00:40 (join) samth_
00:45 (join) mithos28
00:49 (join) misterm
01:11 (quit) mithos28: Quit: mithos28
01:12 (quit) samth_: Ping timeout: 264 seconds
01:14 (join) coyotama|2
01:15 (quit) coyo: Ping timeout: 260 seconds
01:33 (quit) dnolen: Quit: dnolen
03:16 (join) lucian
03:24 (quit) lucian: Remote host closed the connection
03:27 (quit) tfb: Quit: sleeping
04:42 (quit) coyotama|2: Ping timeout: 240 seconds
04:51 (join) tfb
04:59 (quit) abbe: Ping timeout: 276 seconds
05:35 (join) jesusito
05:40 (join) cderici
05:41 (quit) cderici: Client Quit
05:48 (join) tauntaun
05:55 (join) Agari
06:12 bremner: it would be nice to have checksums published for the racket tarballs
06:13 bremner: or better, gpg signatures
06:13 (quit) evhan: Ping timeout: 246 seconds
06:32 (quit) tauntaun: Remote host closed the connection
06:34 (quit) realitygrill: Quit: realitygrill
06:48 (quit) jesusito: Quit: ERC Version 5.3 (IRC client for Emacs)
07:05 (join) mye
07:10 mye: I'm trying to find the definitions of core forms, say 'or': what do I have to grep for? I tried variations of: grep -R '.*(syntax.*[ (]or$' .; sind I presume it's a macro of some kind.
07:11 mye: (grepping from the collects folder)
07:11 rapacity: have you tried using git's grep ?
07:11 mye: no
07:11 (quit) Agari: Ping timeout: 240 seconds
07:12 mye: also, this is in the normal windows binary installation folder
07:12 mye: not a git repo
07:13 rapacity: or
07:13 rapacity: oops
07:13 mye: I tried searching the tags file, and I jumped to (or ... with Vim but that goes to some kind of Or struct definition
07:14 mye: ctags -R . seems to need some modification to what ctags recognizes as definitions
07:16 (join) Snaffu
07:23 (join) tauntaun
07:48 (join) realitygrill
07:50 (quit) realitygrill: Client Quit
07:56 (quit) tauntaun: Ping timeout: 246 seconds
08:12 (join) masm
08:20 (join) MayDaniel
08:29 (quit) MayDaniel: Read error: Connection reset by peer
08:58 (join) realitygrill
09:53 (join) cpach
10:01 (part) shofetim: "ERC Version 5.3 (IRC client for Emacs)"
10:04 (join) mithos28
10:06 (join) dnolen
10:13 (quit) mithos28: Quit: mithos28
10:33 (join) abbe
10:36 (join) tauntaun
10:36 (join) evhan
10:42 (join) coyotama|2
10:42 (join) MayDaniel
10:43 (nick) coyotama|2 -> coyo
10:43 (quit) coyo: Changing host
10:43 (join) coyo
10:58 (join) SinDoc
11:06 (quit) dnolen: Quit: dnolen
11:34 (join) samth_
11:40 (join) corruptmemory
11:46 (join) mceier
12:00 (quit) masm: Ping timeout: 260 seconds
12:03 (quit) MayDaniel: Read error: Connection reset by peer
12:08 (quit) realitygrill: Quit: realitygrill
12:18 (join) jfalcon
12:19 (join) carleastlund
12:19 (quit) mye: Ping timeout: 240 seconds
12:20 (join) anRch
12:23 (join) jonrafkind
12:25 (join) masm
12:33 (join) mithos28
12:36 (quit) SinDoc: Ping timeout: 264 seconds
12:43 (join) realitygrill
12:51 (quit) anRch: Quit: anRch
12:53 (part) jfalcon
13:05 (join) SinDoc
13:18 (quit) tauntaun: Ping timeout: 255 seconds
13:21 (join) lucian
13:22 (join) MayDaniel
13:22 (part) SinDoc
13:25 (join) kenjin2201
13:29 (join) PLT_Notify
13:29 PLT_Notify: racket: master Robby Findler * 7cc3465 (1 files in 1 dirs): clarify the style<%>'s get-delta's docs
13:29 PLT_Notify: racket: master Robby Findler * 87e637a (4 files in 4 dirs): - adjust the teaching language test coverage style implementation ...
