnsha at arm dot com>]

> 
> 
> --On Friday, January 24, 2003 06:05:26 PM -0800 Joe Buck 
> <jbuck@synopsys.com> wrote:
> 
> >
> > I've been looking at our high-priority regressions and have come to a
> > somewhat disturbing conclusion.  Now, regressions aren't the whole story,
> > but based only on the regression data, it seems that 3.3 is in
> > substantially worse shape than either 3.2.x or 3.4, begging the question
> > of what the point of the 3.3 branch is.  (Please read to the end before
> > you flame).
> 
> You raise an interesting point.
> 
> I think it is worth it to spend some time fixing regressions in 3.3 so
> that it *is* better than 3.2.  The 3.4 numbers would look a lot worse
> were it not for the fact that I've been frantically trying to fix new
> parser bugs as quickly as possible; that's kept things down there.  But,
> there are probably going to be a lot more bug reports; I sure don't
> expect that I wrote 15,000 lines of completely bug-free code.
> 
> So, I think partly you're seeing where effort has been focused.
> 
> There's already the new parser and PCH in 3.4, and we might have rtlopt
> as well.  That's a lot of new technology, and my guess is a lot of new
> bugs.
> 
> I'd bet we can cleanup 3.3 without *too* much effort, and provide people
> with something that is better than 3.2, but that does not have a lot
> of the instability from all the major new technology.
> 

Right, and the danger here is that we fall into the same trap that Red Hat 
did with their "2.96" release -- that we only look at the existing bug 
reports and testsuite scores and ignore the likely impact that the changes 
made will have on applications that are "out there" but haven't been fed 
through the new changes yet.

R.