I'm nearing completion of a web application with stateful server
continuations, and am thinking about future scalability.
If we limit the discussion to the specific problem of retention of
server-side continuations in gc'd memory (and not to whether
server-side continuations are of value)
I
Matthew,
I use the MysterX COM layer of sections 2.1-2.2 for working with Microsoft's ADO
library
I have no intentions/ability to use the other sections
I would be happy to share all my ADO related code if it would be of benefit.
Racket Users list:
http://lists.r
Jay,
->dynamic-wind issue
--
The short answer is the code I'm working on is an implementation of
Harel's hierarchical state machine formalism.
I hesitate to provide a long answer as I fear it would be an imposition
on your time and goodwill. Though I'm happy to provid
FYI, C-level implementation appears to be in src\fun.c, line 7722 as
*scheme_dynamic_wind
Jay,
->dynamic-wind issue
--
The short answer is the code I'm working on is an implementation of
Harel's hierarchical state machine formalism.
I hesitate to provide a long ans
Jay,
Thanks very much for the answer.
My current thinking is that the dynamic-wind described on page 17 of Friedman's
Constraining Control may provide enough of dynamic-wind's functionality for my
needs.
It doesn't appear to have any of the forbidden constructs* of the web-server
language.
and
Shiri,
While AJAX and client-side continuations enable significant client-side
computational complexity, I think they complement vs. eliminate the need
for server-side complexity.
I'm in pursuit of both (1) server complexity and (2) server-side
statelessness
I could achieve (1) by giving
Sorry, should have included a brief example.
Here's an example of what I've implemented, which is just Harel's
hierarchical state machines
You can create a Hierarchical State Machine by defining individual nodes
in a flat file, as below.
each node has 6 keyword parameters , state, parent, e
know if
that is too opaque ...)
Robby
On Sat, Dec 31, 2011 at 7:56 PM, Galler
wrote:
Sorry, should have included a brief example.
Here's an example of what I've implemented, which is just Harel's
hierarchical state machines
You can create a Hierarchical State Machine by def
ful to sort out some of the subtleties:
http://www.cs.utah.edu/plt/delim-cont/
hth,
Robby
On Sat, Dec 31, 2011 at 8:18 PM, Galler
wrote:
Robby,
Thanks
Not at all opaque. I was experimenting with that earlier this
afternoon, as
below.
Actuallly, the way you phrased it helps clarify someth
u may
need to use some kind of unique values in the marks to be able to tell
if the shared tails _really_ are shared or if they simply have the
same structure.
You might find this paper useful to sort out some of the subtleties:
http://www.cs.utah.edu/plt/delim-cont/
hth,
Robby
On Sat, Dec 3
On the subject of porting #lang racket to #lang web-server code
I provide a few lessons learned for the archives for the next person to
attempt this.
1) Bindings, Serializability , and anonymous functions
1A) the predicate serializable? from racket/serializable is invaluable
#lang web-se
When using stateless web servlets in #lang webserver with serve/servlet:
Q: Is there a method for identifying the uri-scheme of the inbound
request and redirecting HTTP schemed uris to HTTPS scheme for SSL?
Issue 1)
Examining the (url-scheme (request-uri request)) of the request start
pas
Description:
the stateless webserver allows a user to generate a stuffer to
manipulate the encoding of the ANF sent to the client in the URL
For example
(stuffer-chain serialize-stuffer (stuffer-compose base64-stuffer
(HMAC-SHA1-stuffer #"mysupersecretkey")))
will prepend a digest to the b
Neil V. mentioned his preference for directly calling the OpenSSL
libraries vs. using mzcrypto (planet vyzo/crypto)
I wanted to provide a small example using this technique for
generating and retrieving a 1024 bit unencrypted private key in
PEM format.
(define res (process "openssl genrsa 10
, Robby Findler wrote:
One comment here: you really want to be sure that you close all three
of the ports that process returns, or else you'll run into problems if
you use this in a larger context.
Robby
On Tue, Apr 24, 2012 at 12:14 PM, Galler
wrote:
Neil V. mentioned his preferenc
(body "Some other message"
allows you to inspect the exception and return a different thing.
(The push will happen shortly.)
Jay
On Mon, Apr 23, 2012 at 3:35 PM, Galler
wrote:
Description:
the stateless webserver allows a user to generate a stuffer to
manipulate
the encoding
Eli, Neil
This thread raises an interesting point.