13:29 PLT_Notify: racket: master Robby Findler * 3cb3b9f (1 files in 1 dirs): fix a bug in PLTDRPAR
13:29 PLT_Notify: racket: master commits ca5c061...3cb3b9f - http://bit.ly/ggJNxi
13:29 (part) PLT_Notify
13:31 bremner: eli: I (and the debian ftp-masters) have a few questions about some of the licenses in files in the racket distribution. Should I file a ticket, send to the devel list, or something else?
13:32 (quit) samth_: Ping timeout: 276 seconds
13:34 (quit) MayDaniel: 
13:35 (quit) realitygrill: Quit: realitygrill
13:38 (quit) tfb: Ping timeout: 240 seconds
13:55 (join) Agari
14:03 (join) realitygrill
14:05 (join) MayDaniel
14:10 (join) anRch
14:13 (join) Fare
14:23 jeapostrophe: for some reason "open" doesn't work on one of my macs with drracket
14:24 jeapostrophe: does anyone have any idea why that would be?
14:24 jeapostrophe: it focuses drracket but it doesn't load the file
14:24 clklein: I see that intermittently.
14:25 clklein: Running open a second time always (or almost always) works though
14:25 jeapostrophe: i think mine used to do that but it doesn't work at all now
14:25 clklein: oh, I am several days behind HEAD
14:37 jeapostrophe: naturally when i just updated it works great
14:38 jeapostrophe: weird
14:38 clklein: DrR knows that the next step after "git pull" is "Submit PR" :)
14:39 jeapostrophe: i think the updated queue handling for the repl bug fix did it
14:40 clklein: I suppose that's more likely.
14:40 jeapostrophe: :P
14:43 (join) tauntaun
14:46 (quit) evhan: Quit: Lost terminal
14:48 (quit) mithos28: Quit: mithos28
14:50 (join) evhan
14:51 (join) mithos28
14:53 (join) PLT_Notify
14:53 PLT_Notify: racket: master Robby Findler * 1cf98d3 (1 files in 1 dirs): fix when check syntax is clearing out its error message editor ... - http://bit.ly/hYHXO1
14:53 (part) PLT_Notify
14:54 (quit) mithos28: Client Quit
14:54 (join) saint_cypher
14:57 (join) mithos28
15:06 (quit) mithos28: Quit: mithos28
15:10 (join) mithos28
15:11 (quit) Agari: Ping timeout: 240 seconds
15:12 (join) mithos28_
15:12 (quit) mithos28: Read error: Connection reset by peer
15:12 (nick) mithos28_ -> mithos28
15:28 (join) Agari
15:32 (quit) anRch: Quit: anRch
15:44 (quit) kenjin2201: Ping timeout: 276 seconds
15:52 (join) Slipyx
15:53 (join) samth_
16:06 (join) kenjin2201
16:09 (quit) realitygrill: Quit: realitygrill
16:28 (quit) samth_: Ping timeout: 246 seconds
16:43 (join) jfalcon
16:47 (quit) rapacity: Ping timeout: 260 seconds
16:48 (join) rapacity
16:48 (quit) rapacity: Changing host
16:48 (join) rapacity
17:03 (quit) MayDaniel: Read error: Connection reset by peer
17:05 (quit) coyo: Quit: KVIrc 4.0.2 Insomnia http://www.kvirc.net/
17:06 (join) samth_
17:09 (quit) Snaffu: Quit: Leaving
17:18 (join) Hydrant
17:19 Hydrant: hey all.. I'm still learning racket (from years of C++) and I have some beginner questions as I go through the language :-)
17:19 Hydrant: what is the difference between define and let?  Is it that let is locally scoped and returns its value, whereas define is a global definition ?
17:25 jonrafkind: yea basically
17:26 Lajla: Hydrant, you can also locally use define
17:26 Lajla: In fact, inside a let
17:26 (quit) kenjin2201: Ping timeout: 240 seconds
17:26 Lajla: rudybot, (let ((a 3)) (define b a) b)
17:26 rudybot: Lajla: your sandbox is ready
17:26 rudybot: Lajla: ; Value: 3
17:27 Lajla: the definition of b with define is local to the let block
17:27 Hydrant: what is the real difference between let and define?  It seems that racket has a lot of redundancy... let, let*... define, and it's not clear to me how these are different or why they are
17:27 Lajla: anything outside it can't see it.
