[sage-devel] Re: recompilation with sage -clone
leif wrote: leif wrote: Looks like reinstalling the Sage library spkg would meanwhile also trigger the rebuild of *all* extension modules; not sure since when though... (seen with Sage 5.10.beta5 at least) Yep. This is because #14570, merged into Sage 5.10.beta4, completely breaks sage-sync-build, which is also used in the Sage library's spkg-install. (sage -sync-build now removes all Cython-generated files, hence *all* have to get rebuilt afterwards.) IMHO both issues (complete rebuild upon cloning and after sage-sync-build) should be resolved before the final 5.10 gets out. Note that cythonizing out-of-tree may also break sage-clone (not just sage-sync-build), depending on the (rather random) order in which subdirectories of $SAGE_SRC/build/ are coopied. Cf. http://trac.sagemath.org/sage_trac/ticket/13245#comment:50 ff. -leif -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone
leif wrote: leif wrote: leif wrote: Looks like reinstalling the Sage library spkg would meanwhile also trigger the rebuild of *all* extension modules; not sure since when though... (seen with Sage 5.10.beta5 at least) Yep. This is because #14570, merged into Sage 5.10.beta4, completely breaks sage-sync-build, which is also used in the Sage library's spkg-install. (sage -sync-build now removes all Cython-generated files, hence *all* have to get rebuilt afterwards.) IMHO both issues (complete rebuild upon cloning and after sage-sync-build) should be resolved before the final 5.10 gets out. Note that cythonizing out-of-tree may also break sage-clone (not just sage-sync-build), depending on the (rather random) order in which subdirectories of $SAGE_SRC/build/ are copied. Disabling / postponing cythonizing out-of-tree is now http://trac.sagemath.org/sage_trac/ticket/14721 , a 5.10 blocker ticket; trivial patch, needs review... -leif -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone and after sage -sync-build
leif wrote: leif wrote: 4) Minimally adapt sage-sync-build.py to reflect the (hard-coded) change by #14570. [May be easy as well, but potentially unsafe.] Even that turns out to be non-trivial, AFAICS, i.e., the whole strategy would have to get changed, as opposed to just some file / directory names in a few places. Just for the record: I've meanwhile managed to hack sage-sync-build.py such that it /seems/^TM to work with a Cython build_dir (below $SAGE_SRC/build/) as well, but the patch would need clean-up and thorough testing. And as mentioned, the script needs further, not directly related, fixes anyway, so I'd prefer to postpone changes to sage-sync-build to 5.11. -leif -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone and after sage -sync-build
Hey, Could we at least get the hacked sage-sync-build into 5.10? The combinat queue has a few patch which create new cython files that replace python files (i.e. cythonizes them) which requires the sync-build to be run when people are moving ahead of those patches. Rebuilding all cython files in sage seems like a high price to me for something we can put a quick patch (no pun intended) over. Thanks, Travis On Sunday, June 2, 2013 3:32:12 AM UTC-7, leif wrote: leif wrote: leif wrote: 4) Minimally adapt sage-sync-build.py to reflect the (hard-coded) change by #14570. [May be easy as well, but potentially unsafe.] Even that turns out to be non-trivial, AFAICS, i.e., the whole strategy would have to get changed, as opposed to just some file / directory names in a few places. Just for the record: I've meanwhile managed to hack sage-sync-build.py such that it /seems/^TM to work with a Cython build_dir (below $SAGE_SRC/build/) as well, but the patch would need clean-up and thorough testing. And as mentioned, the script needs further, not directly related, fixes anyway, so I'd prefer to postpone changes to sage-sync-build to 5.11. -leif -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone
John H Palmieri wrote: On Thursday, May 30, 2013 9:57:07 PM UTC-7, leif wrote: leif wrote: Looks like reinstalling the Sage library spkg would meanwhile also trigger the rebuild of *all* extension modules; not sure since when though... (seen with Sage 5.10.beta5 at least) Yep. This is because #14570, merged into Sage 5.10.beta4, completely breaks sage-sync-build, which is also used in the Sage library's spkg-install. (sage -sync-build now removes all Cython-generated files, hence *all* have to get rebuilt afterwards.) IMHO both issues (complete rebuild upon cloning and after sage-sync-build) should be resolved before the final 5.10 gets out. The rebuild is dealt with at #13245 (needs partial review). I'll revisit #13245 now that you've changed the ticket's title... ;-) [Avoiding also the rebuild of the fast_callable interpreters should be feasable by just copying a few more files in sage-clone (I /think/ I already did that), but I agree that's a minor issue, at least compared to rebuilding the whole Sage library.] Is there a ticket for the sage-sync-build problem? Not yet. There are a couple of ways to (quickly or temporarily) fix this: 1) Unmerge #14570, which doesn't fix a bug, but only introduces a new (btw. non-configurable) feature. [A bit odd, especially since #14570 also patches the Cython spkg.] 2) Revert just #14570's change to setup.py, a one-liner (removing the build_dir parameter to cythonize()). [Better than 1), although effectively the same -- feature vanishes until we reenable it on some follow-up ticket.] 3) Make the behaviour configurable (defaulting to the previous; little change to setup.py). No breakage to ordinary users; people using the new feature are expected to know what they're doing, i.e., at least sage-sync-build [still] won't work for them. [IMHO preferable for a temporary / quick fix; easy and safe.] 4) Minimally adapt sage-sync-build.py to reflect the (hard-coded) change by #14570. [May be easy as well, but potentially unsafe.] I'd personally prefer 2) or 3) for now, postponing a proper solution to a follow-up ticket which presumably won't make it into 5.10 (but hopefully 5.11.*), as sage-sync-build needs some rework* anyway, the Cython build directory (kind of VPATH) should IMHO be configurable, and #14570 might cause further issues we haven't noticed yet. -leif __ * the usual can of worms -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone
leif wrote: John H Palmieri wrote: Is there a ticket for the sage-sync-build problem? Not yet. There are a couple of ways to (quickly or temporarily) fix this: 1) Unmerge #14570, which doesn't fix a bug, but only introduces a new (btw. non-configurable) feature. [A bit odd, especially since #14570 also patches the Cython spkg.] 2) Revert just #14570's change to setup.py, a one-liner (removing the build_dir parameter to cythonize()). [Better than 1), although effectively the same -- feature vanishes until we reenable it on some follow-up ticket.] diff --git a/setup.py b/setup.py --- a/setup.py +++ b/setup.py @@ -532,7 +532,7 @@ ext_modules = cythonize( ext_modules, nthreads = int(os.environ.get('SAGE_NUM_THREADS', 0)), -build_dir = 'build/cythonized', +# build_dir = 'build/cythonized', force=force) open(version_file, 'w').write(Cython.__version__) works for me (with Sage 5.10.rc0). 3) Make the behaviour configurable (defaulting to the previous; little change to setup.py). No breakage to ordinary users; people using the new feature are expected to know what they're doing, i.e., at least sage-sync-build [still] won't work for them. [IMHO preferable for a temporary / quick fix; easy and safe.] Similar to the above change, but with (say) diff --git a/setup.py b/setup.py --- a/setup.py +++ b/setup.py @@ -532,7 +532,7 @@ ext_modules = cythonize( ext_modules, nthreads = int(os.environ.get('SAGE_NUM_THREADS', 0)), -build_dir = 'build/cythonized', +build_dir = 'build/cythonized' if os.environ.get(BREAK_SAGE_SYNC_BUILD, no)==yes else None, force=force) open(version_file, 'w').write(Cython.__version__) which *optionally* enables building extension modules out of tree (they're in fact in the same tree, but one level deeper, still hard-coded location) at the expense of breaking (at least) sage-sync-build. (Works for me, too.) -leif 4) Minimally adapt sage-sync-build.py to reflect the (hard-coded) change by #14570. [May be easy as well, but potentially unsafe.] I'd personally prefer 2) or 3) for now, postponing a proper solution to a follow-up ticket which presumably won't make it into 5.10 (but hopefully 5.11.*), as sage-sync-build needs some rework* anyway, the Cython build directory (kind of VPATH) should IMHO be configurable, and #14570 might cause further issues we haven't noticed yet. -leif __ * the usual can of worms -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone and after sage -sync-build
