Re: [sage-devel] OS X Mavericks
I've opened #15433 to trac porting to 10.9. On Sat, Nov 16, 2013 at 10:18 PM, R. Andrew Ohana wrote: > Ok, I've successfully built a fully working copy of sage on 10.9 (i.e. > passes all > doctests with 3 exceptions, see below). > > polybori: I stopped experiencing the issue others are having (and still > have not > been able to figure out what resolved it for me) > > scipy: This seems like more or less the same issue that we had with 10.8 > -- see > #13309. An easy work around until scipy is updated is to add > -D__ACCELERATE__ to the CPPFLAGS. > > r: links against CoreData, which in turn links against sqlite. > Unfortunately, it > seems like CoreData now requires the extra modules Apple has included in > the > system copy of sqlite. As I see it we have 2 options: > 1) include these extra modules to support CoreData > 2) switch to using the system sqlite on osx > I would vote for the 2nd, as I think it would reduce future headache (and > it isn't > crucial for our purposes to have the most recent version of sqlite). Doing > this > currently produces 3 doctests errors that explicitly test our copy of > sqlite > > sage library: the system cblas won't pretend it is atlas anymore, so set > BLAS2 to 'cblas' > > gcc: at least for me, I was finding that the resulting libstdc++ was > missing some > symbols. I haven't looked into this yet (I instead used homebrew's gcc). > > -- > Andrew > -- Andrew -- 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. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] patchbot problem?
I noticed a failed doctest report by a patchbot: http://patchbot.sagemath.org/log/15432/Ubuntu/12.10/x86_64/3.5.0-17-generic/pi/2013-11-17%2018:44:09%20-0800 which I could reproduce on my own computer with 5.13b2 by pulling the branch and doing sage -b. However, after issuing `make` the doctest succeeded (well, it timed out, but sage -t --long succeeded). Does that indicate an error in dependency tracking in sage -b? How do you know if running sage -b is enough or when a full make is required? -- 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. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: [sage-git] cython cache and speeding up rebuild
I just gave #15430 a try and with warm gcc/cython caches I can do "sage -ba" in 15 seconds... nice! -- 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. For more options, visit https://groups.google.com/groups/opt_out.
[sage-devel] Re: [sage-git] cython cache and speeding up rebuild
On Sunday, November 17, 2013 1:08:52 AM UTC-5, robertwb wrote: > > Are you thinking that ccache and cycache won't be sufficiently fast? > I'm not concerned with rebuilding the Sage library, that is already fast enough that I'm more annoyed by the "do you really want to force build y/n" prompt than by actually doing it. Running "sage -ba" takes 2m 14s on my laptop even without cycache. I'm thinking about package building, especially if you switch between older tickets then you'll be recompiling a good chunk of Sage. The compiler cache is not a magic bullet here, just the configure checks and possible tuning inside the packages can be an unpleasant wait. Of course you can be lucky and the older sage library works with the new packages so you don't have to rebuild packages, but I don't think that'll be always true. > Granted some package changes compile non-C code or do expensive > tuning, but those changes should be relatively rare (and cloning the > repo isn't that bad in the exceptional case). > > - Robert > -- 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. For more options, visit https://groups.google.com/groups/opt_out.
