D'oh: keshav@sage ~ $ grep -r '/python' git-1.7.9.4.p1 git-1.7.9.4.p1/src/Makefile: PYTHON_PATH = /usr/bin/python git-1.7.9.4.p1/src/Makefile: PYTHON_PATH = /usr/local/bin/python git-1.7.9.4.p1/src/git_remote_helpers/Makefile: PYTHON_PATH = /usr/local/bin/python git-1.7.9.4.p1/src/git_remote_helpers/Makefile: PYTHON_PATH = /usr/bin/python git-1.7.9.4.p1/src/git_remote_helpers/Makefile: print 'lib/python%i.%i/site-packages' % sys.version_info[:2]")
So we'll need to patch the makefile, I guess. Let's pile this onto #12707... -Keshav On Friday, April 20, 2012 12:47:45 PM UTC+8, Keshav Kini wrote: > > Hmm... > > keshav@sage ~ $ mkdir bin > keshav@sage ~ $ cat - > bin/python > #!/bin/sh > echo "You are likely to be eaten by a grue" > keshav@sage ~ $ chmod a+x bin/python > keshav@sage ~ $ export PATH="/home/keshav/bin:$PATH" > keshav@sage ~ $ ./sage/sage -i ./git-1.7.9.4.p1.spkg | grep grue > keshav@sage ~ $ > > So as far as I can see the only way it could be calling the system python > would be if something in git has a hashbang hard-coded to /usr/bin/python > or something. But all these look in order... > > keshav@sage ~ $ tar xf git-1.7.9.4.p1.spkg > keshav@sage ~ $ grep -r '^#!.*python' git-1.7.9.4.p1 > git-1.7.9.4.p1/src/contrib/p4import/git-p4import.py:#!/usr/bin/env python > git-1.7.9.4.p1/src/contrib/fast-import/git-p4:#!/usr/bin/env python > git-1.7.9.4.p1/src/contrib/fast-import/import-zips.py:#!/usr/bin/env python > git-1.7.9.4.p1/src/contrib/gitview/gitview:#! /usr/bin/env python > git-1.7.9.4.p1/src/contrib/ciabot/ciabot.py:#!/usr/bin/env python > git-1.7.9.4.p1/src/contrib/hg-to-git/hg-to-git.py:#!/usr/bin/env python > git-1.7.9.4.p1/src/git_remote_helpers/__init__.py:#!/usr/bin/env python > git-1.7.9.4.p1/src/git_remote_helpers/util.py:#!/usr/bin/env python > git-1.7.9.4.p1/src/git_remote_helpers/git/git.py:#!/usr/bin/env python > git-1.7.9.4.p1/src/git_remote_helpers/setup.py:#!/usr/bin/env python > git-1.7.9.4.p1/src/git-remote-testgit.py:#!/usr/bin/env python > > Any ideas? > > -Keshav > > ---- > Join us in #sagemath on irc.freenode.net ! > > On Friday, April 20, 2012 12:23:20 PM UTC+8, John H Palmieri wrote: >> >> >> >> On Thursday, April 19, 2012 8:25:15 PM UTC-7, Rajeev wrote: >>> >>> I don't think that should happen because I'm just giving the following >>> command on terminal - >>> >>> sage -i git-1.7.9.4 >>> >>> I am using sage-4.8. I get the same error on two computers (one >>> running SuseLinux and another running Debian). >>> >> I see the same error when running "sage -i git" on >> sage.math.washington.edu (using sage-5.0.beta13). There are also >> suspicious looking warnings like >> >> /scratch/palmieri/sage-5.0.beta13/local/lib/python/linecache.py:127: >> Warning: 'with' will become a reserved keyword in Python 2.6 >> >> So is the installation somehow using the system's Python (which is >> version 2.5.2) rather than Sage's? Anyway, if the spkg doesn't install >> correctly on sage.math, something is seriously wrong with the spkg. >> >> By the way, I see a different error when running "sage -i git" on OS X >> Lion: >> >> AR libgit.a >> /usr/bin/ranlib: file: libgit.a(gettext.o) has no symbols >> LINK git-daemon >> LINK git-fast-import >> ld: warning: ignoring file /opt/local/lib/libz.dylib, file was built for >> unsupported file format which is not the architecture being linked (x86_64) >> ld: warning: ignoring file /sw/lib/libiconv.dylib, file was built for >> unsupported file format which is not the architecture being linked (x86_64) >> ld: warning: ignoring file /opt/local/lib/libcrypto.dylib, file was built >> for unsupported file format which is not the architecture being linked >> (x86_64) >> Undefined symbols for architecture x86_64: >> "_SHA1_Init", referenced from: >> _store_object in fast-import.o >> _stream_blob in fast-import.o >> _sha1fd_throughput in libgit.a(csum-file.o) >> _write_sha1_file_prepare in libgit.a(sha1_file.o) >> _write_loose_object in libgit.a(sha1_file.o) >> _read_index_from in libgit.a(read-cache.o) >> _write_index in libgit.a(read-cache.o) >> ... >> >> and then it goes on like that for a while. >> >> -- >> John >> >> > On Friday, April 20, 2012 12:23:20 PM UTC+8, John H Palmieri wrote: >> >> >> >> On Thursday, April 19, 2012 8:25:15 PM UTC-7, Rajeev wrote: >>> >>> I don't think that should happen because I'm just giving the following >>> command on terminal - >>> >>> sage -i git-1.7.9.4 >>> >>> I am using sage-4.8. I get the same error on two computers (one >>> running SuseLinux and another running Debian). >>> >> I see the same error when running "sage -i git" on >> sage.math.washington.edu (using sage-5.0.beta13). There are also >> suspicious looking warnings like >> >> /scratch/palmieri/sage-5.0.beta13/local/lib/python/linecache.py:127: >> Warning: 'with' will become a reserved keyword in Python 2.6 >> >> So is the installation somehow using the system's Python (which is >> version 2.5.2) rather than Sage's? Anyway, if the spkg doesn't install >> correctly on sage.math, something is seriously wrong with the spkg. >> >> By the way, I see a different error when running "sage -i git" on OS X >> Lion: >> >> AR libgit.a >> /usr/bin/ranlib: file: libgit.a(gettext.o) has no symbols >> LINK git-daemon >> LINK git-fast-import >> ld: warning: ignoring file /opt/local/lib/libz.dylib, file was built for >> unsupported file format which is not the architecture being linked (x86_64) >> ld: warning: ignoring file /sw/lib/libiconv.dylib, file was built for >> unsupported file format which is not the architecture being linked (x86_64) >> ld: warning: ignoring file /opt/local/lib/libcrypto.dylib, file was built >> for unsupported file format which is not the architecture being linked >> (x86_64) >> Undefined symbols for architecture x86_64: >> "_SHA1_Init", referenced from: >> _store_object in fast-import.o >> _stream_blob in fast-import.o >> _sha1fd_throughput in libgit.a(csum-file.o) >> _write_sha1_file_prepare in libgit.a(sha1_file.o) >> _write_loose_object in libgit.a(sha1_file.o) >> _read_index_from in libgit.a(read-cache.o) >> _write_index in libgit.a(read-cache.o) >> ... >> >> and then it goes on like that for a while. >> >> -- >> John >> >> > On Friday, April 20, 2012 12:23:20 PM UTC+8, John H Palmieri wrote: >> >> >> >> On Thursday, April 19, 2012 8:25:15 PM UTC-7, Rajeev wrote: >>> >>> I don't think that should happen because I'm just giving the following >>> command on terminal - >>> >>> sage -i git-1.7.9.4 >>> >>> I am using sage-4.8. I get the same error on two computers (one >>> running SuseLinux and another running Debian). >>> >> I see the same error when running "sage -i git" on >> sage.math.washington.edu (using sage-5.0.beta13). There are also >> suspicious looking warnings like >> >> /scratch/palmieri/sage-5.0.beta13/local/lib/python/linecache.py:127: >> Warning: 'with' will become a reserved keyword in Python 2.6 >> >> So is the installation somehow using the system's Python (which is >> version 2.5.2) rather than Sage's? Anyway, if the spkg doesn't install >> correctly on sage.math, something is seriously wrong with the spkg. >> >> By the way, I see a different error when running "sage -i git" on OS X >> Lion: >> >> AR libgit.a >> /usr/bin/ranlib: file: libgit.a(gettext.o) has no symbols >> LINK git-daemon >> LINK git-fast-import >> ld: warning: ignoring file /opt/local/lib/libz.dylib, file was built for >> unsupported file format which is not the architecture being linked (x86_64) >> ld: warning: ignoring file /sw/lib/libiconv.dylib, file was built for >> unsupported file format which is not the architecture being linked (x86_64) >> ld: warning: ignoring file /opt/local/lib/libcrypto.dylib, file was built >> for unsupported file format which is not the architecture being linked >> (x86_64) >> Undefined symbols for architecture x86_64: >> "_SHA1_Init", referenced from: >> _store_object in fast-import.o >> _stream_blob in fast-import.o >> _sha1fd_throughput in libgit.a(csum-file.o) >> _write_sha1_file_prepare in libgit.a(sha1_file.o) >> _write_loose_object in libgit.a(sha1_file.o) >> _read_index_from in libgit.a(read-cache.o) >> _write_index in libgit.a(read-cache.o) >> ... >> >> and then it goes on like that for a while. >> >> -- >> John >> >> > On Friday, April 20, 2012 