From: Raymond T. <ray...@er...> - 2005-11-17 13:58:45
|
>>>>> "AJ" == A J Rossini <A.J.> writes: AJ> I've evaluated matlisp 2.0beta and it seems like it might work, with AJ> some glue, as the backend replacement for the current numerical linear AJ> algebra code. But of course, there are some questions: AJ> #1 - mixing of ASDF (some of the current tools, cells-gtk and clunit, AJ> that I've started using) and MK:DEFSYSTEM installations? Is it worth AJ> adding a joint ASDF build system for uniformity/install compatibility? I'm not really an asdf user, so it probably won't be kept up-to-date, unless users do so. However, I have no problem mixing and matching asdf with mk:defsys. This works especially well if your lisp supports an extension to require that will load systems using asdf or mk:defsys. AJ> #2 - current status of the 2.0beta? (it looks to be stalled for 2+ years now) It's feature-complete. :-) I don't use it much, so any changes to matlisp these days are driven by user requests or patches. (Which reminds me that here is an outstanding request.) AJ> #3 - interest in taking back some of the possible changes? Yes, of course! AJ> Also, have there been many changes in CVS compared to the 2003 AJ> 2.0 beta release? Probably quite a few, but probably not many that a user would notice. I'd have to check logs to be sure.... Ray |