On 12 June 2010 13:02, kcrisman <kcris...@gmail.com> wrote: > >> I suspect there is a difference between what is needed (base) and what >> is recommended. A failure for a recommended package to build might not >> be reason to break the build of Sage. > > Correct, in fact we only very recently re-enabled building recommended > packages. But in practice they are heavily relied upon by all kinds > of things people actually use R for. > > - kcrisman
In that case, there are 3 conclusions I feel 1) It would be dumb to test in spkg-instal 2) It is good to check in spkg-check, which is what happens now. 3) Given running SPKG_CHECK on Sage is not often done (I've seen comments like "only do it if you really know what you are doing"), there needs to be some test of this in the doctests or similar. >From what others have said, an ideal requirement would be to 1) Get a list of recommended packages from that version of R - preferably in a way that does not require them to be hard-coded in a test script, but generated by R. 2) Test if all those recommended packages have actually built. Does that sound reasonable? If so, I'll try to sort out how to do it, though any help would be appreciated, as I've never written any sort of test in Sage, and I have never used R. But I see this as pretty important, so will put some time in. Dave -- 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