17:27 Lajla: Hydrant, well, define allows for mutual recursion.
17:28 Lajla: As a common example: (define (even? n) (or (zero? n) (odd? (- (abs n) 1)))) (define (odd? n) (even? (- (abs n) 1))))
17:28 Lajla: If you use let, they won't 'see' each other.
17:29 Hydrant: o let is basically a locally scoped thing
17:29 Hydrant: *so
17:29 Lajla: (let ((x <expression>)) ...) basically <expression> is in the scope outside of the let.
17:29 Lajla: It won't see any bindings you make _with_ the let.
17:30 Lajla: Also, let also immediately constructs a block
17:30 Lajla: So any new binding you make is local to the block
17:30 Lajla: which is extremely common in racket to use, often up to very deep.
17:30 Lajla: 'variables' are typically not used to be variable, but rather just shorthands for more complex expressions to keep the code more readable.
17:31 Hydrant: (let ((x <expression>)) ...) <-- can you clarify?
17:31 Hydrant: you mean that <expression> can use things defined outside of the let, right ?
17:31 Lajla: Hydrant, yeah
17:31 Hydrant: but outside scope can't see inside the let
17:31 Lajla: and not things defined inside the let.
17:31 Hydrant: okay
17:32 (part) Slipyx: "Leaving"
17:32 Lajla: Yeah
17:32 Lajla: Like, if you use (let ((x <expression>) (y <expression>)) <body>)
17:32 Lajla: the two bindings, x and y are said to be visible in <body>
17:32 Lajla: But they are not visible in both <expression>
17:32 Lajla: So it doesn't allow for recursive definitions.
17:32 Hydrant: right
17:32 Hydrant: which is what let* is for
17:32 Hydrant: but why?
17:32 Hydrant: it seems to complicate the language
17:32 Lajla: Well, no, let* does not allow recursive definitions either
17:33 Hydrant: I thought let* would allow x to be visible to y, and y to be visible to x
17:33 Lajla: If you have (let* ((x <expression-1>) (y <expression-2>)) <body>)  then x is visible in <expression-2>, but not the reverse.
17:33 carleastlund: letrec makes them all visible to each other; let* makes them all visible below but not above themselves.
17:34 Lajla: No, (let* ((x <foo>) (y <bar>)) <baz>) is just sugar for (let ((x <foo>)) (let ((y <bar>)) <baz>))
17:34 carleastlund: We have them all because some programs call for one, and some for the other.  And they're essential for being able to write different kinds of macros.
17:34 Hydrant: okay.... does anyone else find it odd that there is let, let* and letrec rather than just the most powerful one being used as the norm ?
17:34 carleastlund: Hydrant, none of them are "most powerful".
17:34 tauntaun: Hydrant: you should not be thinking about let* and letrec until you master more basic things.
17:34 Hydrant: it would seem letrec would be best, because it appears to do the same as let* and let
17:35 Hydrant: what more basic things are there to master ?
17:35 carleastlund: You can write a let* with five different bindings of X, for instance, but not a let or letrec.  You can build recursion with letrec, but not with let or let*.  You can swap x and y's bindings with let, but not with let* and letrec.
17:35 Agari: It might be helpful to understand let in terms of an abstraction over lambda too.
17:35 tauntaun: Hydrant: exactly how are you teaching yourself Scheme? (Which book(s)?)
17:35 jonrafkind: i recommend not using let/let* or anything like that, just use define
17:35 Hydrant: tauntaun: a die-hard professor that believes the racket guide is the best way to learn ;-)
17:35 tauntaun: and the name of this professor is...?
17:36 Lajla: Hydrant, well, the point about letrec is that you some-times want to locally use the same name as you use one globally.
17:36 Agari: As in: (let ((x 20)) (* x x)) vs ((lambda (x) (* x x)) 20)
17:36 Lajla: For instance, say you have a variable the-list in your environment and you use (let ((the-list (cdr the-list)) ...)
17:36 Lajla: If you use letrec, this would fail.