An Artifact is herein defined as some piece of Racket code that does a useful
piece of work.
Some of the lesser lights in the community, (such as myself) will, over the
course of a working day, come up with a useful Artifact or two.
Three conc
Sam,
I love it and added a few more.
Thanks
Zack
On Tue, Apr 24, 2012 at 7:35 PM, Sam Tobin-Hochstadt wrote:
On Tue, Apr 24, 2012 at 7:17 PM, Galler
wrote:
My Suggestion:
Is there room, either on the blog, or in the user group, for a Racket
Artifacts
thread
i.e.
subj: How to call
Neil,
Let me respond to your argument with a specific question.
Below is some code that sets up fluid-like dynamic variable bindings using Jay's
web-cells.
I haven't put this code up on GitHub (or elsewhere)
(define/contract (make-dyno-bindings ids)
(-> (listof symbol?) (listof (cons/c symbo
Matthias Felleisen writes:
>
> I don't think Neil is questioning the need for sharing. I think he's arguing
that we should organize the huge
> wealth of sharing that already exists.
>
You're right. I was probably attacking a strawman argument. Sorry, Neil.
The search and retrieval question i
New Artifacts posted: Improved! Low-Fat! Exciting!
https://github.com/plt/racket/wiki/Artifacts
1. Specifying a HMAC-SHA1 stuffer for the stateless web-server
2. Split a string into lines
3. Fetch the contents of a URL
4. Generate a n-byte key for use in MAC authentication (like HMAC-SHA1)
(build-list 22 (λ (x) (/ (* x
(- x 1))
2)))
Discard the first two values from the list.
R./
Zack
Racket Users list:
http://lists.racket-lang.org/users
;and without having to discard first two elements->
(build-list 21 (compose (λ (x) (/ (* x
(+ x 1))
2))
add1))
Racket Users list:
http://lists.racket-lang.org/users
> (sequence-constructor (list 1 2 3)) = list
> (sequence-constructor (vector 1 2 3)) = vector
> etc
>
> I'd like to use it for a declare-mappable macro that extends functions
> of one argument
> to map over sequences. As in (sin (list 1 2 3)) = (list (sin 1) (sin
> 2) (sin 3)).
>
I think you a
ok. then it is a type problem, I was wrong.
On Sun, Apr 29, 2012 at 3:56 PM, Jens Axel Søgaard wrote:
2012/4/29 Galler :
(distribute sin list 1 2 3)
(0.8414709848078965 0.9092974268256817 0.1411200080598672)
This makes the constructor explicit. That's what I want to avoid.
The pr
BACKGROUND:
Per the documentation:
with-continuation-mark causes a key to be associated with a binding
within a continuation frame
continuation-mark-set->list returns the binding(s) associated with that
particular key, for any given continuation-mark-set
while
continuation-mark-set-first
))
'id1)
'(1)))
On Sun, May 6, 2012 at 8:10 PM, Galler wrote:
Thanks Robby.
I'm attempting to implement the behavior described in Section 3.2 of
your
2007 paper, specifically
"each continuation frame can have any number of marks with distinct
keys"
by de
at's why you see only one binding.
continuation-mark-set-first can be more efficient than
continuation-mark-set->list and it is more likely to work in futures.
That's the only reason it exists.
Robby
On Sat, May 5, 2012 at 11:02 PM, Galler
wrote:
BACKGROUND:
Per the document
> but continuation marks are carefully designed to avoid breaking tail
> recursion (they let you understand the tail behavior of your program,
> more accurately) so that's why you see only one binding.
Robby,
Yes. That is exactly what happened.
On further inspection, I was using with-continuatio
This code was generated in response to the user who sought to implement
run-length encoding of a bit-vector on Sunday night.
I didn't post this to the board b/c there's a much easier way to solve
problem using regular expressions, which Eli B. demonstrated.
But, the (infinite-k-pump) function st
ijn wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi Zack,
On 15-05-12 18:56, Galler wrote:
This code was generated in response to the user who sought to
implement run-length encoding of a bit-vector on Sunday night.
I didn't post this to the board b/c there's a much easier w
lers
3) moving to the next signal in the signal stream
4) transitioning between states
Let me sit down tonight and figure out if those are possible purely with
functional composition.
