- Gambit-list - mailman.iro.umontreal.ca

I/O puzzle...
by Marc Feeley 10 Jui '09

10 Jui '09

Puzzle... what's the bug in the following code? See end of message for answer. (define http-server-start! (lambda (hs stop?) (let ((server-port (open-tcp-server (list port-number: (server-port-number hs) backlog: 128 reuse-address: #t server-address: '#u8(127 0 0 1) ; on localhost interface only char-encoding: 'ISO-8859-1)))) (input-port-timeout-set! server-port 3) (accept-connections hs server-port stop?)))) (define accept-connections (lambda (hs server-port stop?) (let loop () (let ((connection (read server-port))) (if (eof-object? connection) (if (stop?) (close-input-port server-port) (loop)) (begin (if (server-threaded? hs) (let ((dummy-port (open-dummy))) (parameterize ((current-input-port dummy-port) (current-output-port dummy-port)) (thread-start! (make-thread (lambda () (serve-connection hs connection)))))) (serve-connection hs connection)) (loop))))))) . . . . . . . . Answer: the (input-port-timeout-set! server-port 3) is at the wrong place. It sets a timeout ***3 seconds in the future***. So after 3 seconds have passed all calls to read in accept-connections will not block (and lead to a tight, CPU intensive loop). The correct approach is to put the call to (input-port-timeout-set! server-port 3) just before the call to read. Puzzle #2: why did Gambit chose to define timeouts that are not relative to the moment that an I/O function is called? Marc

2 3

ignore rest of the file
by lowly coder 09 Jui '09

09 Jui '09

foo.scm includes bar.scm in bar.scm, is there any command I cay say that will tell foo.scm to ignore the rest of the file? I don't want to commnent out all the second half of bar.scm nor do I want to put a lambda around it (complaints about define) is there someway I can access a continuation or something and jump out of the include? thanks!

3 2

compiled modulers fun
by lowly coder 09 Jui '09

09 Jui '09

I've recently discovered how cool gambit is in the following way: I have some interpreted code, it's running kind of slow, so I compiled my code, and got an automatic 5x speed up. Yay!! Here's the problem though: foo.scm -> foo.o1, in the same directory. This is bad for one reason -- it screws up tab completion; I can no longer do things like: vim fo<TAB>; since it gets stuck at "foo." ... so is there anyway to make both compilation and load default to somethign like "foo.scm" -> _foo.o1 or .foo.o1 ? I realize I can manually hack this, but if there was a global default, that'd be really cool too (interpreting the path for load + rewriting it seems a bit hacky) Thanks!

4 3

profiling / exceptions
by lowly coder 08 Jui '09

08 Jui '09

Here's my problem: I am running a multi threaded application on gsi. Thread X executes thunk THUNK. if THUNK throws an exception, I want to be able to use the repl to jump into thread X and see what is wrong (and have the whole interactive repl debugger at my finger tips). However, I also want to profile THUNK. Thus, before THUNK starts, I want to run (profile-start!) and when THUNK finishes (or if an exception is thrown), I want it to call (profile-stop!) Here now is the problem ... if I don't catch the exception (and allow myself to debug it at the GSI), my profiler will colelct data of me sitting at the REPl debugging. On the toher hand, if I catch the exception to call (profile-stop!), I no longer know how I can debug the exception at the REPL. How do I get around this? Thanks!

