Re: GMP transition: 4.3.2 to 5.0.1?
On 17.03.2011 19:42, Julien Cristau wrote:
> On Sun, Feb 6, 2011 at 12:39:23 -0600, Steve M. Robbins wrote:
>> Now that squeeze is out, I'd like to move from GMP 4 to GMP 5. The
>> latter was released upstream about a year ago and the gmp lists
>> aren't buzzing with outrageous bugs, so it appears stable enough.
>> I know GMP is used in gcc itself, so I'd appreciate some guidance
>> from the gcc team as well, since this is a major version change.
>> I uploaded GMP 5 to experimental last year and it builds fine with two
>> exceptions: hppa is BD-Uninstallable (?) and ia64 fails to build with
>> an ICE. The underlying cause is already known  and from reading
>> the bug report I suspect I may be able to work around this by building
>> one file with -O2 rather than -O3. Other suggestions welcome.
>> Main question: should I go ahead and upload the new version when
>> I get a free moment or do we need more investigation?
> So this is going pretty badly. gmp has a *lot* of reverse dependencies.
> That means:
> - an SONAME bump has a big impact on the rest of the archive
> - an SONAME bump is a bad idea
> - an SONAME bump without versioned symbols in both the old and new
> libraries is going to end badly
> - an SONAME bump without versioned symbols in either the old or new
> library, and without an ack from -release before the upload, is fairly
> harsh on the release team and your fellow developers.
> I'm not sure what to do at this point.
what about uploading a gmp3 source package just building the old runtime
libraries? This way packages can migrate on it's own, because build dependencies
are not checked for migration to testing. If that's not enough, build the -dev
packages from this source too, and require build-depending on the generic
libgmp-dev. Then every package maintainer can convert to the new version on
it's own. See readline for a similar transition, which may take more than one
release cycle (yes, the readline transition has license issues on top).