00:02 (quit) mheld: Quit: mheld
00:16 (quit) Fare: Ping timeout: 240 seconds
00:42 (quit) mwolfe: Remote host closed the connection
00:42 (quit) parcs: Ping timeout: 240 seconds
00:43 (join) parcs
03:29 (join) Sgeo
03:37 Sgeo: Are there alternate PLaneTs? How easy/difficult would it be to use one?
03:42 jonrafkind: there are no other ones, in theory i guess its not super hard to set up
03:42 jonrafkind: you need postgres and trac
03:46 Sgeo: How would someone use an alternate planet?
03:48 (quit) askhader: Remote host closed the connection
03:48 (join) askhader
03:54 jonrafkind: im not sure of the planet source is publicaly available though..
03:55 jonrafkind: there is a planet url somewhere in your racket configuration stuff
04:01 Sgeo: Hmm
04:02 Sgeo: Can individual source files override that?
04:02 jonrafkind: im not sure if its a parameter
04:03 jonrafkind: but i guess it could be made so
04:13 (quit) jonrafkind: Ping timeout: 276 seconds
04:45 (join) aLeSD
04:48 aLeSD: hi all
04:51 Sgeo: Hi
05:37 (quit) martinhex: Ping timeout: 240 seconds
05:38 (join) martinhex
05:48 (join) Fare
05:58 (quit) Fare: Quit: Leaving
06:07 (quit) martinhex: Remote host closed the connection
06:07 (join) martinhex
06:17 (quit) martinhex: Read error: Connection reset by peer
06:17 (join) martinhex
06:18 (quit) aLeSD: Ping timeout: 264 seconds
06:28 (quit) martinhex: Ping timeout: 240 seconds
06:38 Sgeo: Why can't I find any actor-based concurrency stuff?
06:38 (join) martinhex
06:38 (quit) martinhex: Remote host closed the connection
06:39 (join) martinhex
06:50 (quit) martinhex: Ping timeout: 255 seconds
07:01 (join) martinhex
07:02 Sgeo: "A type for Racket?s native Unicode strings, which are in UCS-4 format."
07:02 Sgeo: n/m
07:06 (quit) martinhex: Ping timeout: 255 seconds
07:16 (join) martinhex
07:23 (join) masm
07:36 (quit) martinhex: Ping timeout: 240 seconds
07:39 (join) martinhex
07:46 (quit) martinhex: Ping timeout: 240 seconds
07:47 (join) martinhex
08:02 (join) pmurias
08:02 pmurias: why do frtime programs freeze after a while?
08:53 (join) stis
08:54 (quit) martinhex: Ping timeout: 250 seconds
08:58 (join) martinhex
09:04 (join) martin_hex
09:05 (quit) martinhex: Disconnected by services
09:05 (nick) martin_hex -> martinhex
09:08 (join) _p4bl0
09:09 _p4bl0: hi everyone
09:09 (quit) martinhex: Ping timeout: 240 seconds
09:09 (join) martinhex
09:10 _p4bl0: I'm developing a GUI using Racket. I'd like to have an area% (in fact it's a canvas%) that is stretchable but keep a ratio width/height of 10/21. The canvas% is in a horizontal-panel%. Is this possible?
09:18 (quit) martinhex: Ping timeout: 240 seconds
09:28 (join) martinhex
09:31 (quit) martinhex: Excess Flood
09:31 pmurias: did any one here use frtime for anything?
09:31 (join) martinhex
09:36 (quit) martinhex: Ping timeout: 272 seconds
09:38 (join) MayDaniel
09:41 (join) martinhex
09:48 (quit) martinhex: Ping timeout: 264 seconds
09:50 (join) martinhex
09:55 (quit) martinhex: Read error: Connection reset by peer
10:07 Lajla: R6 is actually pretty cool.
10:12 pmurias: R6?
10:12 (join) martinhex
10:12 cky: pmurias: R6RS, perhaps.
