Gambit-list Mars 2006

gambit-list@iro.umontreal.ca

10 participants
22 discussions

Poker server using Gambit
by Greenidge, Leroy 14 Mar '06

14 Mar '06

Hi Joel, I am starting my own poker site & i wonder if you could give me some advice ? Regards L Greenidge. This e-mail (including all attachments) is confidential and may be privileged. It is for the exclusive use of the addressee only. If you are not the addressee, you are hereby notified that any dissemination of this communication is strictly prohibited. If you have received this communication in error, please erase all copies of the message and its attachments and notify us immediately at support(a)gpxs.net <mailto:support@gpxs.net>. Thank You.

2 1

found my first "bug" in gambit: web-server example README
by Samuel Montgomery-Blinn 14 Mar '06

14 Mar '06

Well, it has been a week now tinkering with Gambit, and my first "bug", found when trying to find a valid example of finding the index of a given character in a string, which I found in the web-server example directory. The README stipulates the manner of starting the server: gsi html http web-server 8000 However, the continuation-example calculator does not work unless base64 is also included: gsi base64 html http web-server 8000 Hooray. This glorious contribution will live in infamy, I am sure. I think my pains of learning the "Scheme Way" and Gambit would have been much less had I read more of the examples and paid a thorough attention to: http://www.schemers.org/Documents/Standards/R5RS/HTML/ Instead of playing a guessing game as to which SRFIs are implemented! (string-index from SRFI 13 in particular, for which the web-server has the example I was seeking as "find-char-pos" -- in other words answering my question, "Yes, you do have to write this yourself."). If nothing else, reading through the http.scm "parse-uri" procedure (and others nearby) makes me wonder if this code was generated or if some human actually did sit down and write all the states... -Sam

1 0

Re: [gambit-list] newbie question: output from one port as input to another port
by Marc Feeley 13 Mar '06

13 Mar '06

On 13-Mar-06, at 12:20 PM, Samuel Montgomery-Blinn wrote: > Thank you both for the answers (and the code samples). Just seeing > how one is "supposed" to write such functions is a very good head > start to understanding. > > -Sam No problem. By the way, your message got me thinking about how to improve the input-to-output latency. On a slow input device, it is probably worthwhile to read only as many octets as are currently available and transfer them to the output, in particular when the buffers are big. This can be done by blocking the reading thread only for the reading of the first octet, and use a nonblocking read for the remaining octets. Here's the new code (the copy-octet-port- to-octet-port-asap procedure). Give it a try and see if it improves performance. Marc (define buf-size 4096) (define (copy-octet-port-to-octet-port in out) (let ((buf (make-u8vector buf-size))) (let loop () (let ((n (read-subu8vector buf 0 (u8vector-length buf) in))) (if (> n 0) (begin (write-subu8vector buf 0 n out) (force-output out) (loop))))))) (define (copy-octet-port-to-octet-port-asap in out) (let ((buf (make-u8vector buf-size))) (let loop () (input-port-timeout-set! in #f) ; block thread until at ; least 1 octet available from in (let ((n (read-subu8vector buf 0 1 in))) (if (> n 0) (begin (input-port-timeout-set! in 0) ; get whatever is still in the ; buffer without blocking (let ((m (read-subu8vector buf n (u8vector-length buf) in))) (write-subu8vector buf 0 (+ n m) out) (force-output out) (loop))))))))

1 0

re: Gambit-C on Windows
by Greg Lee 13 Mar '06

13 Mar '06

On January 19, 2006 Darren Pegg reported: >Perhaps I'm missing something, but if I build a release version of >Gambit-Cb17 VS 2003 V 7.1.3088 >running from the windows command prompt I get the following :- >C:\gambc40b17\gsi>Release\gsi >Gambit Version 4.0 beta 17 >> (+ 1 2) >3 >> >*** EOF again to exit >doesn't seem to matter what I enter, I always get the EOF exit, both under >gsi and gsc Has there been any follow-up to this? As anyone gotten the MSVS port to work?

