At Mon, 11 Feb 2013 16:41:15 +0100, Philipp Dikmann wrote:
> It appears that the GUI canvas% refreshes itself whenever it receives a
> mouse-event%.
> I intend to refresh the canvas at a steady pace (using a timer%) and
> independently of any mouse-events instead.
> Is there a reasonable way to s
At Thu, 14 Feb 2013 13:42:38 +, Norman Gray wrote:
> (parameterize ((current-subprocess-custodian-mode #f))
> [...])
>
> I get an error "current-subprocess-custodian-mode: expects argument of type
> <'interrupt, 'kill, or #f>; given: #f".
Repair pushed to the git repo. Thanks for that repo
the errno are not so descriptiv. By the way, how can I make a printf
in dll visible through rackets ffi. When I make this for my function
>
> (define-dreamlog DisplayHelloFromDLL (_fun -> (r : _void)))
>
> and eval (DisplayHelloFromDLL)
>
> I get nothing or I rewrite the function which returns a c
Last time I looked at that benchmark, I changed the hash table to hold
boxed values, so that the updating `hash-ref' plus `hash-set!' could be
replaced with `hash-ref', `unbox' (which is fast), and `set-box!'
(which is also fast).
To confirm other comments: `hash-update' is implemented as just
`ha
Yes, that fixed it. Also I tried what Robby suggested, and that worked too.
On 02/14/2013 01:05 PM, Matthew Flatt wrote:
> I don't know whether it will solve your problem, but a cross-device
> issue could be real for some users, so I've changed `net/ftp' to download
> to the same directory as the
I don't know whether it will solve your problem, but a cross-device
issue could be real for some users, so I've changed `net/ftp' to download
to the same directory as the destination path.
At Thu, 14 Feb 2013 13:56:37 -0600, Robby Findler wrote:
> You can just put the llncs2e.zip into the right pl
On Thu, Feb 14, 2013 at 06:06:44AM -0700, Jay McCarthy wrote:
> The answer to this question depends entirely on the site. Most sites
> will have an HTML form which expects you to send the credentials to
> via the query string or maybe post data. The response is likely to
> contain a cookie, in its
You can just put the llncs2e.zip into the right place and then scribble
won't try to download it anymore. I forget if you have to unpack it first,
but a quick peek at the source should fill you in (in the error message).
On Thu, Feb 14, 2013 at 1:49 PM, Jon Rafkind wrote:
> Strange scribble err
Also, this is with the latest git as of right now:
294bed209ed903e9ac97095c369edfc9d01ff555
On 02/14/2013 12:49 PM, Jon Rafkind wrote:
> Strange scribble error today. I haven't debugged it much yet, in case someone
> has an idea of whats going on.
>
> $ scribble --pdf main.scrbl
> File not found
Strange scribble error today. I haven't debugged it much yet, in case someone
has an idea of whats going on.
$ scribble --pdf main.scrbl
File not found: /home/jon/.racket/llncs.cls
Downloading via ftp://ftp.springer.de/pub/tex/latex/llncs/latex2e/llncs2e.zip...
rename-file-or-directory: cannot re
Also on more than one occasion I've had to work around issues with entry
scaling with even mutable hashtables, A few 100K entries, they are fine
but rapidly grind to a halt. The Python folks have hash tables nailed,
über performance and scaling.
On Thu, Feb 14, 2013 at 2:20 PM, James Bergstra w
I was actually trying to understand the performance bottlenecks of the
"knucleotide" challenge of the language shootout:
http://benchmarksgame.alioth.debian.org/u64/program.php?test=knucleotide&lang=racket&id=4
Most of the time seems to be spent in the loop that builds the hash table.
(Incidenta
On Thu, Feb 14, 2013 at 2:15 PM, Robby Findler
wrote:
> Is this a bottleneck in some real program? Perhaps you'd share that (or some
> inner loop of it)? I wonder if there is a better way to speed it up than
> this.
`hash-set!` (which, like `hash-update`, is written in Racket and
hashes twice) is
hash-update is defined in Racket and, besides the error checking, is
basically the composition of those two functions. I expect it was written
as something to abstract over a common pattern, not speed things up.
Is this a bottleneck in some real program? Perhaps you'd share that (or
some inner loo
If I am right, then it would make sense to use hash-update when the hash is big
enough that a log-time traversal takes longer than a general procedure call's
set-up and tear-down.
