Re: Guile base64
() l...@gnu.org (Ludovic Courtès) () Mon, 06 Sep 2010 00:47:46 +0200 > Could someone suggest a replacement? Maybe (rnrs bytevectors) and (rnrs io ports)? Thanks for the tip; i'll look into those modules.
scm_with_continuation_barrier returning #f on throws ?
Hello !
With guile-1.8.7, scm_with_continuation_barrier does not seam to return
SCM_BOOL_F on error. I did this :
static void *load_file(void *filename)
{
scm_c_primitive_load((char const *)filename);
scm_force_output(scm_current_output_port());
return NULL;
}
...
SCM res = scm_with_guile(load_file, "a_file_that_doesnt_exist");
Thus, scm_c_primitive_load displays an ERROR message and throws out.
But scm_with_guile, ie. the underlying scm_with_continuation_barrier,
returns 0x0.
Is the bug in the doc, in the code or once again in my head ?
Re: Give up Guile
Marek Kubica writes: > On Thu, 02 Sep 2010 14:14:24 -0400 > Joel James Adamson wrote: > >> > After several day's investigating Guile, I decide to give up and >> > return to PLT-scheme/MIT-scheme. The experience is so frustrating. >> >> And why are you telling us??? > > Obviously to make it better. While it might be a bit troll-ish, all > critique holds some truth and is an occasion to improve on. I definitely see that, however all he offered was negative criticism. As I told the OP, if he wants a *particular user experience* then he should use the platform that specializes on and tries to deliver that user experience. Just saying "Guile sucks because I want PLT Scheme" is not offering any improvements for Guile. Joel -- Joel J. Adamson Servedio Lab University of North Carolina at Chapel Hill FSF Member #8164 http://www.unc.edu/~adamsonj pgpFOUf5xc1Fa.pgp Description: PGP signature
Re: make-module question.
Hello,
On Sat 04 Sep 2010 23:10, Ian Hulin writes:
> If #:key is a superset of #:optional
#:key is not a superset of optional. Keyword parameters are bound by
name; optional parameters are bound by position. Keyword parameters are
not bound by position.
((lambda* (#:optional foo) foo) 10) => 10
((lambda* (#:key foo) foo) 10) => Error.
> If this is the case, could I request an enhancement for this the base
> code to declared using
> (define* (make-module #:key (size 31) (uses '()) (binder #f))
>...)
Unfortunately we cannot, as this would be an incompatible change.
> If make-module is declared using #;key, how would this affect calling it
> from code with scm_call_1, scm_call_2, scm_call_3 or scm_call_n?
> SCM keyname = scm_str2symbol ("#;uses");
In modern Guile (>= 1.8), one makes a symbol from a string via
scm_from_locale_symbol. However in this case you want to make a keyword,
so use scm_from_locale_keyword ("uses").
(Actually, we should be using scm_from_latin1_keyword here, but that
doesn't exist yet.)
> SCM modlist = scm_list_2 (scm_str2sym("ice-9 syncase"),
> scm_str2sym("ice-9 debug"))
Neither of these modules are needed in current Guile. scm_str2sym is not
a part of Guile either; assuming it is one of your functions, you should
use your own namespace (ly_ for example), not Guile's (scm_).
> scm_call_2( SCM_VARIABLE_REF (scm_make_module_x), keyname, modlist):
Yes, you would call it like this.
Regards,
Andy
--
http://wingolog.org/
Re: Guile base64
Hi, On Sun 05 Sep 2010 23:58, Thien-Thi Nguyen writes: > () Romel Sandoval > () Fri, 03 Sep 2010 15:15:58 -0500 > >Do you think any of mentioned implementations should be >included with Guile 2.0 or it's better to keep it apart? > > Probably next week (2010-09-13 onward) i can find some time to add > (ice-9 base64) from Guile 1.4.x Cool! Perhaps we could combine interfaces -- the stream-based (I presume; I haven't seen your code yet, but I know your Scheme code looks good) interfaces from yours, and the map-3-to-4 stuff from Andreas' R6RS port of my base64.scm. I say this because the r6rs port already uses bytevectors, and compiles down to fairly tight VM code. I haven't run benchmarks though. Cheers, Andy -- http://wingolog.org/
currying
Hello, 1. I have curried (compose) so that I can define (not2) below, (define compose(lambda(f) (lambda(g) (lambda(x) (f(g x)) (define not2 (compose not) ) ((not2 null?)'()) This is OK. 2. I do this also, (define both(lambda(f) (lambda(a b) (and(f a)(f b)) ) )) (define neithernull (both ((compose not)null?) ) ) (neither '()'()) This is OK. 3. But both here below are ERR, (define neither (both ((compose not)) ) (define neither (both (compose not)) The first one doesn't compile. The second one does, but this is not what I meant; It suspects two arguments, the '() '() for example , instead of one, the funtion null? for example. Can I define (neither) is some way that I can do something like ((neither null?)'(a)'(a)) Thanks, Eric J.
