|
Boost Users : |
From: Jens Seidel (jensseidel_at_[hidden])
Date: 2008-03-16 14:06:01
On Sun, Mar 16, 2008 at 10:50:24AM -0700, Jeff Garland wrote:
> The bottom line is that we can't hold a release that's pretty much ready every
> time a new compiler version is released -- we'd never finish...and we already
You forget that gcc 4.3 snapshots are available since years? Most of the
time the problem is simple that existing patches were ignored for some
libraries.
> have enough trouble with that. Now, I'd be all for a 1.35.x to add support
> for additional compilers -- assuming *someone new* wants to volunteer to do
> the work involved. We have a limited number of people that have been willing
> to step up and run a release. All those folks are very busy, and so I'd prefer
> to move right toward a quick 1.36 release...of course new blood to help with
> that would be appreciated too.
I may be wrong but I guess gcc is not just an arbitrary compiler but a
very good one which is also very standard conform. How often does it
complain about valid code? So performing all tests not only with a
single version of gcc but also with snapshots and stable older versions
seems to be no problem for me.
Debian provides for example various gcc versions (but this will change
soon?) and also a snapshot of gcc and the whole archive with more than
15000 packages was already multiple times compiled with it. This helps
to get the archive in a good shape and to detect errors in the snapshot
early.
Jens
Boost-users list run by williamkempf at hotmail.com, kalb at libertysoft.com, bjorn.karlsson at readsoft.com, gregod at cs.rpi.edu, wekempf at cox.net