Re: gcc-2.95.3 is released.

[Marc Espie <espie at quatramaran dot ens dot fr>]

In article <Pine.LNX.4.30.0103161438360.1311-100000@host140.cambridge.redhat.com> you write:
>I'd like to ask everyone who works on OS distributions to be a bit more
>careful with version numbers.  If you apply patches for your release,
>_please_ make sure that your patched version clearly identifies itself,
>e.g. as "2.95.3 (debian)", "2.95.3 (OpenBSD)", or whatever.  Please do
>not increment the version number, but also do not leave the version string
>unchanged.  People have done strange things to gcc-2.95.2, and this has
>been a source of problems while doing regression tests for gcc-2.95.3.

I'm not quite sure how to handle that.
On one hand, sure, it is fine to tag gcc so that you know the bug-report
you've got is not quite from a vintage compiler.

On the other hand, I'm wary. I'd like some warranties that a bug-report
that says "2.95.3 (OpenBSD)" won't be summarily discarded as in
`oh, it's not really our fault. Let the OpenBSD people deal with it.'
or (worse)
`oh, well, it's not the true redhat linux compiler. We don't need to support
it.'

I guess that it will come down to good faith, but I'm serious about it.
I would really appreciate some guarantees that support will still happen.

For starters, you know as I do that OpenBSD 2.9 will include a 2.95.3 plus
the sjlj fix-ups, as we're neither HP-UX nor AIX.