2 1

Re: [gambit-list] newbie question: output from one port as input to another port
by Marc Feeley 12 Mar '06

12 Mar '06

On 12-Mar-06, at 2:22 PM, Samuel Montgomery-Blinn wrote: > > Doesn't this waste time not reading the input port while writing to > the output port if the input port could be read while the output > port is being written, and similarly waste time not writing to the > output port while reading from the input port? > > -Sam If you have a single CPU, it will either be filling the buffer (reading) or draining the buffer (writing). So it can't really do both at the same time. However, your disk, network card, and CPU can all be working in parallel because the network card and disk have internal buffers (i.e. when the write-subu8vector returns, the network card (and TCP stack) will still be actively sending packets out, which leaves enough time for the next disk access to occur). Of course YMMV depending on the speed of the CPU, network card and disk. Just measure how close you can get to the bandwidth of your network. I'm confident you will be pretty close. Marc

1 0

Re: Gambit-C on OS X with Developers tools 2.2.1 (make check fails)
by Marc Feeley 12 Mar '06

12 Mar '06

On 9-Mar-06, at 3:29 PM, Mark Smith wrote: > Hi, > > I hope this is the right address for this. I'm trying to get Gambit up > and running on OSX. I've tried a few times in the past. It seems I can > never build a version that will pass the checks before install. > > I recently upgraded my development tools I thought I'd give compiling > it another go :). It compiles fine and it doesn't cause a heap > overflow now (which has to be a good thing) so I consider it a step in > the right direction however this time It fails with > > ------------ TEST 2 (error handling) > ../gsi/gsi -:h4000 -f error.scm < error.scm > test2.out > diff test2.ok test2.out && rm -f test2.out > 1218c1218 > < (string?-read-line-open-tcp-client) => #t > --- >> ;;; CALL EXPRESSION (force-output '#<input-output-port #15 (tcp- >> client "www.iro.umontreal.ca" 80)>) IS WRONG: (string?-read-line- >> open-tcp-client) > make[1]: *** [test2] Error 1 > make: *** [check] Error 2 > > Is it safe to ignore this because I'd really like to try gambit out > seriously? If not then is there any advice you could give me to fix > this? > > Thank you, > > Mark. Hi. Yes it is highly probable that Gambit was compiled properly. Some people on other systems have also reported the same error on this test. I may remove or change this test because it assumes you are building Gambit on a machine currently connected to the internet (and properly configured) and that "www.iro.umontreal.ca" is up and running, and reachable. You might try a different web site. Marc

1 0

newbie question: output from one port as input to another port
by Samuel Montgomery-Blinn 12 Mar '06

12 Mar '06

So, I have been trying to learn Gambit-C's port mechanisms. This is likely an infantile question, but it seems that it should be the most basic thing to be able to take one input port and "connect" it to an output port. The specific use for me is sending the contents of a file over a network (tcp) port. I have tried a few methods, mostly "busy" and likely lossy methods. 1. read the entire contents of the file port, then output the file to the network port. (wastes time in which earlier parts of the file could be written to the network port) 2. read a line from the file port, and output the line to the network port. repeat until eof. (wastes time for many reasons) I was guessing that something to do with the procedures "call-with-input-file" or "with-output-to-file" might be something towards what might be most succinctly expressed as some sort of UNIX "sendfile" functionality, but as I am new both to Scheme (or any Lisp for that matter) and Gambit-C, I thought I would both say "hello" to this list and ask for its help and guidance. If this sort of question is not a faux pas, I certainly have more! -Sam

3 2

Bug fix in html-lib.scm
by Bradley Lucier 10 Mar '06

10 Mar '06

Marc: html-lib.scm had only a one-character bug, but having #t instead of #f as the last argument of (define-input-tag) gave <input> tags a closing tag, which was wrong. Sorry. Here's a corrected version. Brad

1 0

Couple more inlining suggestions
by Bradley Lucier 08 Mar '06

08 Mar '06

I'm just going through the O(N^2) parts of my image smoothing codes and listing the things that come up. Brad 1. (sqrt x) => (if (and (##flonum? x) (##flonum.<= 0. x)) ;; OK for infinities and NaNs (##flonum.sqrt x) (sqrt x)) 2. (max x y) => (cond ((and (##fixnum? x) (##fixnum? y)) (##fixnum.max x y)) ((and (##flonum? x) (##flonum? y)) (##flonum.max x y)) (else (max x y))) 3. Same for min.

1 0

Suggestions for inlining
by Bradley Lucier 07 Mar '06

07 Mar '06

Marc: Here are a few things that could be inlined productively (I won't write the tests to see whether the functions when called are the standard functions). Brad 1. (exact->inexact x) => (if (##fixnum? x) (##flonum.<-fixnum x) (exact->inexact x)) 2. (vector-ref v i) (and all the [suf][8,16,32]vector-ref's) => (if (and (##vector? v) (##fixnum? i) (##fixnum.<= 0 i) (##fixnum.< i (##vector-length v))) (##vector-ref v i) (vector-ref v i)) 3. (vector-set v i x) (and all the [suf][8,16,32]vector-set's with appropriate checks for the value of x) (if (and (##vector? v) (##fixnum? i) (##fixnum.<= 0 i) (##fixnum.< i (##vector-length v))) (##vector-set v i x) (vector-ref v i x))

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Gambit-list Mars 2006