I'm just a dumb user ;-) and no Sage system coder, so I don't know which implications my suggestions may have on Sage integrity, but anyway:
On 1 Feb., 16:48, Pat LeSmithe <qed...@gmail.com> wrote: > On 02/01/2010 07:15 AM, xtian wrote: > > On 1 Feb., 11:12, Pat LeSmithe <qed...@gmail.com> wrote: > >> Could you please give a specific example (e.g., file contents, commands > >> to evaluate in sequence) that exhibits the problem? > Thanks! I think this happens because > sage.misc.reset.reset > calls reset_attached(). According to > hg log reset.pyx -p > we added this line at #7514. At this point, I cannot see the benefit I would get as a Sage user from the current effect of reset() on attached files -- I just see that attach() does not work as I would expect anymore. The documentation in reset?? is about vars, globals and interfaces (okay, a call to reset_attached() is mentioned, but not explained). As stated before, I don't have any deeper insight into the Sage sources, and I just wish things to be the way they used to be. :) > I'm not very familiar with Cython, but... Are there any objections to > adding a keyword option attached=True? Then you could put > > reset(attached=False) > > in attached scripts. As a user, I'd prefer it the other way around, i.e. reset() would behave as in Sage 4.3 and before, something like reset(attached=True) could then do what the recent changes to attach() were meant to do (let there be a keyword for the newly introduced feature so "old" code will behave as before). > I've also noticed this two-timing behavior, just as I was experimenting > with your example above, but it's sporadic and I can't reproduce it > reliably. With my 4.3.1 compiled from source, I can reproduce the stuff described above reliably, on i386 as well as on amd64. For more system info, see my original posting. Just 2 and a half pence, xtian -- 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