12:23:20 PM UTC+8, John H Palmieri wrote: >> >> >> >> On Thursday, April 19, 2012 8:25:15 PM UTC-7, Rajeev wrote: >>> >>> I don't think that should happen because I'm just giving the following >>> command on terminal - >>> >>> sage -i git-1.7.9.4 >>> >>> I am using sage-4.8. I get the same error on two computers (one >>> running SuseLinux and another running Debian). >>> >> I see the same error when running "sage -i git" on >> sage.math.washington.edu (using sage-5.0.beta13). There are also >> suspicious looking warnings like >> >> /scratch/palmieri/sage-5.0.beta13/local/lib/python/linecache.py:127: >> Warning: 'with' will become a reserved keyword in Python 2.6 >> >> So is the installation somehow using the system's Python (which is >> version 2.5.2) rather than Sage's? Anyway, if the spkg doesn't install >> correctly on sage.math, something is seriously wrong with the spkg. >> >> By the way, I see a different error when running "sage -i git" on OS X >> Lion: >> >> AR libgit.a >> /usr/bin/ranlib: file: libgit.a(gettext.o) has no symbols >> LINK git-daemon >> LINK git-fast-import >> ld: warning: ignoring file /opt/local/lib/libz.dylib, file was built for >> unsupported file format which is not the architecture being linked (x86_64) >> ld: warning: ignoring file /sw/lib/libiconv.dylib, file was built for >> unsupported file format which is not the architecture being linked (x86_64) >> ld: warning: ignoring file /opt/local/lib/libcrypto.dylib, file was built >> for unsupported file format which is not the architecture being linked >> (x86_64) >> Undefined symbols for architecture x86_64: >> "_SHA1_Init", referenced from: >> _store_object in fast-import.o >> _stream_blob in fast-import.o >> _sha1fd_throughput in libgit.a(csum-file.o) >> _write_sha1_file_prepare in libgit.a(sha1_file.o) >> _write_loose_object in libgit.a(sha1_file.o) >> _read_index_from in libgit.a(read-cache.o) >> _write_index in libgit.a(read-cache.o) >> ... >> >> and then it goes on like that for a while. >> >> -- >> John >> >> > On Friday, April 20, 2012 12:23:20 PM UTC+8, John H Palmieri wrote: >> >> >> >> On Thursday, April 19, 2012 8:25:15 PM UTC-7, Rajeev wrote: >>> >>> I don't think that should happen because I'm just giving the following >>> command on terminal - >>> >>> sage -i git-1.7.9.4 >>> >>> I am using sage-4.8. I get the same error on two computers (one >>> running SuseLinux and another running Debian). >>> >> I see the same error when running "sage -i git" on >> sage.math.washington.edu (using sage-5.0.beta13). There are also >> suspicious looking warnings like >> >> /scratch/palmieri/sage-5.0.beta13/local/lib/python/linecache.py:127: >> Warning: 'with' will become a reserved keyword in Python 2.6 >> >> So is the installation somehow using the system's Python (which is >> version 2.5.2) rather than Sage's? Anyway, if the spkg doesn't install >> correctly on sage.math, something is seriously wrong with the spkg. >> >> By the way, I see a different error when running "sage -i git" on OS X >> Lion: >> >> AR libgit.a >> /usr/bin/ranlib: file: libgit.a(gettext.o) has no symbols >> LINK git-daemon >> LINK git-fast-import >> ld: warning: ignoring file /opt/local/lib/libz.dylib, file was built for >> unsupported file format which is not the architecture being linked (x86_64) >> ld: warning: ignoring file /sw/lib/libiconv.dylib, file was built for >> unsupported file format which is not the architecture being linked (x86_64) >> ld: warning: ignoring file /opt/local/lib/libcrypto.dylib, file was built >> for unsupported file format which is not the architecture being linked >> (x86_64) >> Undefined symbols for architecture x86_64: >> "_SHA1_Init", referenced from: >> _store_object in fast-import.o >> _stream_blob in fast-import.o >> _sha1fd_throughput in libgit.a(csum-file.o) >> _write_sha1_file_prepare in libgit.a(sha1_file.o) >> _write_loose_object in libgit.a(sha1_file.o) >> _read_index_from in libgit.a(read-cache.o) >> _write_index in libgit.a(read-cache.o) >> ... >> >> and then it goes on like that for a while. >> >> -- >> John >> >> -- 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