leif wrote: leif wrote: John H Palmieri wrote: Is there a ticket for the sage-sync-build problem? Not yet. There are a couple of ways to (quickly or temporarily) fix this: 1) Unmerge #14570, which doesn't fix a bug, but only introduces a new (btw. non-configurable) feature. [A bit odd, especially since #14570 also patches the Cython spkg.] 2) Revert just #14570's change to setup.py, a one-liner (removing the build_dir parameter to cythonize()). [Better than 1), although effectively the same -- feature vanishes until we reenable it on some follow-up ticket.] diff --git a/setup.py b/setup.py --- a/setup.py +++ b/setup.py @@ -532,7 +532,7 @@ ext_modules = cythonize( ext_modules, nthreads = int(os.environ.get('SAGE_NUM_THREADS', 0)), -build_dir = 'build/cythonized', +# build_dir = 'build/cythonized', force=force) open(version_file, 'w').write(Cython.__version__) works for me (with Sage 5.10.rc0). 3) Make the behaviour configurable (defaulting to the previous; little change to setup.py). No breakage to ordinary users; people using the new feature are expected to know what they're doing, i.e., at least sage-sync-build [still] won't work for them. [IMHO preferable for a temporary / quick fix; easy and safe.] Similar to the above change, but with (say) diff --git a/setup.py b/setup.py --- a/setup.py +++ b/setup.py @@ -532,7 +532,7 @@ ext_modules = cythonize( ext_modules, nthreads = int(os.environ.get('SAGE_NUM_THREADS', 0)), -build_dir = 'build/cythonized', +build_dir = 'build/cythonized' if os.environ.get(BREAK_SAGE_SYNC_BUILD, no)==yes else None, force=force) open(version_file, 'w').write(Cython.__version__) which *optionally* enables building extension modules out of tree (they're in fact in the same tree, but one level deeper, still hard-coded location) at the expense of breaking (at least) sage-sync-build. (Works for me, too.) -leif 4) Minimally adapt sage-sync-build.py to reflect the (hard-coded) change by #14570. [May be easy as well, but potentially unsafe.] Even that turns out to be non-trivial, AFAICS, i.e., the whole strategy would have to get changed, as opposed to just some file / directory names in a few places. -leif I'd personally prefer 2) or 3) for now, postponing a proper solution to a follow-up ticket which presumably won't make it into 5.10 (but hopefully 5.11.*), as sage-sync-build needs some rework* anyway, the Cython build directory (kind of VPATH) should IMHO be configurable, and #14570 might cause further issues we haven't noticed yet. -leif __ * the usual can of worms -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone
On Thursday, May 30, 2013 9:57:07 PM UTC-7, leif wrote: leif wrote: Looks like reinstalling the Sage library spkg would meanwhile also trigger the rebuild of *all* extension modules; not sure since when though... (seen with Sage 5.10.beta5 at least) Yep. This is because #14570, merged into Sage 5.10.beta4, completely breaks sage-sync-build, which is also used in the Sage library's spkg-install. (sage -sync-build now removes all Cython-generated files, hence *all* have to get rebuilt afterwards.) IMHO both issues (complete rebuild upon cloning and after sage-sync-build) should be resolved before the final 5.10 gets out. The rebuild is dealt with at #13245 (needs partial review). Is there a ticket for the sage-sync-build problem? -- John -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone
