Hi...
This doesn't seem to be the hottest topic of all time. More below. On 12/07/2008, at 12:16 AM, mabshoff wrote: > On Jul 10, 11:23 pm, David Philp <[EMAIL PROTECTED]> wrote: > > Hi David, > >> R does not work inside sage if LD_LIBRARY_PATH includes /sw/lib. >> (Which it ordinarily wouldn't, that was a mistake of mine.) That bug >> is pretty trivial and easily fixed. > > Well, it is likely a problem that we do not ship our own libjpeg. We > will do so in the future (Tom Boothby has been poking me to review > that ticket for a while). It may be a problem, but it seems an inappropriate fix to the problem I had. >> But I found that R is the only component of sage that links to system >> frameworks. (Aside from python.) That seems to be a bit >> inconsistent >> for sage, i.e. R obviously does not link to Apple frameworks on other >> platforms. > > Well, R is still somewhat in flux inside Sage and there have been > people who want Sage to become a framework, too. Could be very useful. But a sage framework would presumably be a platform-specific way of exposing 'standard sage', which is ideally platform-independent everywhere it makes sense. It makes sense to me that R is 'standard sage', and should not provide, expose or require platform-specific features. I like the idea that a sage framework would only provide features that are translatable to other platforms, hence a Mac-sage application would be reasonably portable. (E.g. providing graph-drawing to Mac Sage Apps via R's aqua device would make them essentially unportable.) That's getting ahead though. My issue with libJPEG and CoreFoundation is only really that the current spec as I perceive it is that 'standard sage' should be as uncontaminated by the host platform as possible, and R don't conform to that spec. IOW I think 'sage becoming a framework' is irrelevant to the issue at hand, and should continue to be so. OTOH I suppose if your vision for Sage includes "sage is the best way to get the latest and greatest version of any of its components" then you would aim to include platform-specific features. >> The aqua device is compiled in Mac-sage-R, though my feeling is that >> it wouldn't be a drama to remove it. There is plenty inside the R >> configure script about detecting CoreFoundation, though no simple >> option. I think it would be a good idea to switch off CF as well as >> Aqua. >> >> So the two questions are: >> * would switching off Aqua have any benefit, or inconvenience anyone? >> * would switching off CF be a good idea too? > > I do not think it would matter if we turned them off, but other people > might have an opinion there, too. I am already turning off the Aqua > device in 64 bit mode on OSX for R anyway. But R 2.7.1 might have a > fix for those issues. Passing configure --disable-corefoundation doesn't seem to do anything, but I don't know much about autoconf so I might not have been doing it right. D --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