1 1

Is gambit *just* a compiler/interpreter?
by Adrien Piérard 07 Jui '09

07 Jui '09

Hello everyone. As others here too I guess, I do not program only in Scheme, and I also keep myself informed of what happens around in the Common Lisp, Haskell or ML communities (since I have a functional programming bias). I believe that most of you are aware of the impressive fame of Haskell and its main implementation GHC, and I would even consider betting that "do it in a monad" will soon belong to the dialect of the IT marketing teams and deciders. Let's see what I think helped the buzz spread, let alone the language itself: - cabal: convenient and community friendly way to share, downloade and builp code. - Haskell platform: on click under windows, and you have a fully set up system with all-you-need-ready-to-use (c). - spare time: every single library coded by haskell users somehow ended up in the community shared code base. No need to redo what's already done. - communication: the haskell weekly news: state of the art, blogs, and IRC quotes. The community is alive and shows it to the world. I guess that the closest to this in "our" (very split) community is PLT-scheme. Yet, I think that Gambit too is somehow evolving along a bigger and bigger user base, but may benefit from more development targetting the (potential) users rathers that its speed or efficiency. =========== Suggested improvements to Gambit: - We have a dumping ground, which slowly but surely grows. I think that it is now time to reconsider doing a CPAN-like or cabal-like system for Gambit. Users/programmers should be suggested to put there any single line of once-useful code, without fearing comments regarding quality or else. If someone wants this code but finds it slow, it'll be patched… Snow has tried to do that, but portability killed i.(I gave up portabililty a while ago). - I haven't tried blackhole yet, don't know its limitations, but if it is good, let's make in part of Gambit (even as an external branch or whatever). - A better explanaition of of it diverges from pure RnRS (that is, emphasize on define-type, non-blocking IOs, and so on), and how it makes it relevent for most real-life programming tasks. - Involve the community and code the wish-list (or at least, give a guided tour of gambit's internals and of its where-is-whats) so that some may consider implementing those requests themselves. =========== Suggested improvements to Scheme: - We have #scheme on Freenode that is lively and wealthy. We have scheme.dk/planet which is lively and wealthy too (I just wished Grant could stop twittering on it and make real posts with more than 134 characters that all start like "here is ...". Grant, if you read this :P) Now, maybe a publication could be useful too, monthly at first, maybe unofficial, that sums up recent activity, from "XXX now runs on the iPhone, to "John Doe got a Scheme job at Great Compagny Inc", not forgetting Prof. Farnsworth's latest paper on continuations". I may be stating the obvious, but I felt the urge to share it (so much that I wrote this too quickly and will probably regret having sent it)… P! -- Français, English, 日本語, 한국어

3 3

eval
by lowly coder 07 Jui '09

07 Jui '09

Why does gambit's eval ignore the second argument (environment)? Is there some way I fake this by grabbing environments & evaluating in them? Thanks! (eval being able to choose an environment, in my mind, is a bug plus of scheme) Thanks!

2 2

Termite + Black Hole
by Jamie Briggs 07 Jui '09

07 Jui '09

On the Black Hole page of the Gambit wiki, in the "additional Libraries" section there is the line: Termite: A version suited for usage in Black Hole is found at X. Are BH and Termite currently incompatible? And if so, how? I was just going to try using them together but then realized I would probably notice gross problems but might be not catch subtle ones. - Jamie

3 3

gambit, iphone, and LOAD
by James Long 06 Jui '09

06 Jui '09

Hey, folks. I'm using Gambit on the iPhone. There's one thing I can't get working on it, and that's the LOAD procedure. I've heard reports that you can't load dynamic libraries on the iPhone, but I can't find any documentation about it. Besides, LOAD doesn't work to load in raw scheme code either. When my application attempts to use LOAD in an app running on the iPhone simulator, it simply crashes without any messages (both when the file is compiled and not). I'm not sure how to debug this, and I was wondering two things: do you use LOAD on the iphone? is there anything in LOAD which I should look at when running Gambit on a different platform than usual? I think loading dynamic libraries should work in the simulator. And EVAL definitely works, so loading in raw source should work too. I just looked around the Gambit source and everything around these two mechanisms seems fine. - James

5 6

Performance of the interpreter
by Lam Luu 06 Jui '09

06 Jui '09

Hello everyone, I am trying to write a web service in Scheme, so I wish to keep the code in source form (not compiled). I know that Gambit has an very good compiler, but I have not heard much about the interpreter (gsi). Does anyone have an idea how efficient it is, especially in comparison to the compiler as well as other interpreted languages like Perl, Python, and Ruby? Thank you!

2 1

scheme scriptable unicode terminal?
by lowly coder 05 Jui '09

05 Jui '09

Hi! I've been a big fan of rxvt-unicode for a while now; but I really really want a terminal with scheme as it's extension language. Anyone using one? Thanks!

3 2