10:14 Lajla: Well, of course.
10:14 Lajla: I would like to see a split in it though as they proposed.
10:14 Lajla: Like, declare some libraries 'optional'
10:15 Lajla: Because it's perhaps a bit too extensive.
10:15 Lajla: Also, requiring the whole numerical tower is just madness.
10:15 Lajla: I always liked the fact that they didn't.
10:15 Lajla: Some implementations can be geared towards tasks that have no need for exact numbers or complex numbers
10:16 (quit) martinhex: Read error: Connection reset by peer
10:16 (join) martinhex
10:40 (join) molbdnilo
10:43 (join) corruptmemory
10:47 (quit) MayDaniel: Read error: Connection reset by peer
10:47 (join) mheld
11:08 (quit) martinhex: Ping timeout: 246 seconds
11:10 (join) martinhex
11:12 (quit) molbdnilo: Quit: molbdnilo
11:13 (part) pmurias
11:20 (join) martin_hex
11:20 (quit) martinhex: Disconnected by services
11:20 (nick) martin_hex -> martinhex
11:29 (quit) martinhex: Remote host closed the connection
11:29 (join) martinhex
11:34 (quit) martinhex: Ping timeout: 264 seconds
11:36 (join) martinhex
11:40 (join) mceier
11:44 (quit) martinhex: Remote host closed the connection
11:45 (join) martinhex
11:47 (quit) masm: Quit: Leaving.
11:49 (quit) martinhex: Ping timeout: 255 seconds
11:50 (quit) bremner: Read error: Operation timed out
11:51 (join) bremner
11:53 (join) martinhex
12:02 (quit) martinhex: Ping timeout: 272 seconds
12:03 (join) martinhex
12:06 (quit) martinhex: Excess Flood
12:26 (join) MayDaniel
12:26 (join) masm
12:42 (quit) mheld: Quit: mheld
13:16 (join) mheld
13:24 (join) Demosthenes
13:24 (quit) MayDaniel: 
13:31 (join) jonrafkind
13:34 clklein: jonrafkind: I think I figured out yesterday's mystery.
13:34 jonrafkind: oh nice
13:35 clklein: http://paste.lisp.org/display/118118
13:35 clklein: call-with-exception-handler handlers run in the context of the raise
13:36 clklein: I'm guessing that the uncaught-exception-handler does the same
13:37 jonrafkind: yea
13:37 jonrafkind: i guess there is some reason for things being that way.. but it sure is confusing
13:39 clklein: Yea, it caught me off guard
13:39 (join) samth
13:48 (join) martinhex
14:03 Lajla: jonrafkind,
14:03 Lajla: you did an argument from ignorance.
14:03 Lajla: I thought I'd just wait about 10 hours after.
14:03 Lajla: to tell you that.
14:03 Lajla: Because it has been itching since.
14:04 jonrafkind: about planet?
14:04 Lajla: jonrafkind, yeah
14:04 Lajla: I see you remember your own arguments from ignorance
14:04 jonrafkind: its in the chat logs
14:04 jonrafkind: it wasn't an argument, i was trying to be helpful. no one else was responding
14:05 jonrafkind: i tried to find where the planet source is but i couldnt find it. i thought it was in iplt
14:05 jonrafkind: i attempted to set up my own planet instance once, which is why i sort of know one or two things
14:05 Lajla: I am so good with this, like, yesterday, I messaged a linguist whom I once talked to in a café 3 years prior per text message with the text 'It's not a third person passive imperative, it's a fourth person active imperative, which actually makes sense', I wish I could see the look on his face.
14:05 Lajla: Well, then you should say 'not that I know off'
14:05 Lajla: And I know you were.
14:06 Lajla: But it has been itching since.
14:06 Lajla: So I thought I'd tell you.
