On Aug 2, 1:54 pm, William Stein <wst...@gmail.com> wrote: > On Mon, Aug 2, 2010 at 10:52 AM, kcrisman <kcris...@gmail.com> wrote: > > >> > Could you formulate some more precise descriptions of experimental and > >> > optional, for discussion? > > >> I think that, among other things, optional packages should build on > >> all "supported platforms" (whatever that means -- this phrase needs to > >> be defined carefully somewhere: see #9487), while experimental ones > >> need not. > > > +1 - but do our current optional packages do so? > > Definitely not. Ensuring this would be a massive project, which would > be well worth doing. > > I used to at least test regularly that they installed on sage.math, > but I haven't even done that lately. > Regularly ensuring at least that all the optional spkg's install > *somewhere* would be a step forward from the > current situation. >
We need to ask the R people how they do this. Somehow, they have a system that ensures that packages are not accepted as optional packages on CRAN unless they build on some build bot - I don't know any of the details, but it sounds like the difference between the CRAN and http://www.rforge.net/, which specifically does not have "make check on-commit, nighlty builds of packages, testing on various plarforms and full CRAN-like reposity access". The R foundation is also a lot bigger (and it's an official FSF project), so maybe that is how they do it. - kcrisman -- 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
