There have been some attempts to build termite with gambc-4.0b22, and I had tried it earlier without success, so I thought I'd report my attempt more clearly so somebody might be able to help (hi, Guillaume!). Basically, I think the definition of ##make-macro-descr in _eval.scm has changed, so the use of it in termite should be updated. If I'm right, perhaps someone can tell me why I get
[brad:~/programs/termite/examples] lucier% gsi cell.scm *** ERROR IN "cell.scm"@21.24 -- Datum or EOF expected
after I install gsi+termite. cons.scm at least finishes:
[brad:~/programs/termite/examples] lucier% gsi cons.scm #<kons #2> 1 2 (1 . 2) (123 . 2) done
Shat start[123].sh are supposed to do is anybody's guess ;-).
Brad
I used the darcs command at toute.ca to download termite.
To debug, I changed termite.scm as follows:
[brad:~/programs/termite] lucier% rcsdiff termite.scm =================================================================== RCS file: RCS/termite.scm,v retrieving revision 1.1 diff -r1.1 termite.scm 43c43 < (compile-time-load "termite/match-support.scm") ---
(compile-time-load "match-support.scm")
70a71
#|
82a84,85
|#
I put the the time-related stuff in termite-time.scm:
(c-declare "#include <time.h>")
(define formatted-current-time (c-lambda () char-string "time_t t; struct tm* tm; char str[64]; time(&t); tm = localtime(&t); strftime(str, 64, "20%y-%m-%d %H:%M:%S", tm); ___result = str;"))
So now I can compile and load termite-time separately and I load termite.scm into the interpreter, and I immediately get:
^C[brad:~/programs/termite] lucier% gsi Gambit Version 4.0 beta 22
(load "termite-time")
"/Users/lucier/programs/termite/termite-time.o1"
(load "termite.scm")
*** ERROR IN "match.scm"@6.1 -- Wrong number of arguments passed to procedure (##make-macro-descr -4 '#<procedure #2> #f) 1>
Looking at the definition of ##make-macro-descr in _eval.scm, perhaps you need:
--- termite.scm 2007/04/19 01:57:02 1.1 +++ termite.scm 2007/04/19 02:11:02 @@ -30,6 +30,7 @@ ##interaction-cte ',(car pattern) (##make-macro-descr + #f ',(form-size (cdr pattern)) ,src #f)))))
Afficher les réponses par date
On 4/18/07, Bradley Lucier lucier@math.purdue.edu wrote:
There have been some attempts to build termite with gambc-4.0b22, and I had tried it earlier without success, so I thought I'd report my attempt more clearly so somebody might be able to help (hi, Guillaume!).
Hi Brad :)
I've applied your fix, thanks! It's in the repository and I've updated the tarball.
I've also updated the 'INSTALL' file. It can now be used as a shell script to install the current version of Gambit along with the current version of Termite. As an example:
$ lynx --source http://toute.ca/termite/INSTALL | sh
will get Gambit and Termite and will compile them properly.
[brad:~/programs/termite/examples] lucier% gsi cell.scm
*** ERROR IN "cell.scm"@21.24 -- Datum or EOF expected
An extra parenthesis... how embarassing :/
Shat start[123].sh are supposed to do is anybody's guess ;-).
You can use those scripts to quickly start nodes and have them named as node1, node2 and node3 in your programs.
For example:
[in console 1] sh start1.sh
[in console 2] sh start2.sh
(remote-spawn node1 (lambda () (print 'yay)))
[in console 1]
yay
Guillaume
On 4/19/07, Guillaume Germain guillaume.germain@gmail.com wrote:
On 4/18/07, Bradley Lucier lucier@math.purdue.edu wrote:[...]
Note that I've also changed Termite's license to BSD so that people feel free to do whatever they want with it, but Gambit remains LGPL, of course. I will also gladly accept contributions to Termite and suggestions for improvement.
Guillaume
On Apr 19, 2007, at 12:45 AM, Guillaume Germain wrote:
On 4/19/07, Guillaume Germain guillaume.germain@gmail.com wrote: On 4/18/07, Bradley Lucier lucier@math.purdue.edu wrote: [...]
Note that I've also changed Termite's license to BSD so that people feel free to do whatever they want with it, but Gambit remains LGPL,
Well, dual LGPL and Apache 2.0.
Brad
Hi Guillaume,
I am working on Windows and I have followed the instructions in the INSTALL link and sadly gsi still just returns without doing anything (seemingly).
Gambit 4.0b22 is okay but the termite build seems to do something odd. Is there any way I can get some more information to investigate this? (running in debug mode? a logging flag?)
Thanks
Tom
Guillaume Germain wrote:
On 4/18/07, Bradley Lucier lucier@math.purdue.edu wrote:
There have been some attempts to build termite with gambc-4.0b22, and I had tried it earlier without success, so I thought I'd report my attempt more clearly so somebody might be able to help (hi, Guillaume!).
Hi Brad :)
I've applied your fix, thanks! It's in the repository and I've updated the tarball.
I've also updated the 'INSTALL' file. It can now be used as a shell script to install the current version of Gambit along with the current version of Termite. As an example:
$ lynx --source http://toute.ca/termite/INSTALL | sh
will get Gambit and Termite and will compile them properly.
[brad:~/programs/termite/examples] lucier% gsi cell.scm
*** ERROR IN "cell.scm"@21.24 -- Datum or EOF expected
An extra parenthesis... how embarassing :/
Shat start[123].sh are supposed to do is anybody's guess ;-).
You can use those scripts to quickly start nodes and have them named as node1, node2 and node3 in your programs.
For example:
[in console 1] sh start1.sh
[in console 2] sh start2.sh
(remote-spawn node1 (lambda () (print 'yay)))
[in console 1]
yay
Guillaume
Gambit-list mailing list Gambit-list@iro.umontreal.ca https://webmail.iro.umontreal.ca/mailman/listinfo/gambit-list