R./
Zack
On Thu, May 24, 2012 at 11:59 AM, Eli Barzilay wrote:
More than a week ago, Galler wrote:
Y, that's closer to what I had in mind.
On Thu, May 24, 2012 at 1:18 PM, Eli Barzilay wrote:
Just now, Galler wrote:
Eli,
I fully agree with you that any FSM is equivalent to functional
composition and can be implemented in the manner you show.
However , in the way you've imple
ago, Galler wrote:
Y, that's closer to what I had in mind.
On Thu, May 24, 2012 at 1:18 PM, Eli Barzilay wrote:
Just now, Galler wrote:
Eli,
I fully agree with you that any FSM is equivalent to functional
composition and can be implemented in the manner you show.
However , in th
Is there a function or syntax to obtain the module-level binding of a
function?
For example:
(define (my-function)
(display
(some-useful-function-which-captures-name-of-module-level-binding)))
;-> 'my-function
Thanks very much.
R./
Zack
Racket Users list:
http://l
ad its own define
that cooperated with some-useful-... to do that.
You would probably want to do this with a macro, tho, so that you get
the lexical property. I guess.
Robby
On Tue, Aug 7, 2012 at 1:15 PM, Galler wrote:
Is there a function or syntax to obtain the module-level binding of a
func
Thanks Danny
On Tue, Aug 7, 2012 at 4:58 PM, Danny Yoo wrote:
;; Question to others: how can this be simplified?
(splicing-let-syntax ([get-literal-metadata
(lambda (stx)
#`(#%datum . #,(format "~s"
I did a slight s
7, 2012, at 3:36 PM, Galler wrote:
> Is 'Semantics Engineering with PLT Redex' where I should get started on
language-engineering?
No, this book is about engineering the mathematical semantics of a language
(existing or to be designed) so that you can test properties (such as
well-def
Eli, Matt, and Jay
Under 5.3, there's been a near order-of-magnitude improvement in
compilation time for the web-server language.
I didn't notice prior discussion of this, although Jay has mentioned
compilation time as a potential drawback in implementing stateless code.
Its a very meaningf
This may have been asked and answered. Apologies if redundant.
The improvement in stateless compilation time in 5.3 has made the
debugger a practical tool for me. (thanks again)
Is it technically feasible to evaluate simple expressions in the dynamic
extent of the point at which code is pause
Robby and Matt,
I found one could very easily use the pict library to create bitmaps and
disseminate over HTTP as png-encoded binary streams
Which is really quite nice.
Thanks much for your work. The 2006 paper is very clear.
Recursively defined picts structures are powerful abstractions
;
Neil,
This library is enormously useful.
Thanks very much for the work required to implement.
R./
Zack
Racket Users list:
http://lists.racket-lang.org/users
call-with-continuation-prompt is your friend in the definitions window
(call-with-continuation-prompt (λ _ (+ 100
(call-with-composable-continuation
(λ (k) (+ 1000 (k 1
Greg Graham writes:
>>do you think this is a crazy idea or not?
I won't call it crazy.
Can I call it overengineered?
I would write the following in T-SQL to achieve what you want:
sp_configure 'Ad Hoc Distributed Queries', 1
reconfigure
USE [your-database]
GO
insert into OPENROWSET('Micros
Good evening,
I'm using Racket 5.3.6 under windows XP
When I obtain the whalesong package as follows:
#lang racket/base
(require (planet dyoo/whalesong:1:21/make-launcher))
then attempt to run whalesong at the command prompt as follows
C:\Documents and Settings\Galler>whalesong
I rec
Joe,
I've been pulling and building directly
Thanks very much for this, but I'm not at all competent on racket
internals
Could you please sketch out where to place the files and how to build
them?
I won't need help after that.
Sorry to bother.
Thanks again.
R./
Zack
Joe: building from github worked. Thanks.
If anyone from the future interested:
Navigate to
https://github.com/dyoo/whalesong
and download the master branch (choose 'download ZIP' if unfamiliar)
Extract zipped archive and move the entire whalesong directory into the
Racket/collects directory
Nathaniel Rudavsky-Brody writes:
> Hi Zack and everyone,
> I'm going back to this old thread
(http://lists.racket-lang.org/users/archive/2013-January/056176.html), since
I have a similar problem with raco and flomap, and can't tell if anyone
found a solution. Are there any new ideas?