14:08 (join) anRch
14:10 (quit) martinhex: Read error: Connection reset by peer
14:10 (join) martinhex
14:45 (quit) Lajla: Ping timeout: 276 seconds
15:12 (quit) anRch: Quit: anRch
15:23 (join) Lajla
15:26 (join) MayDaniel
16:18 (quit) jonrafkind: Ping timeout: 265 seconds
16:26 (quit) MayDaniel: Read error: Connection reset by peer
16:42 (quit) mheld: Quit: mheld
17:02 (quit) tv|z: Ping timeout: 240 seconds
17:04 (join) tv|z
17:37 (join) mheld
17:51 Sgeo: Why doesn't Racket have actors?
17:52 samth: Sgeo, actors are, in most modern languages, a programming pattern rather than a feature
17:52 samth: you can certainly use threads as actors in Racket
18:02 Lajla: Aren't actors basically just functions in CPS.
18:02 Lajla: Or am I missing my theory.
18:02 samth: Lajla, actors also involve concurrency
18:03 Lajla: Hmm
18:03 Lajla: I rephrase
18:03 samth: in a sequential setting, actors = cps functions
18:03 Lajla: Aren't actors basically functions in a lazy language.
18:03 samth: no, lazyness != concurrency != CPS
18:03 Lajla: And isn't the case that if you defined your functions in such a way that they don't depend on any order that it comes down to the same.
18:03 Lajla: CPS can't work with laziness?
18:04 samth: no, it can, but they aren't the same thing
18:04 Lajla: Oh yeah
18:04 Lajla: No, I mean lazy + CPS
18:04 Lajla: At least, then you have your basic building blocks of that they can only send messages and accept them and all that.
18:04 samth: no, actors are a strict but nondeterministic and concurrent paridigm
18:05 Lajla: You don't say.
18:05 Lajla: Then I have totally gotten this wrong.
18:07 (join) danbrown
18:12 Lajla: samth, is concurrency a formal concept?
18:13 samth: yes
18:14 Lajla: samth, do you have a formal definition, because my google skills are insufficient at tihs juncture.
18:15 _p4bl0: Lajla: here is an entry point to formal concurrency: http://en.wikipedia.org/wiki/Pi_calculus :-)
18:17 samth: Lajla, there are lots of formal models of concurrent systems, _p4bl0 has pointed to one influential one
18:17 Lajla: _p4bl0, I know that, but that is not a formal definition of concurrency.
18:17 Lajla: It is a formal description of concurrent models.
18:17 samth: it's not clear what a formal definition of concurrency would mean
18:17 Lajla: Well, yeah
18:17 Lajla: it wasn't to me either.
18:17 samth: what's the formal definition of computation?
18:17 _p4bl0: samth: yup that's why I said it's only an entry-point
18:18 Lajla: I would not say that computation is a formal category.
18:18 samth: concurrency is like computation
18:18 Lajla: I don't know a formal definition that can seperate a computation from a non-computation.
18:18 Lajla: Ah, so like, these models basically for sake of argument assume that 'all is concurrent' in the same way that ZFC assumes that all is a set?
18:18 Lajla: All they can describe
18:19 Lajla: *
18:19 samth: no, that's not right at all, either about concurrency or about zfc
18:19 Lajla: I am pretty sure that the interpretation of ZFC is that every object in it is a set and ZFC does not assume the existence of non-sets.
18:20 samth: no, zfc is a collection of axiom schemes for describing set formation and relations between sets
18:20 samth: it doesn't make any assumptions
18:20 Sgeo: Do people often type the unicode lambda, or is plain lambda preferred?
18:20 Lajla: THat is saying the same thing, like I said, ZFC does not assume non sets.
18:20 Lajla: samth, not assuming a non set is part of not making assumptions.
18:20 Lajla: And different from assuming that there are no 'non sets'
18:21 Lajla: It does not treat the quaestion what is a set and what not.