Hey, I noticed that after running sync-build that it was always taking a very long time. I'm moving through a couple of patches which create new cython files and having to watch sage rebuild because I need to run sync-build slows me down significantly. This definitely should be a blocker. Best, Travis On Friday, May 31, 2013 11:57:48 AM UTC-7, John H Palmieri wrote: On Thursday, May 30, 2013 9:57:07 PM UTC-7, leif wrote: leif wrote: Looks like reinstalling the Sage library spkg would meanwhile also trigger the rebuild of *all* extension modules; not sure since when though... (seen with Sage 5.10.beta5 at least) Yep. This is because #14570, merged into Sage 5.10.beta4, completely breaks sage-sync-build, which is also used in the Sage library's spkg-install. (sage -sync-build now removes all Cython-generated files, hence *all* have to get rebuilt afterwards.) IMHO both issues (complete rebuild upon cloning and after sage-sync-build) should be resolved before the final 5.10 gets out. The rebuild is dealt with at #13245 (needs partial review). Is there a ticket for the sage-sync-build problem? -- John -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone
Eric Gourgoulhon wrote: I confirm this behavior on a fresh install of version 5.9 from the sources: sage -clone triggers the recompilation of the Cython sources. In version 5.8, it did not. Don't know the reason for this... Looks like reinstalling the Sage library spkg would meanwhile also trigger the rebuild of *all* extension modules; not sure since when though... (seen with Sage 5.10.beta5 at least) -leif Le vendredi 10 mai 2013 14:53:27 UTC+2, vdelecroix a écrit : Hi, It seems that the behavior of sage -clone changes around the version 5.9. I freshly compiled a version of 5.10.beta2 and when I do a clone it recompiles the wholse sage sources (but hopefully not the auxilliary spkg). The first line in the log of sage -clone which differs from previous version is the following Updating Cython code Compiling sage/algebras/quatalg/quaternion_algebra_element.pyx because it depends on ./sage/structure/sage_object.pxd ... I was not able to find some information elsewhere. Best, Vincent -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone
leif wrote: Looks like reinstalling the Sage library spkg would meanwhile also trigger the rebuild of *all* extension modules; not sure since when though... (seen with Sage 5.10.beta5 at least) Yep. This is because #14570, merged into Sage 5.10.beta4, completely breaks sage-sync-build, which is also used in the Sage library's spkg-install. (sage -sync-build now removes all Cython-generated files, hence *all* have to get rebuilt afterwards.) IMHO both issues (complete rebuild upon cloning and after sage-sync-build) should be resolved before the final 5.10 gets out. -leif -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone
On 2013-05-31, leif not.rea...@online.de wrote: IMHO both issues (complete rebuild upon cloning and after sage-sync-build) should be resolved before the final 5.10 gets out. +1 It should be a blocker, IMHO. Cheers, Simon -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone
For the impatient, the following at least avoids rebuilding of (most of the) Cython-generated files: Thanks leif, it works quite well on sage-5.9. Do you have an idea for avoiding documentation regeneration (see [1]) ? Best, Vincent [1] http://groups.google.com/group/sage-devel/browse_thread/thread/3c0a371b78568d90 -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone
leif wrote: leif wrote: Eric Gourgoulhon wrote: I confirm this behavior on a fresh install of version 5.9 from the sources: sage -clone triggers the recompilation of the Cython sources. In version 5.8, it did not. Don't know the reason for this... For the impatient, the following at least avoids rebuilding of (most of the) Cython-generated files: diff --git a/sage-clone b/sage-clone --- a/sage-clone +++ b/sage-clone @@ -53,6 +53,10 @@ cpdir(os.path.abspath('sage/sage'), os.path.abspath(branch + '/sage')) +if os.path.isfile('sage/.cython_version'): +print Copying over hidden Cython version file... +os.link('sage/.cython_version', branch+'/.cython_version') + def copy_dtree(src_dir, dest_dir): src_root = os.path.abspath(src_dir) dest_root = os.path.abspath(dest_dir) It seems the interpreters for fast_callable get regenerated (and hence also re-cythonized) because sage-clone does not (or no longer) properly copy devel/sage/sage/ext/interpreters/timestamp either: Building interpreters for fast_callable Updating Cython code Building sage/plot/plot3d/parametric_surface.pyx because it depends on sage/ext/interpreters/wrapper_rdf.pxd. Building modified file sage/ext/interpreters/wrapper_rdf.pyx. Building modified file sage/ext/interpreters/wrapper_cdf.pyx. Building modified file sage/ext/interpreters/wrapper_rr.pyx. Building modified file sage/ext/interpreters/wrapper_py.pyx. Building modified file sage/ext/interpreters/wrapper_el.pyx. ... (This is what I meant by avoids rebuilding of /most of/ the Cython-generated code.) -leif -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone
I confirm this behavior on a fresh install of version 5.9 from the sources: sage -clone triggers the recompilation of the Cython sources. In version 5.8, it did not. Don't know the reason for this... Eric. Le vendredi 10 mai 2013 14:53:27 UTC+2, vdelecroix a écrit : Hi, It seems that the behavior of sage -clone changes around the version 5.9. I freshly compiled a version of 5.10.beta2 and when I do a clone it recompiles the wholse sage sources (but hopefully not the auxilliary spkg). The first line in the log of sage -clone which differs from previous version is the following Updating Cython code Compiling sage/algebras/quatalg/quaternion_algebra_element.pyx because it depends on ./sage/structure/sage_object.pxd ... I was not able to find some information elsewhere. Best, Vincent -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone
Eric Gourgoulhon wrote: I confirm this behavior on a fresh install of version 5.9 from the sources: sage -clone triggers the recompilation of the Cython sources. In version 5.8, it did not. Don't know the reason for this... FWIW, it did still work in 5.9.beta2. Probably #13031 [1] (merged into 5.9.beta4), or one of its follow-ups(?) is to blame... -leif [1] http://trac.sagemath.org/sage_trac/ticket/13031 Le vendredi 10 mai 2013 14:53:27 UTC+2, vdelecroix a écrit : Hi, It seems that the behavior of sage -clone changes around the version 5.9. I freshly compiled a version of 5.10.beta2 and when I do a clone it recompiles the wholse sage sources (but hopefully not the auxilliary spkg). The first line in the log of sage -clone which differs from previous version is the following Updating Cython code Compiling sage/algebras/quatalg/quaternion_algebra_element.pyx because it depends on ./sage/structure/sage_object.pxd ... I was not able to find some information elsewhere. Best, Vincent -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: recompilation with sage -clone
leif wrote: Eric Gourgoulhon wrote: I confirm this behavior on a fresh install of version 5.9 from the sources: sage -clone triggers the recompilation of the Cython sources. In version 5.8, it did not. Don't know the reason for this... FWIW, it did still work in 5.9.beta2. Probably #13031 [1] (merged into 5.9.beta4), or one of its follow-ups(?) is to blame... For the impatient, the following at least avoids rebuilding of (most of the) Cython-generated files: diff --git a/sage-clone b/sage-clone --- a/sage-clone +++ b/sage-clone @@ -53,6 +53,10 @@ cpdir(os.path.abspath('sage/sage'), os.path.abspath(branch + '/sage')) +if os.path.isfile('sage/.cython_version'): +print Copying over hidden Cython version file... +os.link('sage/.cython_version', branch+'/.cython_version') + def copy_dtree(src_dir, dest_dir): src_root = os.path.abspath(src_dir) dest_root = os.path.abspath(dest_dir) (To be applied to the Sage scripts repo, in $SAGE_ROOT/local/bin/.) It seems also all(?) Python files get re-(byte-)compiled; don't know yet why... (but the latter doesn't take as long as re-cythonizing* does) -leif __ * What an ill-chosen name! [1] http://trac.sagemath.org/sage_trac/ticket/13031 Le vendredi 10 mai 2013 14:53:27 UTC+2, vdelecroix a écrit : Hi, It seems that the behavior of sage -clone changes around the version 5.9. I freshly compiled a version of 5.10.beta2 and when I do a clone it recompiles the wholse sage sources (but hopefully not the auxilliary spkg). The first line in the log of sage -clone which differs from previous version is the following Updating Cython code Compiling sage/algebras/quatalg/quaternion_algebra_element.pyx because it depends on ./sage/structure/sage_object.pxd ... I was not able to find some information elsewhere. Best, Vincent -- () The ASCII Ribbon Campaign /\ Help Cure HTML E-Mail -- You received this message because you are subscribed to the Google Groups sage-devel group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en. For more options, visit https://groups.google.com/groups/opt_out.