-Ian
- Original Message -
From: James Bergstra
To: J. Ian Johnson
Cc: users@racket-lang.org
Sent: Thu, 14
Thanks, I suppose I just had my hopes up. I thought the more compact
form might involve fewer expression evaluations and require just one
hash lookup instead of two, and therefore might run something like
twice as fast, rather than slightly slower.
On Thu, Feb 14, 2013 at 1:36 PM, J. Ian Johnson
My guess would be the higher-orderness on hash-update causes the (minor)
slowdown.
-Ian
- Original Message -
From: James Bergstra
To: users@racket-lang.org
Sent: Thu, 14 Feb 2013 13:16:47 -0500 (EST)
Subject: [racket] Question about hash table efficiency
Hi list, just discovered racket a
Jay and Robby,
Thanks much.
I found that discussion very helpful.
R./
Zack
Racket Users list:
http://lists.racket-lang.org/users
Whenever I encounter this situation, I tend to switch to Racket's class system:
(define vehicle
(class object%
(define vehicle-state
(class object%
(field position ;; 3D point
velocity ;; real
)
...))
(field state ;; vehicle-s
Hi list, just discovered racket and having fun learning about scheme
again. Haven't used it since undergrad. Thanks!
I was wondering about the efficiency of hash-update vs. hash-set and
hash-ref. I thought the former would be faster, otherwise why have it?
A little benchmark script reveals the sam
On Feb 13, 2013, at 9:47 AM, Danny Yoo wrote:
>> The sentence: "Every child is unique; items never share their children
>> even if the latter have the identical content."
>
> Looking at primary sources... Yup, the original specification:
>
>http://okmij.org/ftp/Scheme/SXML.html#SXML Tree
>
Nice idea ;-)
Racket Users list:
http://lists.racket-lang.org/users
Ah yes. We agree. I presume that Zack is concerned with the normal
circumstance of running a racket program from the command line, but I
recognize that you could do lots of creative things that will behave
properly.
Jay
On Thursday, February 14, 2013, Robby Findler wrote:
> That is also what I m
Hello friends,
how can I catch an com exception during execution of some com calls
and how can I redirect the error messages. At the moment when a com exception
occurs it is placed in Drracket.
Yours,
Racket Users list:
http://lists.racket-lang.org/users
That is also what I meant by code (well, I was speaking generally about all
bytecode).
The point I'm making is that this is not a limitation of our runtime
system, it is just because you are imagining there are still references to
that bytecode.
In the specific instance you mention below, the nam
When I say "code", I mean module top-level binding byte-code. There's
no way for Racket to remove
(define (f x) (lambda (y) (printf "Gigantic string: ~a" (+ x y
at the top-level. The byte-code will always contain the gigantic
string. If you ever call f and get back something, that closure cou
Greetings.
I think that current-subprocess-custodian-mode may have some error in its
argument-checking.
If I evaluate the following in DrRacket:
(parameterize ((current-subprocess-custodian-mode #f))
(subprocess (current-output-port)
(current-input-port)
(current-
I'm pretty sure code is collected in (regular) Racket. Small numbers aren't
collected, true, but they also don't really take any space. Big numbers are
collected, too. (I'm not sure about interned symbols.)
Robby
On Thu, Feb 14, 2013 at 7:26 AM, Jay McCarthy wrote:
> In the Web server language,
In the Web server language, all anonymous functions are turned into
top-level structures masquerading as functions where the fields are
their free variables. For instance:
(define (f x) (lambda (y) (+ x y))
becomes, roughly,
(struct a-lambda42 (x)
#:property prop:procedure
(lambda (the-structu
The answer to this question depends entirely on the site. Most sites
will have an HTML form which expects you to send the credentials to
via the query string or maybe post data. The response is likely to
contain a cookie, in its headers, which you would include in the
headers on future requests.
J
Hello Matthew,
you are right. My RAcket is x64 but my dll is x86. That is the problem, but the
errno are not so descriptiv. By the way, how can I make a printf in dll visible
through rackets ffi. When I make this for my function
(define-dreamlog DisplayHelloFromDLL (_fun -> (r : _void)))
and
Also, on further reflection, I'm happy to concede that my placing anonymous
functions in a web-cell may not have been a great idea.
If the ultimate answer is 'don't do that', that would be ok.
Racket Users list:
http://lists.racket-lang.org/users
32 matches
Mail list logo
