Decision on gccbug?

[Volker Reichelt]

Ben Elliston wrote:
> Wolfgang Bangerth <bangerth@ices.utexas.edu> writes:
> 
> > think were reasonable given the typical length and style of a
> > PR. But people were quite set on this and I gave up when I became
> > frustrated. If you ask me: nuke it.
> 
> I think my point was missed.
> 
> In the same way that we don't simply remove target support without
> warning users for a release or two, I think it would be poor form to
> simply nuke gccbug.  At least one release should be made with a
> version of gccbug that explains that it is no longer in use and that
> users should file bug reports into Bugzilla.
> 
> If necessary, I volunteer to write such a version of gccbug.

I think gccbug should just be deleted. It's not a part of the compiler
that is used regularly: You only use it, if you find a bug. You don't have
to prepare your code for a transition in the future.
Therefore, a deprecation period is bogus IMHO.

It's just a tool for convenience that costs maintenance time.
So just don't waste more time on it.

When we switched over from GNATS to bugzilla we didn't keep the GNATS
interface either. We just turned it off.
Taking into account the number of actual uses of gccbug, we should just
remove it IMHO and document the removal in the bug reporting instructions.

Regards,
Volker