18:21 samth: Sgeo, people regularly use the unicode lambda
18:21 samth: DrRacket even has a keybinding for it: Ctrl-\
18:26 _p4bl0: BTW, I'm studiying computer science (in France, I'm in master 1, that's the 4th year of uni) and I'm currently looking for an internship in the US for 4 to 6 month starting mid-march 2011, I'd like to work on programming languages. I already contacted Guy Steele at the Sun Labs (now Oracle) to work in the Programming Langue Research Group (currently working on Fortress), he said he's reading my résumé but he doesn't know yet if the
18:26 _p4bl0: group will take intern next year. So if anyone here knows of another place where I could go, I'm interrested :-)
18:26 Sgeo: Suppose I only wanted to do a define in a certain condition (say, a version number passed in earlier was greater than something). I'd make a macro that may or may not expand into define?
18:27 samth: _p4bl0, i recommend contacting MS Research in redmond as well, and Mozilla Labs (they do a bunch of language stuff)
18:27 samth: Sgeo, i don't recommend doing that
18:27 _p4bl0: Oh and I'd love to work in an academic research team but I'll need to be paid to be able to live in the US, that's why I primarily look at companies
18:27 samth: how would you know when you could refer to the variable?
18:28 _p4bl0: samth: Mozilla is a good idea! MS research work too much on type system from what I heard, and I have enough of that in France if you know what I mean
18:28 Sgeo: The calling code should know that that variable doesn't exist in the older version
18:29 Lajla: Sgeo, some people would say it is safer to always defined it to #f and set! it.
18:32 Sgeo: Hmm
18:32 Sgeo: Safer how, exactly?
18:32 Sgeo: Easier, I guess, since it doesn't involve macros, but safer?
18:32 Sgeo: s/doesn't involve/doesn't rely on/
18:33 Sgeo: Probably easier for me as implementor to use macros anyway
18:33 samth: Sgeo, i don't recommend set!, but it's better to have your scope readily apparent
18:33 Lajla: Sgeo, well, like samth said, how would you know if you can refer to it.
18:33 Lajla: Well, some people would also say that you should be hung by the throat for using set! =)
18:34 Sgeo: I'm thinking of wrapping a C API which gets new functions frequently
18:34 Sgeo: If some code says that the intended version is 77, it either doesn't know of, or otherwise can't use, functions defined later than that
18:35 Lajla: Sgeo, a much simpler way though is just use something like (define da-var (cond ((version-seven? vers) 7) ((version-eight? vers) 8))) or soemthing like that.
18:35 Lajla: Making it #void (I believe) when none is true.
18:35 samth: i would define it to a version that dynamically errors when the function is not available
18:36 Sgeo: Whatever I do, I'd probably write a macro that expands into it, for simplicity?
18:37 samth: Sgeo, i'd write it the long way first, and then write a macro to abstract over the common parts
18:44 (nick) samth -> samth_away
18:51 Lajla: Sgeo, another thing re-reading that part, at what point is the version number known.
18:51 Lajla: Like, you enter it statically into the macro context as parametre.
18:53 Sgeo: Hmm, not entirely sure
18:53 Sgeo: Can you pass arguments into modules?
18:53 Sgeo: Can the parameter stuff do that?
18:54 Lajla: Well, macros can only operate on information that is known at compile time as far as I know.
18:54 Lajla: At least, in what way they expand.
18:54 Lajla: If they expand to some form that features a conditional in it, I'm sure you could hack around.
18:54 Sgeo: Huh, didn't realize that
18:56 (quit) corruptmemory: Ping timeout: 240 seconds
19:02 (join) jonrafkind
19:14 Lajla: Sgeo, well, that is what a macro does.
19:14 Lajla: But to be hones,t I'm trying to make an example.
19:14 Lajla: But for some reason it doesn't work and I'm not sure why
19:14 Lajla: http://codepad.org/mO19cRmF
19:14 Lajla: BAsically, if you do (version 3.5) or any non integer, it starts to cry, whic is what I wanted.