Nathaniel
...
> 3. comment out flomap-struct.rkt's provide form
...
For clarity, that should be in the pasted version of flowmap-struct, not the
original.
Racket Users list:
http://lists.racket-lang.org/users
Danny,
define-resource is generating a unexpected specialize! error when
compiled.
Here's the minimal setup: Please note that I'm using the latest github
version of your source, not the Planet package.
Thanks for taking a look
R./
Zack
Program resource-error.rkt
Galler writes:
>
> Danny,define-resource is generating a unexpected specialize! error when
compiled.
Clarification: the error is occuring at run time, not at compile time. I
believe its originating in file whalesong/resource/compile-time.rkt
The following very-naive patch eliminat
Henri
Is this what you're after?
#lang racket/base
(require (for-syntax syntax/parse
racket/base
))
(provide make-script)
(begin-for-syntax
(define-syntax-class unit-dictionary
#:attributes (str)
(pattern (~literal p)
#:wit
Konrad,
Adding the super-bar syntax-class below
and the following #with directive to the foo syntax class
#:with sb:super-bar #'((foo-symbol ...) (bar.symbol ...))
should accomplish what you want without recourse to explicit parameters
Let me know if you need/want further discussion.
Thank
Happy birthday,
Thanks for all your contributions and intellectual generosity.
R./
Zack
Racket Users list:
http://lists.racket-lang.org/users
>
> Another possible workaround for now would be to cast the field to a
> varchar and call string->number on the result.
>
numeric is an exact datatype in SQL Server; you'd also want to compose
inexact->exact with string->number to preserve exactness (if that's a concern)
i.e (inexact->exact
Use (define/contract...
to enforce a contract within a module
(define/contract (amount)
(-> positive?)
-7)
#|
amount: broke its contract
promised: positive?, produced: -7
in: the range of
(-> positive?)
contract from: (function amount)
blaming: (function amount)
at: unsav
>
> Even if you use define/contract, you may violate the contract on recursive
calls. This is intentional.
>
Thanks.
I hadn't run into that behavior before and wasn't aware.
Could you possibly elaborate on the thinking wrt this design choice?
(define/contract (my-natural-number n)
(-> exa
I've been thinking about a reified continuation:
The documentation (See Guide s10.3) says "A continuation is a value that
encapsulates a piece of an expression context"
I'm coming to the belief that the continuation is actually "an ordered
collection of computation-branches, with facilities prov
>>What you can NOT do even in this world is visit the past and perform a
different computation than (+ 1 2).
Yes. I agree that Racket's eval order combined with which branch the
continuation is reified on uniquely determines which branches are in the
past and which are in the future.
I acknowled
>>Now imagine that we lied. We don't really reify evaluation contexts as
[continuation] procedures but as [continuation] objects with additional
operations for inspecting and possibly mutating them. One could imagine that
continuations are organized as a [abstract] sequence of [abstract] frames:
N
> The phrase 'evaluation context' has been in the literature since 1986 for just
this purpose. I know, I coined it.
>
I wonder if there's been terminology drift in the reference
For example:
In the documentation, 'expression context' appears and is defined (Ref. 1.2.3.3)
'continuation frame'
Hello,
Is there a destructuring-bind for regular expressions?
Something along the lines of the purely fictional function
"with-destructuring-regex", below:
(with-destructuring-regex ([(list #px"[\\D]*" #px"[\\d]+\\.[\\d]+"
#px"[\\D]*") "This is a 40.2 Song about Japan" (x y z)])
bout Japan"
[(regexp #px"([\\D]*)([\\d]+\\.[\\d]+)([\\D]*)" (list _ x y z))
(list x y z)])
;; produces
'("This is a " "40.2" " Song about Japan")
The `_` is ignoring the first result of `regexp-match`, which is the
full match.
Sam
On Fri
how () in regular expressions work
Thanks much.
On Fri, Jan 4, 2013 at 3:44 PM, Sam Tobin-Hochstadt wrote:
On Fri, Jan 4, 2013 at 3:37 PM, Galler wrote:
Could you possibly briefly elaborate on how the parentheses in the
regular
expression make the destructuring work?
The () in the
Hello,
I encountered difficulty compiling a #lang web server application after
upgrading to v.5.3.1 from v5.3 last night. O/S is WinXP service pack 3
The executable would compile into an .exe file.
