On Feb 21, 2:20 pm, William Stein <wst...@gmail.com> wrote: > There are many (important) potential users who do not use Sage at all > just because the startup time is so bad. It's really unfortunate > how much I let the startup time of Sage slip over the years. Sorry.
I see that Jeroen has relaxed the startup time requirement from 1.5 seconds to 2.0 seconds as of changeset c67488a4ece1 (2011-01-13), so that should help (once the patchbot starts testing from newer releases containing this changeset). In any case I am not complaining about Sage being slow to start up, or even about Sage considering itself to have failed doctesting if it doesn't start up within a certain time, I suppose. But conditions can vary, I would imagine, and startup delays encountered while testing a patch (thereby leading to a "TestsFailed" mark on the patch) might not actually be due to anything the patch changed in Sage's codebase. Does the patchbot doctest the vanilla distribution from time to time as a control sample? If not, that might be an idea... > The startup-by-forking demo that Jeroen wrote is really an awesome way > of partly attacking this problem... Yeah, that is nice! -Keshav -- 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
