Re: Pending soname bumps for m4ri, m4rie, and ntl
2013/5/6 Jerry James <loganjerry(a)gmail.com>:
On Mon, May 6, 2013 at 8:18 AM, Paulo César Pereira de Andrade
<paulo.cesar.pereira.de.andrade(a)gmail.com> wrote:
> I started working on updating to sagemath 5.9 that was just
> released. But if the 5.8 build finished, and most likely did, as
> most of the time is spent building documentation, it should
> be ok to update.
No, the build failed because of the new version of NTL. Sagemath's
ntl_wrap.{h,cpp} assume that many of the fundamental types (ZZ, ZZ_p,
Ok. This will not be one of the easy updates, as sagemath 5.9
besides not changing much "build requires" from sagemath 5.8,
had a lot of refactoring on the key portions of the rpm package
build.
ZZX, etc.) are structs. In NTL 6.0.0, they are classes, not
structs.
I've got a patch to adapt sagemath to this, but didn't have time to
test it over the weekend. I've just started a test build.
If it works for sagemath 5.8, updating for sagemath 5.9 should
be trivial.
If the build succeeds, what would you like me to do? I can send you
the patch, and you can work it into the 5.9 update, or I can do a
build of 5.8 with the patch.
This is fine, feel free to rebuild sagemath 5.8 in rawhide if you
think it is required to avoid breakage for some time/days. If
everything goes fine, I will add your patch to the sagemath 5.9
package.
> The Singular abi/api is somewhat volatile, so, I prefer to
keep
> at the version used by sagemath. Testing/updating after the m4ri
> and m4rie updates should be a better plan.
OK, that makes sense.
Rex, are you okay with me going forward with the rebuilds, or would
you like to handle your own?
--
Jerry James
http://www.jamezone.org/
Paulo