19:15 Lajla: But availability flag is still unbound it says.
19:15 Lajla: But jonrafkind probably knows what's wrong.
19:15 Lajla: Because he has this omniscient click wheel he can ask quaestions
19:15 Lajla: When I check the expansion and macro stepper it looks fine
19:16 Sgeo: I'm fairly knew with Racket
19:16 Sgeo: Read about Scheme before, got discouraged by all the implementations and inconsistencies between such
19:17 Sgeo: s/knew/new/
19:18 Lajla: Sgeo, R6RS actually does better at unifying it.
19:18 Lajla: the main thing is that there are multiple scheme standards.
19:18 Lajla: And an implementation can like claim to support R5, R6, and SRFI 1 through 30
19:18 jonrafkind: it works if you use #'vers instead of #'_
19:18 jonrafkind: hm, im not exactly sure why
19:19 jonrafkind: maybe because #'_ isn't really bound
19:19 jonrafkind: also you don't need to rebind define
19:19 Lajla: jonrafkind, in the availability flag?
19:19 jonrafkind: yea
19:19 jonrafkind: also you should use syntax-local-introduce instead of datum->syntax
19:19 Lajla: jonrafkind, oh yeah, it searches for the binding in the enviroment where the syntax case lay.
19:19 Lajla: jonrafkind, the whot?
19:20 Sgeo: I kind of only skimmed the stuff about macros >.>
19:20 Lajla: This is why I don't understand syntax-case.
19:20 Sgeo: I was reading Racket stuff all night instead of sleeping
19:20 Lajla: I just use it and it usually works, it's my very own C++
19:20 Lajla: Sgeo, ahh
19:20 Lajla: well, macros happen at compile time.
19:20 Lajla: THey cannot access runtime information
19:20 Lajla: And syntax-case is pretty difficult an macro system.
19:20 Lajla: a*
19:21 Lajla: Hmm
19:21 Lajla: actually
19:21 Lajla: I should in all means not rebind define.
19:21 jonrafkind: yes if you're just starting with macros you should stick to syntax-rules
19:21 Lajla: It would like break the syntax in (let ((define +)) ...)
19:21 Lajla: Well, sytnax rules doesn't dow hat he wants.
19:21 Lajla: He wants to extract the version numer.
19:21 Lajla: number
19:22 Sgeo: extract?
19:22 Lajla: Well, feed it to the syntax transformer.
19:22 Lajla: Give it access to it.
19:22 jonrafkind: (define availability-flag (if (> version 4) #t #f)) whats wrong with that
19:22 Lajla: Sgeo, but the thing with scheme standards is that contrary to CL, in Scheme standards and ideas sort of prove their worth and fight for acceptance.
19:23 Lajla: Whereas in CL some autocratic organ seems to decide most.
19:23 jonrafkind: rudybot: eval (define-syntax-rule (make-available name flag) (define name (if (> flag 4) #t #f))
19:23 rudybot: jonrafkind: error: eval:1:0: read: expected a `)' to close `('
19:23 jonrafkind: rudybot: eval (define-syntax-rule (make-available name flag) (define name (if (> flag 4) #t #f)))
19:23 rudybot: jonrafkind: Done.
19:23 Lajla: jonrafkind, it doesn't check for exactness and integerness.
19:23 jonrafkind: rudybot: eval (begin (make-available f1 8) f1)
19:23 rudybot: jonrafkind: ; Value: #t
19:24 jonrafkind: rudybot: eval (begin (make-available f2 2) f2)
19:24 rudybot: jonrafkind: ; Value: #f
19:24 Lajla: And I also felt demonstrating just how cool syntax ase is, but I failed. =(
19:24 Lajla: Luckily the click wheel helped me out.