17:36 (quit) mithos28: Quit: mithos28
17:37 Hydrant: so let will hide variables with new scope
17:37 carleastlund: I think the main answer, though, is that Racket has a huge amount of tools because sometimes, you need just the right tool for the job.  99% of the time?  I use define for recursion at the top level of a module, and let* for pretty much all other bindings.  I rarely use explicit let or letrec.  Just because all the options are there, doesn't mean you should be using them all, all the time.
17:37 tauntaun: hide??
17:38 tauntaun: Hydrant: it's simple.  let creates bindings, roughly like declaring and defining local variables in C/C++ in a curly-braced block.
17:38 carleastlund: shadow
17:38 tauntaun: As for let* and letrec, don't think about them until you understand let thoroughly.
17:38 Lajla: That's a good thing to do for now yeah.
17:39 Lajla: Basically, you shouldn't use letrec or let* if you don't have a reason as they are more expensive.
17:39 tauntaun: Lajla: exactly how is let* more expensive than let?
17:39 carleastlund: Ex... expensive?
17:40 tauntaun: (let* just gets macro-rewritten into nested lets, as you probably know...)
17:40 Lajla: Yeah
17:40 Lajla: So you have two lets instead of one.
17:40 Lajla: Have to create two environments
17:40 carleastlund: Untrue.  Neither let nor let* are primitive in Racket.
17:40 Lajla: don't know if the compiler optimizes for this.
17:40 tauntaun: carleastlund: yes, they in turn get macro-rewritten into lambda applications.
17:40 carleastlund: Please do not give out blanket advice on which macros to use based on unfounded assumptions about their expansion.
17:41 carleastlund: tauntaun: also false, neither's expansion involves lambda.
17:41 tauntaun: what?
17:41 Lajla: Yeah, but like, (let ((a b) (c d)) ...) becomes ((lambda (a c) ....) c d), which I reckon is cheaper than a double application
17:41 tauntaun: carleastlund: okay, I believe in Racket they expand to letrec*
17:41 carleastlund: Nope.
17:41 tauntaun: or let-values
17:41 carleastlund: Yes.
17:41 tauntaun: something like that
17:41 tauntaun: ok fine.
17:41 Lajla: And let-values is a primiive?
17:41 carleastlund: But more importantly?  Don't give advice about "don't use this macro because of its expansion" as a programming tip.  It's like the last thing one should ever worry about.
17:42 tauntaun: argh!  I never did that!
17:42 eli: bremner: You can just mail me.  (Or ask here, etc.)
17:42 tauntaun: Please read my words carefully.
17:42 carleastlund: It's like saying "don't drive that car because it has fuzzy dice in the window, the mass of which may hurt fuel economy".
17:42 carleastlund: You didn't.  That part was to Lajla.
17:43 (quit) samth_: Ping timeout: 240 seconds
17:43 jonrafkind: bremner, i am at a loss for words as to 5.0.2 being rejected in debian
17:44 eli: 5.0.2 got rejected?
17:44 eli: Why?
17:45 jonrafkind: because of the licensing stuff that debian requires
17:45 jonrafkind: im just reading james vega's email
17:46 Hydrant: this might be a simple question... but I can't find it in the docs... what does _ represent ?  I saw some code today like (lambda (_)...)
17:46 jonrafkind: its just a convention meaning 'i dont care about this argument'
17:46 Hydrant: as I recall ML has something like that to indicate "I don't care"
17:47 Hydrant: that expands to one argument, or many?
17:47 Fare: one
17:47 Hydrant: okay, thanks a lot :-)
17:47 Fare: unless it's (lambda _ ...)
17:47 carleastlund: It's a regular argument.  You can use it just like x or whatever, but people tend to use _ when they aren't going to refer to it.
17:47 Hydrant: you're helping to fill in gaps in my mind from documentation and lectures ;-)
17:47 eli: jonrafkind: What was the licensing issue?
17:47 Hydrant: soon I will be a racket jedi ;-)
17:47 jonrafkind: no idea
17:48 jonrafkind: but i mean who cares.. we have a ppa
17:48 Fare: eli: maybe what I need is (1) a parser and (2) a system of tree rewrite rules
17:48 Lajla: carleastlund, that is true.
17:48 Lajla: I require copious amounts of spanking for my crimes.
17:48 Fare: (3) a pretty printer
17:49 Lajla: I still think using let* or letrec without a reason is some-what unhealthy.
