[CM] cvs and whatnot
Fri, 12 May 2006 11:56:27 -0700
Sourceforge claims it will have CVS working sometime today, but any
such users will need to recheck-out the sources under the new name
cvs -d:pserver:email@example.com:/cvsroot/snd login
cvs -z3 -d:pserver:firstname.lastname@example.org:/cvsroot/snd co cvs-snd
I hope that will work -- I haven't tried it, and my connection to sourceforge has
been broken for several days.
I've looked at a bunch of extension languages; here's my quick take on some:
* ecl: (CL) doable; will require that all direct refs be through
* xen.c (can't include its header files!); it also
* requires config.h (colliding with mine), and uses
* the field "complex" which confuses C.
* eel: (C) a commercial product
* elastic: (C) looks dead (no change since 2001), like Lua in calling sequences
* elk: (Scheme) looks dead (no change since 1996) and is worse than stklos
* in terms of name-space problems
* GameMonkey: () c++, windows oriented (no linux I think)
* librep: (CL) looks dead, but I still have the macros (xen.h)
* lua: () doable, but it looks like akcl to me (and it's not freeware?)
* mzscheme: (Scheme) doable (I have the xen.h macros for it), but...
* ocaml: (ML) not an extension language, as far as I can tell
* octave: (Matlab) c++, probably doable -- I'm looking into this currently
* [2.1.73 won't build -- g++ trouble?]
* pike: (C) not an extension language
* python: () looks like ruby to me -- why duplicate?
* rscheme: (Scheme) won't build on my systems
* s-lang: (C) probably doable -- would need to wrap everything in my own
* struct, and 7 args max is too few.
* squirrel: () c++, like lua in call sequence
* stklos: (Scheme) doesn't build libstklos yet, and has many non-unique names
* in its headers
So, octave is the only one that looks interesting (despite being c++ based -- gah).