19:24 jonrafkind: its true it lacks the ability to do compile-time checking of values
19:24 jonrafkind: for that you could use syntax-parse
19:24 jonrafkind: syntax-case is quite poor at doing checks
19:25 jonrafkind: syntax-parse is closer to syntax-case though, im not sure if there is an analog to syntax-rules, or define-syntax-rule
19:25 Lajla: syntax case is quite unexpressive for soemthing hat was the standard so long I guess.
19:25 Lajla: I wonder who invented it.
19:25 jonrafkind: no its quite expressive
19:25 jonrafkind: its just a pain to use
19:25 Lajla: Ehh
19:25 Lajla: syntax-rules.
19:25 Lajla: pardon
19:25 jonrafkind: yea
19:26 (quit) stis: Remote host closed the connection
19:26 Lajla: But say, jonrafkind what if you were like 'let's create a hygienic macro system!'
19:26 Lajla: Would you then invent syntax-rules?
19:26 jonrafkind: its a good starting point
19:26 Lajla: Wouldn't you like when it was done kind of had the idea 'This is unfinished work'
19:26 Lajla: But like, it couldn't even insert ... in expansions in the first stages.
19:26 Lajla: I mean, it's kind of a big hack, how it worked.
19:27 Lajla: Especially its treatment of the symbol ...
19:27 jonrafkind: i mean its almost trivial once you understand the parts, transcription + hygiene
19:27 Lajla: I'm stlil not quite sure I like how syntax-case uses it. One could argue that it is the 'Scheme way' to let peope choose their epilipses.
19:28 Lajla: (syntax-case x ... ()
19:28 Lajla: jonrafkind, but like, would you really make it like that?
19:29 Sgeo: Is Racket hotswappable at all?
19:29 Lajla: I find it kind of hard to believe what went to the person's mind who stopped at that part.
19:29 Lajla: Sgeo, wat's that?
19:29 jonrafkind: are you wondering why syntax-case has some special support for ellipses?
19:29 Lajla: jonrafkind, well, some would say that letting people choose their own symbol for that is a better practice.
19:29 jonrafkind: syntax-parse allows that
19:29 Lajla: Ahh
19:29 jonrafkind: and in racket, you can choose a different symbol for ellipses with syntax-case
19:29 Lajla: Never heard of that.
19:29 Lajla: How?
19:29 Sgeo: Lajla, changing any code of a running system while it's running, I think
19:29 jonrafkind: (require (rename-in racket/base [... my-ellipses]))
19:30 jonrafkind: its just an identifier
19:30 Lajla: As you all know, I know nothing of racket.
19:30 jonrafkind: like anything else
19:30 Lajla: Ah
19:30 Lajla: But that doesn't work either.
19:30 Lajla: I mean, per expansion
19:30 Lajla: So that you can est them inside each other.
19:30 jonrafkind: you can escape ellipses
19:30 Lajla: Without having to make really clumsy escape.
19:30 jonrafkind: (... ...) inside a template becomes ...
19:30 Lajla: Yeah, I know.
19:30 jonrafkind: yea, i sort of hate that too
19:30 jonrafkind: but such is life
19:30 Lajla: But still, wouldn't you say it would be more scheme if you could just choose it?
19:30 Lajla: Sgeo, ehh
19:30 Lajla: you mean self-modifying code?
19:31 Lajla: And yeah, to some extend.
19:31 Lajla: Lisp traditionally has a close correspondence between code and data.
19:31 Lajla: The line between code and the data it operates on is somewhat opaque.
19:31 Sgeo: Lajla, not... quite self-modifying, although I guess with some work, that can enable it. Programmer-modifyable, while the program's running
19:31 jonrafkind: if you used eval you could do it
19:31 Lajla: Sgeo, you mean with I/O?
19:32 Lajla: jonrafkind, well, apply is also a light example.
19:32 Lajla: Sgeo, are you familiar with eval in most languages?