17:49 carleastlund: I agree that sticking to one, most of the time, is a good idea.  I prefer let*; I find myself needing it more often, and rarely do I need let in specific.
17:50 eli: jonrafkind: Well, it would be much more convenient if it was in the official repo.
17:50 eli: But I'm really annoyed at their whole thing with the licensing.
17:50 eli: They get really anal about it.
17:50 jonrafkind: yes
17:50 (quit) Fare: Quit: Leaving
17:51 jonrafkind: i think its appualing how close to their own specifications they adhere to, and the fact that they have such specifications at all
17:51 jonrafkind: it sounds like a complete waste of time to me
17:51 eli: They even had an issue in the past about us distributing R5RS -- I had to get an OK from from Will and Kent.  (Clarifying that they really intend it to be public, or something like that.)
17:51 jonrafkind: lulz
17:52 Hydrant: Debian being anal about licensing is a very important thing... despite how annoying they can be
17:53 carleastlund: Yeah, hilarious that they insist on being careful about not violating the law, and getting explicit permission for intellectual property.  Hilarious that they should not be intimately familiar with Scheme standards.
17:53 eli: Hydrant: It's important to get a completely free distro, but it's extremely annoying for a project like this where the project is obviously trying to do its best to be as open as possible.
17:53 carleastlund: How should it be obvious to them?
17:53 eli: The R5RS example was this sentence: "We intend this report to belong to the entire Scheme community, and so we grant permission to copy it in whole or in part without fee."
17:54 tauntaun: also in R6RS
17:54 carleastlund: If their concern is an ability to hold up in court, I can see why they'd want that clarified.  The terms "entire Scheme community" seem shaky.
17:55 carleastlund: Also it says copy, and not (re)distribute.
17:55 tauntaun: ?? out of curiosity, where is the ambiguity in "entire Scheme community"?
17:55 eli: It is still annoying.
17:55 carleastlund: Do the debian package repository maintainers qualify as part of the Scheme community?
17:56 carleastlund: Is the permission granted to just the Scheme community, or to anyone?
17:56 carleastlund: etc., etc.
17:56 eli: bremner: Feel free to send licensing questions to carleastlund here who just volunteered by not being annoyed.
17:56 tauntaun: the *belonging* is to the entire Scheme community; the *copying* is for everyone.
17:56 carleastlund: eli, yes, I volunteer to not be annoyed by the questions.  I also assert that I won't know any of the answers.
17:57 eli: In that case I reserve the right to be annoyed, and to be even more annoyed at spending more time when I'm expected to justify my "annoyement".
17:58 carleastlund: Pardon me.  I inferred from your statement about annoyance that you thought they were unjustified in their practices.  Apparently I was in error, and you meant no more than you said.
17:59 eli: Not more.  People can be justified or think that they're justified or whatever.
18:00 eli: I could always raise the point of the difficulty of suing someone over something like the r5rs wording or any number of such similarly vague licenses, which gained a v2.0 as a result -- but that only contributes to the mess and forces me to spend more time on something that I really don't want to know anything about.
18:00 eli: (DWTFYW is a nice example of such a license.)
18:05 (nick) tauntaun -> tautaun_away_to_
18:08 (join) MayDaniel
18:09 Hydrant: can someone suggest a good book or reference for learning about what exactly S-expressions are?  None of my usual CS books seem to have it... a textbook on compilers maybe?
18:11 carleastlund: Hydrant, an s-expression is just something written with symbols (basically variable names) and parenthesis-delimited lists.  So banana is an s-expression.  (apple (banana (pear))) is an s-expression.  (apple banana pear) is an s-expression.  Depending on context, other elements like strings and numbers may be allowed.  That's all there is to it.
18:11 samth: Hydrant, http://en.wikipedia.org/wiki/S-expression might be a useful reference
18:13 Hydrant: yah, I looked at wikipedia and it has no references and is completely unclear
18:14 Hydrant: parenthesis-delimited lists is all I need... thx
18:15 (join) kisom_dev
18:20 (quit) corruptmemory: Ping timeout: 240 seconds
18:21 (nick) tautaun_away_to_ -> tauntaun
18:22 (quit) Hydrant: Remote host closed the connection
18:23 (quit) mceier: Quit: leaving
18:27 tauntaun: Funny.  Wikipedia says sexp stands for "symbolic expression" whereas I always thought it meant "syntax expression."  I assume I was wrong the whole time...
