It seems very fragile to me to reply on a particular syntax in the
operating system's way of displaying a symbolic link to find out the
name of the current branch. Why not store the branch's name in a file
and read that? Then it would not matter whether it was a link or not.
John
On 23 August 2010 17:35, Ryan Hinton <iob...@email.com> wrote:
> This seems to work. Thanks!
>
> On Aug 23, 10:17 am, Ryan Hinton <iob...@email.com> wrote:
>> You're right, the symbolic link had been dereferenced, so devel/sage
>> and devel/sage-main were both normal directories. In my sandbox, it
>> worked to remove devel/sage and recreate it as a link to sage-main.
>> We'll try it on the real install soon.
>>
>> Thanks!
>>
>> - Ryan
>>
>> On Aug 21, 2:34 am, Mitesh Patel <qed...@gmail.com> wrote:
>>
>> > On 08/21/2010 02:12 AM, Jason Grout wrote:
>>
>> > > On 8/20/10 2:44 PM, Ryan Hinton wrote:
>> > >> I have a build of Sage at school that I occasionally update when a bug
>> > >> or upgrade affects me. I'm the only one who uses it, but it's
>> > >> installed in a public location (i.e. not writable by me) in case
>> > >> someone else *might* use it. And because I don't have enough disk
>> > >> space in my personal quota. :-)
>>
>> > >> So I went through the motions again: download, build, test, notify the
>> > >> sysadmin that a new build is ready to be copied over. But this time
>> > >> we're having trouble with Sage fixing the install paths. There are
>> > >> multiple levels of symbolic links in the path, and some components
>> > >> (e.g. Zope) are picking up the absolute path (w/o symlinks), while
>> > >> others are picking up the cd'ed path where "./sage" is being run (w/
>> > >> symlinks). The symptom is that we get
>>
>> > >> /share/apps/contrib/sage-4.5.2/local/lib/python2.6/site-packages/sage/
>> > >> misc/misc.py in branch_current_hg()
>> > >> 1877 i = s.rfind('->')
>> > >> 1878 if i == -1:
>> > >> -> 1879 raise RuntimeError, "unable to determine branch?!"
>> > >> 1880 s = s[i+2:]
>> > >> 1881 i = s.find('-')
>>
>> > >> RuntimeError: unable to determine branch?!
>> > >> Error importing ipy_profile_sage - perhaps you should run %upgrade?
>> > >> WARNING: Loading of ipy_profile_sage failed.
>>
>> > >> while Sage is trying to fix the paths (next output after the "Do not
>> > >> interrupt this" message).
>> > >> Any suggestions?
>>
>> > Could you tell us the result of
>>
>> > cd SAGE_ROOT
>> > ls -l devel/
>>
>> > ? For example, I get
>>
>> > total 8
>> > drwxr-xr-x 2 mpatel mpatel 4096 2010-08-14 19:31 old
>> > lrwxrwxrwx 1 mpatel mpatel 9 2010-08-14 19:33 sage -> sage-main
>> > drwxr-xr-x 7 mpatel mpatel 4096 2010-08-19 18:25 sage-main
>>
>> > The branch_current_hg function looks for the symbolic link, which may be
>> > missing because sage is an actual directory.
>>
>> > Perhaps the copy command didn't preserve symlinks? What happens if you use
>>
>> > cp -dR OLD_SAGE_ROOT NEW_SAGE_ROOT
>>
>> > ? The '-d' flag tells cp not to dereference links in the source and to
>> > preserve them in the destination.
>>
>> > On 08/21/2010 02:12 AM, Jason Grout wrote:
>>
>> > > This probably does not directly address your problem, but after looking
>> > > at the code in sage-location (which is the code that fixes paths when
>> > > Sage is moved), and looking at the things that are not fixed (e.g., in
>> > > $SAGE_ROOT/local/lib/pkgconfig), I decided that I wouldn't compile and
>> > > then move Sage any more. A patch for the problems I found in pkgconfig
>> > > is up at
>>
>> > >http://trac.sagemath.org/sage_trac/ticket/9210
>>
>> > > which involves a pretty substantial restructuring of sage-location. I
>> > > think the patch is still ready for review. I don't know if it will fix
>> > > your problem.
>
> --
> To post to this group, send an email to sage-devel@googlegroups.com
> To unsubscribe from this group, send an email to
> sage-devel+unsubscr...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/sage-devel
> URL: http://www.sagemath.org
>
--
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to
sage-devel+unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org