19:32 Sgeo: I mean, like Smalltalk or Erlang
19:32 Sgeo: Lajla, yes
19:32 Lajla: Because in Lisp, it is slightly different.
19:32 Lajla: Well, in Lisps, eval does not evaluate strings.
19:32 Lajla: But data.
19:32 Lajla: (+ 1 2) is a scheme program, it's also a list of one symbol and two numbers as a data structure.
19:33 Lajla: Scheme programs are data. So (eval (list '+ 1 2)) considers that datastructure as a program.
19:33 Sgeo: That's great. Anyone built a system that, if I have a long running Racket program, I can make a change without stopping it?
19:33 Sgeo: eval would probably be an important party
19:34 Lajla: You can do funky stuff like (eval (list (if some-condition '+ '*) 1 2))
19:34 Lajla: Sgeo, overuse of eval is considered bad practice though.
19:36 Sgeo: Suppose, in a program, I (define (double x) (+ x x x))
19:36 Sgeo: And I'm running some program
19:36 Sgeo: Then realize the mistake, and want to fix it without stopping the program
19:37 Lajla: Sgeo, that is not soemthing anyone would recommend doing with eval.
19:37 Lajla: Just use (set! double johnrafkind)
19:37 Lajla: And even set! is frowned upon.
19:37 Lajla: In SCheme, we don't make mistakes.
19:37 Sgeo: Lajla, and can I do that to a running program?
19:37 Lajla: You can have a running program that does that.
19:38 Lajla: Or you can make a program that accepts untrusted input and considers it code.
19:38 Lajla: You can make a program that has an input which takes code and modifies some parts of itself yeah
19:38 Sgeo: Lajla, is there some library or framework in Racket to make that easy?
19:39 Lajla: Sgeo, I am not sure.
19:39 Lajla: Most people would tell you that that is generally not what you want to do.
19:39 Sgeo: In Newspeak, there was an issue in the IDE that I wanted to fix. The code of the IDE was running, I went and fixed it, and saw the change immediately
19:40 Lajla: Sgeo, that would be an interpreter and not a compiler then I guess.
19:40 Lajla: That has little to do with any language.
19:40 Lajla: It's more that the interpreter supports modyfing code at runtime.
19:40 Lajla: Which is pretty dangerous
19:40 Lajla: Racket is mainly a compiler.
19:41 Lajla: You produce an executable.
19:42 Sgeo: Can my code at least load or reload modules dynamically?
19:43 Lajla: I'm not sure.
19:43 Lajla: Depends on how hey are compiled I suppose.
19:43 Lajla: I see no reason why dynamic linking should not be able o happen
19:44 Sgeo: I appear to have caused an infinite loop of crashiness in DrRacket
19:44 Lajla: How?
19:45 Sgeo: Tried to insert a fraction right after #lag racket without any space between fraction and racket
19:45 Sgeo: Exact fraction was 5 1/4
19:45 Sgeo: Though I don't think that matters
19:48 (quit) Sgeo: Read error: Connection reset by peer
19:49 (join) Sgeo
19:49 Sgeo: Anything happen after <Sgeo> Though I don't think that matters
19:49 Sgeo: ?
19:49 jonrafkind: no
20:07 (quit) masm: Quit: Leaving.
20:16 Sgeo: Do most libraries that expose objects leave (send) as something the client has to do?
20:17 jonrafkind: yea I think so
20:17 jonrafkind: which libraries are you looking at?
20:17 Sgeo: Just wondering in general
20:17 jonrafkind: anything that uses classes will most likely use `send'
20:45 Sgeo: Does the graphical stuff like XML boxes actually see use?
20:47 (join) dnolen
20:47 jonrafkind: no idea
20:47 jonrafkind: a lot of the stuff in drracket is designed to use in the classroom
20:48 Lajla: jonrafkind, I got it.
20:48 Lajla: _ and ... are special identifiers.
