On Thu, Jul 30, 2009 at 9:13 AM, King, Simon <simon.k...@nuigalway.ie>wrote:
> Hi William, > > -----Original Message----- > From: William Stein [mailto:wst...@gmail.com] > Sent: Thu 30/07/2009 16:31 > To: Simon King > Subject: Re: Cohomology package: Time regression fixed > > |On Wed, Jul 29, 2009 at 8:04 PM, Simon King <simon.k...@uni-jena.de> > wrote: > |... > |> Should I use another _sig_off in the line before raising the error? > | > |Yes, I think so. As a general rule, I try to only use _sig_on/_sig_off > to > |wrap a block of "pure C" code, which would never contain lines like > | "if error_condition: > | raise RuntimeError" > > OK. But I have blocks of "pure C" code in *loops*. Wouldn't it be time > consuming to have _sig_on/_sig_off pairs inside a loop? Yes, it is time consuming. You can do benchmarks. > Or is it something that is only relevant at compile time? It's relevant at runtime. > > > |> If this is the case and if your answer to the > |> first > |> question is "Another _sig_off is not needed", then please replace the > |> 1.0.1.spkg by > |> > |> > http://sage.math.washington.edu/home/SimonKing/Cohomology/p_group_cohomology-1.0.2.spkg > |> in the "experimental" repository. Otherwise I'll do an update tomorrow. > | > |Done and updated. > > Thank you. But -- since you say that I should change the usage of _sig_off > -- I will produce an update version, without a new version number. > OK. > > > |> Can you please again test, how the new package version installs on the > |> different > |> platforms (in particular on the one which had a doc test error, last > time)? > | > |In progress. So far so good! In fact, the install finished on OS X 10.4 PPC and all tests pass! (Total time: 85.98 minutes). > > > Thank you very much! > > I am really positively surprised that the code is portable, thanks to > distutils! Note that the MeatAxe homepage does not explicitely mention > support for OS X even today (they just say it is "developed for the UNIX > operating system. Supported platforms include Linux (x86), SunOS/Solaris > (Sparc), HP/UX, DEC OSF/1 (Alpha), and Windows NT 4.0 (x86, Alpha, PPC)." > > The MeatAxe code that I had in San Diego in fact didn't build on your > Macbook, even after Michael Abshoff trying to guess the right compiler > directives. It's nice to hear something positive about distutils for a change :-) -- William - -- William Stein Associate Professor of Mathematics University of Washington http://wstein.org --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
