On Feb 21, 3:45 pm, John Cremona <john.crem...@gmail.com> wrote:
> Thanks for the diagnosis and suggestions. Evidently I did the upgrade
> from a clone, one I made yesterday to review one of those patches....
>
> Suggestion: since the upgrade already gives a prompt and will not
> continue unless you reply "yes", why not add to that prompt a line
> saying don't go any further unless you are not in a clone. Even
> better, detect whether or not you are in a clone and quite if you are
> (if that is possible).
>
> I'm now doing an upgrade from my rc0 build, not from any clone, and
> we'll see what happens!
>
The upgrade failed in exactly the same way as above. To recap: I had
a working successful build of sage-3.3.rc0 and was in the main branch
(so starting sage did not give me a message about bing in a different
mercurial branch). I typed "sage -upgrade" followed by "y" at the
prpmpt, and went to work on something else. When I came back it had
finished, so I typed "sage" and got the error already reported: the
banner still says rc0 and there's an error on startup, saying
ImportError: libpynac-0.1.so.2: cannot open shared object file: No
such file or directory
Error importing ipy_profile_sage - perhaps you should run %upgrade?
WARNING: Loading of ipy_profile_sage failed.
I particularly like the suggestion that my problem is to upgrade
(which does nothing, of course).
I am going to need a lot of persuading before trying an upgrade again,
which is a pity since I'll only have time to download and build sage
while at home for a while, and my internet connection is not so fast
in rural Warwickshire (especially when my daughter is watching TV on
her laptop!) so downloading the whole distribution takes a couple of
hours.
John
> John
>
> On Feb 21, 2:49 pm, mabshoff <mabsh...@googlemail.com> wrote:
>
> > On Feb 21, 6:36 am, Simon King <k...@mathematik.uni-jena.de> wrote:
>
> > > On 21 Feb., 15:27, mabshoff <mabsh...@googlemail.com> wrote:
>
> > > > > rm -rf devel/sage
>
> > > > oops:
>
> > > > rm -rf devel/sage*
>
> > > > Note that this will nuke all your changes, but that should be fairly
> > > > obvious :)
>
> > > > > ./sage -f spkt/standard/sage-x.y.z.spkg
>
> > > And I guess in the previous line it is spkg/standard/sage-x.y.z.spkg,
> > > not spkt/...
>
> > Yep, I saw that too late after I send the other correction and thought
> > it was so obvious that I did not need to send another email to correct
> > this. Well, it too two now ;)
>
> > > Cheers,
> > > Simon
>
> > Let me drive home this point again: Aside from a very few cases where
> > a dependency problem cause build failures (i.e some spkg with
> > dependencies had unbumped dependencies or I did something else stupid)
> > I had never fail an upgrade for me from a clean tree to a clean tree.
> > If you have missing symbols like in this case that John describes the
> > root cause is nearly always a screwed up merge that can most often be
> > corrected my merging multiple heads and doing a Sage library build via
> > "sage -b".
>
> > To avoid failed merges don't develop in the tree you upgrade in, use
> > either a completely separate build tree to develop for example by
> > building your vanilla tree, then -bdist and use that tree in another
> > place to do work or use clones of the Sage library to do development.
> > Once you want to upgrade, switch back to main, upgrade, clone again
> > and import your changes from the old clone for development into the
> > newly cloned upgraded vanilla tree. Anything else will likely cause
> > pain and suffering unless you have excellent hg Merge-Fu.
>
> > Cheers,
>
> > Michael
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to
sage-devel-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
