As far as I can tell, 1) this is fixed upstream http://trac.sagemath.org/sage_trac/attachment/ticket/6927/6927-disk_cache-ref.patch (and in the last upstream release)
2) the (upstream) patched file is the only one currently importing processing ( sage/parallel/decorate.py already imports multiprocessing). So once the upstream fix is imported (for convenience, I'm attaching a stripped patch which, if saved to debian/spkg-patches/sage , should apply only the desired change), the dependency can be dropped (and the bug closed). Assuming that's really a drop-in replacement, of course: I tried to build and verified the patch does compile, but space on device ended before finishing compilation, so I couldn't test the package. Pietro
diff -ur sage-3.0.5/sage/parallel/multiprocessing.py sage-3.0.5_p/sage/parallel/multiprocessing.py --- sage-3.0.5/sage/parallel/multiprocessing.py 2010-02-17 15:58:32.000000000 +0100 +++ sage-3.0.5_p/sage/parallel/multiprocessing.py 2010-02-17 16:17:28.000000000 +0100 @@ -11,7 +11,9 @@ # http://www.gnu.org/licenses/ ################################################################################ -from processing import Pool +from __future__ import absolute_import + +import multiprocessing from functools import partial from sage.misc.fpickle import pickle_function, call_pickled_function import ncpus @@ -59,7 +61,7 @@ [(((2,), {}), 4), (((3,), {}), 6)] """ if processes == 0: processes = ncpus.ncpus() - p = Pool(processes) + p = multiprocessing.Pool(processes) fp = pickle_function(f) result = p.imapUnordered(call_pickled_function, [ (fp, t) for t in inputs ])
signature.asc
Description: Questa รจ una parte del messaggio firmata digitalmente