But when run, a console window would briefly open then close.
The program did not execute.
I
no longer occurred.
Based on this, I would assert the issue is specific to 5.3.1 vs 5.3 compilation
behavior.
R./
Zack
- Original Message -
From: Haiwei Zhou
To: Galler
Cc: users
Sent: Sunday, January 27, 2013 6:58 AM
Subject: Re: [racket] raco exe difficutly with
or,
and then closes. If you ran from the shell, you'd be able read the
error. My guess is that the error will say that some file can't be
opened that is part of the configuration of the Web server and I
haven't properly communicated its need to the exe builder.
Jay
On Sat, Jan 26, 20
is in the next release.
Jay
On Mon, Jan 28, 2013 at 7:46 AM, Galler
wrote:
Jay, Thanks.
Here's the result run under Racket 5.3.1 with windows XP SP3
I wasn't aware of the windows command line trick. Thanks for
suggestion.
C:\SOURCE\Simplest RACO FAILURE>simplest-raco-failu
The (require (for-syntax racket/match/parse)) solved the executable
problem. Thanks.
I have a second raco exe issue--
and I don't think this is related to 5.3.1 b/c I believe I encountered
similar problem earlier today using raco with 5.3
This is also using different underlying source code
s. I can't find the
thread though.
Jay
On Mon, Jan 28, 2013 at 8:25 AM, Galler < lzgal...@optonline.net
>
wrote:
The (require (for-syntax racket/match/parse)) solved the executable
problem. Thanks.
I have a second raco exe issue--
and I don't think this is related to 5.3.1 b/
flomaps. I can't find the
thread though.
Jay
On Mon, Jan 28, 2013 at 8:25 AM, Galler < lzgal...@optonline.net
>
wrote:
The (require (for-syntax racket/match/parse)) solved the executable
problem. Thanks.
I have a second raco exe issue--
and I don't think this is related
I wasn't able to isolate out the code that was causing the exception.
Below is the stack trace. This problem does not occur when running the
app in DrRacket (the app works fine there),
but appears after having compiled the application via raco exe, when the
server is handling a response.
flomap and #lang web server above might be
dispositive towards identifying the problem
On Tue, Jan 29, 2013 at 7:44 AM, Jay McCarthy wrote:
[Adding Neil] Do you know what commit fixed the missing require
for-syntax in flomap?
Jay
On Mon, Jan 28, 2013 at 12:09 PM, Galler
wrote:
I wasn&#
traversal of the require hierarchy
cannot
"see".
hth,
Robby
On Tue, Jan 29, 2013 at 10:31 AM, Galler mailto:lzgal...@optonline.net>> wrote:
Some additional detail:
1) the flomap exception below shows up when attempting to run
compiled code in both releases 5.3 and 5.3
Robby
On Tue, Jan 29, 2013 at 11:14 AM, Galler < lzgal...@optonline.net
>
wrote:
Neil,
I was surprised that flomap showed up at all. I'm not calling it
explicitly (or to my knowledge implicitly)
Is there a technique to figure out what part of my code is dependent
upon it?
On T
Preliminary Solution:
I will send more info in more detail when I've got a test case, also I
need to test against 5.3.1 (works on 5.3) but:
1) The cause of the flomap exception was a user-code dependency on the
racket library stepper.rkt in a #lang web server module, complied under
raco e
Racket 5.3.1 is amazing.
Compiled, it runs 20 update queries per second on a local SQL Server
data store, whereas 5.3 ran, at best 5 queries per second. Both figures
inclusive of garbage-collection. Also it is stable (5.2 and earlier had
memory leaks, probably from mysterx library)
I don't h
>> DrRacket's GUI is more responsive.
Eli, Matt and Robby,
Qualitative comments:
I don't know if change occurred in v5.3.1 or v5.3.2 but DrRacket GUI is now
materially more responsive when editing large racket source code files (~2500
lines).
I thank you very much for the improvement.
I recomp
So, more details would be welcome.
I was wrong.
Minor cycles are consuming ~110ms and recovering ~25MB, major cycles are
consuming ~400ms and recovering 40-60MB.
All is as it should be. Apologies.
R./
Zack
Racket Users list:
http://lists.racket-lang.org/users
Good evening,
I've been investigating lang web-server, performance and garbage
collection
I have a question regarding anonymous functions and web cells
Examples below require:
#lang web-server
(require racket/serialize
web-server/lang/web-cells)
Constants appear serializable by th
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
Jay and Robby,
Thanks much.
