On Sun, Jan 3, 2010 at 5:29 PM, David Kirkby <david.kir...@onetel.net> wrote: > 2010/1/4 William Stein <wst...@gmail.com>: >> On Sun, Jan 3, 2010 at 5:02 PM, Dr. David Kirkby >> <david.kir...@onetel.net> wrote: >>> You might have seen my previous post about a problem with ATLAS not knowing >>> how >>> many bits my system is. There's a file 'bitwidth.py' in the top level >>> directory >>> of ATLAS. (i.e where spkg-install and SPKG.txt normally go). The contents of >>> that file are: >>> >>> #!/usr/bin/env >>> import ctypes >>> print str(8*ctypes.sizeof(ctypes.c_long)) > >> I think it is *only* used in the notebook, and then oly indirectly by >> some dependency of the notebook. >> >> William > > But it is called and failing in the ATLAS build system. What is it > supposed to be doing there? > > Can it be removed from ATLAS?
Yes. > It would appear it is unstable at best > on many platforms. I assume that bit of code is computing the number > of bits, but can we avoid this? Yes. Hmm. That appears to be a bit of code that Josh Kantor wrote (see http://trac.sagemath.org/sage_trac/ticket/1497). d-69-91-159-111:atlas-3.8.3.p9 wstein$ more bitwidth.py #!/usr/bin/env import ctypes print str(8*ctypes.sizeof(ctypes.c_long)) d-69-91-159-111:atlas-3.8.3.p9 wstein$ hg log bitwidth.py changeset: 6:ffa12e062a44 user: mabsh...@sage.math.washington.edu date: Thu Dec 13 19:57:44 2007 -0800 summary: Check in fixes for 3.8.p5 - see #1497 --- Anyway, the following non-ctypes-code should do the same thing: #!/usr/bin/env python import sys if sys.maxint > 2**32: print "64" else: print "32" -- William -- 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
