just for info: Sage-5.0.1 was the last version of Sage that did build on OS X 10.4 Intel. I don't know (yet) whether OS X 10.5 on Intel is affected. OS X 10.6 (Intel) is *not* affected, and also on OS X 10.4 PPC, Sage still builds fine. But on OS X 10.4 Intel, Sage-5.1, Sage-5.2 and Sage-5.3.rc1 (and thus I do believe Sage-5.3) do not build. I hope to be able to create a ticket for Sage-5.4 soon (with a patch --- it's a one line change in module_list.py, adding 'pari' to the list of libraries for libs.cremona.mat). See also the report:
> just for info: > Sage-5.0.1 was the last version of Sage that did build on OS X 10.4 Intel. > I don't know (yet) whether OS X 10.5 on Intel is affected. OS X 10.6 > (Intel) is *not* affected, and also on OS X 10.4 PPC, Sage still builds > fine. But on OS X 10.4 Intel, Sage-5.1, Sage-5.2 and Sage-5.3.rc1 (and thus > I do believe Sage-5.3) do not build. I hope to be able to create a ticket > for Sage-5.4 soon (with a patch --- it's a one line change in > module_list.py, adding 'pari' to the list of libraries for > libs.cremona.mat). See also the report:
I tried testing an upgrade on this release, from 5.0 built on my
previous Intel Mac to 5.3 upgraded on this Lion Mac. All seems to go
well, but then it wants to access mirrors one more time, claims that
gap and R need to be upgraded, and then (when I say yes), proceeds to
upgrade nearly the entire installation - or at least a large portion
of it - again! It's done this twice now. I'm just going to decline
the R and GAP upgrades next time, but I'm wondering whether there is
some bug introduced in the latest changes to upgrade that would cause
this to happen. I assume the hardware change in the meantime isn't
the issue, and I think 5.0 is where the gcc spkg was introduced,
right?
> I tried testing an upgrade on this release, from 5.0 built on my
> previous Intel Mac to 5.3 upgraded on this Lion Mac. All seems to go
> well, but then it wants to access mirrors one more time, claims that
> gap and R need to be upgraded, and then (when I say yes), proceeds to
> upgrade nearly the entire installation - or at least a large portion
> of it - again! It's done this twice now. I'm just going to decline
> the R and GAP upgrades next time, but I'm wondering whether there is
> some bug introduced in the latest changes to upgrade that would cause
> this to happen. I assume the hardware change in the meantime isn't
> the issue, and I think 5.0 is where the gcc spkg was introduced,
> right?
There are some bugs with upgrading related to GCC, which is #13395
(merged in sage-5.4.beta1).
The fact that everything is built twice is #13415 (currently "new" but
close to "needs review").
> > I tried testing an upgrade on this release, from 5.0 built on my
> > previous Intel Mac to 5.3 upgraded on this Lion Mac. All seems to go
> > well, but then it wants to access mirrors one more time, claims that
> > gap and R need to be upgraded, and then (when I say yes), proceeds to
> > upgrade nearly the entire installation - or at least a large portion
> > of it - again! It's done this twice now. I'm just going to decline
> > the R and GAP upgrades next time, but I'm wondering whether there is
> > some bug introduced in the latest changes to upgrade that would cause
> > this to happen. I assume the hardware change in the meantime isn't
> > the issue, and I think 5.0 is where the gcc spkg was introduced,
> > right?
> There are some bugs with upgrading related to GCC, which is #13395
> (merged in sage-5.4.beta1).
Sorry, I thought that upgrading was now completely fixed, my
apologies.
> The fact that everything is built twice is #13415 (currently "new" but
> close to "needs review").
Good.
I do wonder about the final errors I'm getting. Everything seems fine
except R, which has given me this problem several times now. I think
that GAP may have failed the first time due to a race condition, but
I'm not going to track it down now.
> > I tried testing an upgrade on this release, from 5.0 built on my
> > previous Intel Mac to 5.3 upgraded on this Lion Mac. All seems to go
> > well, but then it wants to access mirrors one more time, claims that
> > gap and R need to be upgraded, and then (when I say yes), proceeds to
> > upgrade nearly the entire installation - or at least a large portion
> > of it - again! It's done this twice now. I'm just going to decline
> > the R and GAP upgrades next time, but I'm wondering whether there is
> > some bug introduced in the latest changes to upgrade that would cause
> > this to happen. I assume the hardware change in the meantime isn't
> > the issue, and I think 5.0 is where the gcc spkg was introduced,
> > right?
> There are some bugs with upgrading related to GCC, which is #13395
> (merged in sage-5.4.beta1).
Can you give me a URL for upgrading to that (unreleased beta) with
sage -upgrade so I can test that here as well? I don't know whether
the http://sage.math.washington.edu/home/release/sage-5.4.beta1/ directory has the right stuff - if it does, let me know.