I found that discussion very helpful.
R./
Zack
Racket Users list:
http://lists.racket-lang.org/users
Dear Racketeers,
*SMALL PREAMBLE*
I'm trying to understand the observed GC behavior so I can create better
application performance, and I'm focusing on the major GC-cycle.
For usability, my desire would be to have a major-garbage-collection
cycle last about the same time as a typical request
Matt, Ryan
Thanks very much for the answers, let me think about.
I'm looking at the issue Ryan brings up with module references to
lru.rkt, which surprised me.
R./
Zack
On Fri, Feb 15, 2013 at 2:39 PM, Matthew Flatt wrote:
At Fri, 15 Feb 2013 14:16:38 -0500 (EST), Galler wrote:
Wh
Matthew,
A major collection is triggered when current memory use is twice
the use at the last major collection.
I'm not following this sentence, as it implies logarithmic growth in
memory, which is not observed.
At some point in the executable's life, the major GC cycles consistently
occur
Matthew and Ryan and Robby,
MAJOR COLLECTIONS WITHOUT ACTIVITY
The issue regarding major collections occuring with no activity was due
to the collect function executing in lru.rkt on a timer. Now no
activity equals no collections. Thanks Ryan.
BACKGROUND AND NEXT QUESTION
Purpose of fo
Dear all,
I'm having difficulties compiling racket/place code using both raco exe
and DrRacket menu item
I've isolated the simplest example reproducing the behavior
The parent and worker files are as below, taken from the racket
reference
http://docs.racket-lang.org/reference/places.html
Bot
Galler writes:
Hello again.
The problem seems to be headed towards resolution with some combination of
using
1) (define-runtime-path ) in the parent file to reference the worker file
and
2) using the ++lib parameter when compiling with raco exe for libraries
called by the worker file
I&#x
Here's what I determined.
RULES For Compiling code that will use dynamic-places with RACO
Racket: 5.3.5
O/S Windows XP sp3
terminology:
worker = module that will run in the new virtual machine
parent = module calling the dynamic-place procedure
1. In the worker module explicitly use the (
Two issues:
1: wrap the function-under-test in a thunk
2. check for exception type, here exn:fail:contract? which is raised by
raise-argument-error. the more general exn:fail? would also work.
(require rackunit)
(test-exn "negative coin" exn:fail:contract? (λ _ (sum-coins -1 3 5 7) ))
__
Also, here's a small additional bit of code to demonstrate what test-exn is
actually checking.
(require rackunit)
;pass
(test-exn "negative coin" exn:fail:contract? (λ _ (sum-coins -1 3 5 7) ))
;fail -- "Wrong exception raised"
(test-exn "negative coin" exn:fail:filesystem? (λ _ (sum-coins -1 3
>>why a thunk?
The inferential leap to be made is that Racket uses applicative order,
meaning the arguments of a procedure are each evaluated from left-to-right
before the procedure is itself evaluated
for example:
(+ 1 (+ 2 3))
is reduced to
(+ 1 5)
before being evaluated as
6
the procedur
Racketeers,
I'm working through eval, namespace, and syntactical macros
My question concerns re-definition of identifiers.
I've worked through three cases below, demonstrating Racket's default
behavior.
Case 1 shows that identifiers defined at module (or is it top?) level
cannot be redefine
I don't see any obstacle to adding a more fine-grained reflective
operation, such as `namespace-set-variable-constant!`. Would that be
useful?
Matt,
I don't think the introduction of namespace-set-variable-constant! worth
the effort, though I thank you for the solution.
Let me explain my reason
Sam Phillips writes:
>
> On Fri, Oct 8, 2010 at 4:40 PM, Jay McCarthy wrote:
> > I use dynamic-wind for this. If there is something better, I don't
> > know what it is. dynamic-wind is a little bit funny though because if
> > you capture continuations then the in/out handlers can run multiple
>
Sam Phillips writes:
>
> Hi All,
>
> I understand using call-with-exception-handler/with-handlers/etc. to
> catch exceptions in a block of code, but I'm at a loss to what the best
> way to do a "finally" or cleanup action after a block of code. My
> intuition is to use dynamic-wind, but I figu
96 matches
Mail list logo