20:48 Lajla: So you can't just use #' on it
20:48 jonrafkind: well ... is defined in some module by `define-syntax'
20:49 jonrafkind: well yes they are special to the expander
20:49 Lajla: I mean that my define syntax didn't work because I used #'_
20:49 jonrafkind: and (syntax blah) invokes the expander on blah
20:49 jonrafkind: yea
20:49 Lajla: I always thought it was just a dummy used for the first to say 'I don
20:49 Lajla: care'
20:49 jonrafkind: im not sure exactly what the expander does with _
20:49 jonrafkind: i thought it usually threw an error or something
20:51 Lajla: Sgeo: http://codepad.org/5vN0MXsd
20:52 Lajla: This one owns.
20:52 Lajla: It works
20:53 Sgeo: What's that k?
20:53 Sgeo: I don't remember much about syntax-case
20:53 Sgeo: Or anything, actually
20:53 Lajla: Sgeo, k can basically be anything except ... or _
20:53 Lajla: Which have speciual meaning
20:53 Lajla: You can replace the k with VOTE-PALIN-2012
20:54 Lajla: And it will go fine
20:54 Lajla: As long as you do it everywhere
21:43 (quit) Demosthenes: Quit: leaving
21:53 (quit) parcs: Ping timeout: 240 seconds
21:55 (join) parcs
22:07 dnolen: in Racket what is the equivalent of range where (range 10) -> (0 1 2 3 4 5 6 7 8 9) ?
22:08 jonrafkind: (in-range 0 10)
22:09 jonrafkind: is a sequence, not a list though
22:09 dnolen: how do I get a list from that ?
22:10 jonrafkind: rudybot: (stream->list (in-range 0 10))
22:10 rudybot: jonrafkind: error: reference to undefined identifier: stream->list
22:10 jonrafkind: cmon..
22:10 dnolen: nice
22:10 jonrafkind: rudybot: (require racket/stream) (stream->list (in-range 0 10))
22:10 rudybot: jonrafkind: error: default-load-handler: cannot open input file: "/usr/local/plt/collects/racket/stream.rkt" (No such file or directory; errno=2)
22:11 jonrafkind: well if rudybot wasn't so old you could use stream->list
22:12 Sgeo: So streams are like Haskell lazy lists?
22:12 Sgeo: s/So/Are/
22:13 jonrafkind: yea
22:13 Sgeo: :D
22:13 jonrafkind: you can make an infinitely long straem and just get parts of it that you want
22:13 jonrafkind: there is an entire lazy language in racket if you want real laziness
22:13 jonrafkind: #lang racket/lazy I think
22:13 Sgeo: in #lang racket, do you need to use a special map?
22:14 jonrafkind: to map over streams?
22:14 Sgeo: yes
22:14 Sgeo: Lazily
22:14 jonrafkind: yea, http://docs.racket-lang.org/reference/sequences.html?q=stream#(def._((lib._racket/stream..rkt)._stream-map))
22:15 jonrafkind: most constructs in racket are eager by default
22:16 Sgeo: Can I make my own non-list thing that works with regular map?
22:16 jonrafkind: I guess not, map uses cdr to get the next item in your list
22:16 jonrafkind: ive read that clojure has some ISeq interface that any data structure can implement
22:17 Sgeo: And how does cdr work?
22:17 jonrafkind: but racket has no such thing
22:17 Sgeo: Oh, n/m
22:17 Sgeo: Misread you
22:17 jonrafkind: cdr operates on the primitive datatype, pairs
22:17 jonrafkind: you can always throw out map entirely and just use sequences
22:18 jonrafkind: you can use the `for' forms to iterate over them
22:18 jonrafkind: list comprehensions is their name
22:19 (join) Demosthenes
22:22 Sgeo: I guess I was more interested in the general question of functions that don't care about the underlying implementation
23:08 (join) evhan
23:29 (part) evhan
23:56 (join) offby1