Re: [racket] Chipmunk FFI bindings on Github! (v0.1)

2012-06-07 Thread gabor papp
Hi Vince, Thank you for your work on this. Have you looked into Jay McCarthy's Chipmunk package on Planet? http://planet.racket-lang.org/package-source/jaymccarthy/chipmunk.plt/1/0/ Best, Gabor Racket Users list: http://lists.racket-lang.org/users

Re: [racket] Embedded Racket 5.1.3 on Windows

2011-10-06 Thread gabor papp
Meanwhile, if you move the call to Interpreter::Register() to just after scheme_basic_env(), like this: does it work? Yes, it works! Thank you. One of our modules 'collada-import' is giving an error during boot, which you need to uncomment in modules/scheme/fluxus.ss if you want to experiment

Re: [racket] Embedded Racket 5.1.3 on Windows

2011-10-05 Thread gabor papp
Can you say more about the crash, perhaps something along the lines of a stack trace? I'm not sure if it's helpful, but here it is from OS X: Seg fault (internal error) at 0xb7 Program received signal EXC_BAD_ACCESS, Could not access memory. Reason: KERN_PROTECTION_FAILURE at address: 0x00

Re: [racket] Embedded Racket 5.1.3 on Windows

2011-10-05 Thread gabor papp
There was no change related to the stack size, but maybe it has to do with support for places? If you build with `--disable-places' and maybe also `--disable-futures', does it work? Thanks. Compiling with these two options made it work. Is this a bug? Or should we change something with the embedd

Re: [racket] Embedded Racket 5.1.3 on Windows

2011-10-04 Thread gabor papp
Racket recently changed to assume an 8 MB stack under Windows, instead of the default 1 MB stack, and I did not properly update the documentation. Has something also changed on OS X and Linux? We see Fluxus crashing if compiled with Racket versions above 5.1.1, although the stack size there is 8

Re: [racket] Regexp character classes

2011-08-27 Thread gabor papp
On 8/27/11 10:04 AM, Neil Van Dyke wrote: #lang racket/base (define str "17.4 25.4 15.7 13.7 19.4 20.9 ") (regexp-split #rx"[ \t]+" str) or with posix character class: (regexp-split #px"[[:space:]]+" str) _ For list-related administrative tasks:

Re: [racket] using planet from embedded application

2011-03-28 Thread gabor papp
hi Imran, Could you elaborate a little? i used the mysql package, so it might be different in your case. i suggest that you go to the planet page of the module and download the source manually: http://planet.racket-lang.org/package-source/untyped/net-repl.plt/1/0/ save the files into a folder,

Re: [racket] using planet from embedded application

2011-03-27 Thread gabor papp
Resurrecting this from a month ago, as I'm hitting the same problem. I'm using vim-7.3, with racket embedded. I'm trying to use the net-repl library, but am hitting the same error that Gabor did. i ended up downloading the planet module and required the main module file instead. i had to download

[racket] using planet from embedded application

2011-02-13 Thread gabor papp
I'm trying to require jaz's mysql package like this: (require (planet jaz/mysql:1:7)) It works fine if I use it from racket, but if I use it from fluxus, which embeds racket, I got the following error: default-module-name-resolver: the kernel's resolver works only on `quote' forms; given: '(lib

Re: [racket] scheme_make_vector gc macros

2011-01-01 Thread gabor papp
Assuming that `src' never refers to GCable memory, this looks fine. You don't really need to register `tmp', but it's ok to do so. Thank you. 'src' does not refer to GCable memory, which i think is not the case in the opposite direction when we are receiving vectors from scheme. Should this be d

[racket] scheme_make_vector gc macros

2011-01-01 Thread gabor papp
We are experiencing some small memory leaks in Fluxus, and I am wondering if the garbage collector registration of the following code is right. It should return an array of float values as a scheme vector. I read the part on memory allocation in Inside: Racket C API, but I'm not sure how it shou

[racket] (build-vector) segfault

2010-09-22 Thread gabor papp
typing the following code in racket generates a segfault in osx 10.5.8. i believe it's a bug. Welcome to Racket v5.0.1. > (define (v) (build-vector (vector-length 10) add1)) > (v) Seg fault (internal error) at 0x15 Bus error without the define it returns an error as expected: > (build-vector (v

Re: [racket] FFI callback object pointer

2010-07-17 Thread gabor papp
I think you should consider encapsulating the type of values that can be passed around and the box to store them in the closure of the Thank you Thomas and Matthew for your suggestion. It works great and it's really much simpler than the possible solution I was thinking about. Best, Gabor

[racket] FFI callback object pointer

2010-07-17 Thread gabor papp
I'm trying to write an FFI wrapper for a C library that sets and gets variables via callback functions. The callbacks look like this: void callback(void *value, void *clientData); which I translated to this: define _callback (_fun (value : _pointer) (client-data : _pointer)

[racket] Embedding Racket into a Program

2010-06-30 Thread gabor papp
According to the Inside Racket C API (http://docs.racket-lang.org/inside/overview.html) example embedding Racket into a program requires the 'raco ctool --c-mods base.c ++lib racket/base' command, but it gives an error: raco ctool: unknown switch: ++lib I guess it is different with raco than it