These are the reasons I think we might be getting ahead of ourselves pushing for a binary PDL package of some sort before PDL actually builds portably on win32, linux, *bsd, macosx,...
Not to mention the bad impression that could come from PDL "working" on some platforms and "not working" on others because of various issues with the binary releases and lack of full-time maintainers. --Chris On 7/30/2010 6:47 PM, Karl Glazebrook wrote: > > The issue is every binary bundle needs a > maintainer with access to the system in > question and a virgin test system. And it's > all non-trivial. > > Volunteers for Linux and Win32 are welcome. > Though Linux might be a nightmare. > > Karl > On 30/07/2010, at 10:20 PM, Daniel Carrera wrote: > >> Hi Chris, >> >> On Fri, Jul 30, 2010 at 2:10 PM, Chris Marshall wrote: >>> Daniel Carrera wrote: >>>> >>>> Maybe we should make a "SciPerl" distribution >>>> that includes PDL plus other stuff, >>>> statically compiled if necessary. Add >>>> PDL::Stats, and whatever else seems useful. >>> >>> I would expect a SciPerl (from the name) to be >>> more cross-platform than just working on the Mac. >> >> That is exactly how I meant it. Cross-platform. It doesn't see right >> to make a special Mac-only distribution with more than just PDL (as >> Karl said) and not do the same for other platforms. So I wanted to >> talk about making a cross-platform SciPerl that has the extra >> features. >> >> >>> PDL-2.4.7 will be a big improvement over >>> PDL-2.4.6 but the one thing that *won't* be >>> ready is making PDL build cross-platform >>> without expert intervention or a manual build >>> by recipe. >> >> :-( >> >>> How about something like a SciPDL bundle for >>> Macs? I'm not sure we're ready to take on >>> the scientific distribution for all of perl. >> >> Why not call it "PDL Bundle" then? >> >> >> Daniel. _______________________________________________ Perldl mailing list [email protected] http://mailman.jach.hawaii.edu/mailman/listinfo/perldl