18:34 (join) stamourv
18:34 (quit) stamourv: Remote host closed the connection
18:34 (join) stamourv
18:44 (join) samth_
18:52 (quit) samth_: Ping timeout: 240 seconds
18:53 (nick) samth -> samth_away
19:04 (quit) jfalcon: Quit: jfalcon
19:05 (join) mithos28
19:06 (join) PLT_Notify
19:06 PLT_Notify: racket: master Kevin Tew * ad0c22a (1 files in 1 dirs): parallel-do refactor
19:06 PLT_Notify: racket: master Kevin Tew * bdbb9fe (1 files in 1 dirs): Parallel Places Build
19:06 PLT_Notify: racket: master commits 1cf98d3...bdbb9fe - http://bit.ly/eqZjhJ
19:06 (part) PLT_Notify
19:09 (join) jao
19:09 (quit) jao: Changing host
19:09 (join) jao
19:10 (quit) MayDaniel: Read error: Connection reset by peer
19:15 (quit) jao: Read error: Connection reset by peer
19:21 (quit) mithos28: Quit: mithos28
19:22 em: any plans to get racket into gentoo?
19:25 (quit) lucian: Remote host closed the connection
19:25 (quit) Agari: Quit: Quit
19:26 (join) lucian
19:28 jonrafkind: em, fwiw, plt scheme 4.2.2 is in there
19:29 jonrafkind: http://gpo.zugaina.org/dev-scheme/racket/Bugs seems you can use this overlay to get 5.0
19:30 Lajla: What does fwiw stand for?
19:30 Lajla: For what it's worth?
19:31 jonrafkind: yea
19:42 em: jonrafkind: is PLT Scheme 4.2.2 very similar to racket?
19:42 jonrafkind: its made by the same people, racket is the rename of plt scheme
19:42 jonrafkind: 4.2.2 is about a year old now
19:42 em: Yeah I know that, just wondered if Racket has changed much since then.
19:42 (join) mithos28
19:43 jonrafkind: 4.2.2 -> 5.0 is a minor change more or less, the major change is 5.0 -> 5.1
19:43 jonrafkind: it would have been nice for racket and the gui rewrite to happen in the same version change, but alas.. twas not to be
19:43 (quit) masm: Ping timeout: 255 seconds
19:44 jonrafkind: i guess the only real major changes in 4.2.2 -> 5.0 are name changes, #lang scheme -> #lang racket, and stuff like that. mzc -> raco
19:47 (quit) mithos28: Quit: mithos28
19:48 (join) masm
19:49 (join) mithos28
19:52 (quit) masm: Ping timeout: 240 seconds
20:01 (quit) carleastlund: Quit: carleastlund
20:02 (quit) mithos28: Quit: mithos28
20:04 (join) mithos28
20:13 (join) masm
20:16 (quit) mithos28: Quit: mithos28
20:32 (join) realitygrill
20:35 (quit) jonrafkind: Ping timeout: 276 seconds
20:53 (join) mithos28
20:55 (join) Hydrant
21:04 tauntaun: eli: sorry for the typo.
21:05 (quit) mithos28: Quit: mithos28
21:07 (quit) lucian: Remote host closed the connection
21:37 (quit) tauntaun: Quit: Ex-Chat
21:44 Hydrant: another question... how do I accumulate results as things get processed?
21:45 Hydrant: i.e. if I want to walk a tree, and perform actions recursively as I walk it (like XSLT)
21:45 Hydrant: do I just do string-append and pass around a result.... or just return the appending I guess.. or perhaps a global variable?
21:46 Hydrant: if there is a doc which better answers my question, feel free to let me know ;-)
21:46 (quit) realitygrill: Quit: realitygrill
22:04 (join) jfalcon
22:10 (quit) jfalcon: Quit: jfalcon
22:13 (quit) jmillikan: Remote host closed the connection
22:22 (join) realitygrill
22:23 (join) dnolen
22:29 Hydrant: I'm a bit confused on what the . operator actually does... the guide doesn't describe it well... any other references ?
22:30 Lajla: Hydrant, you mean (x . y) ?
22:30 Hydrant: yes
22:30 Lajla: reader notation for a pair, basically.
22:31 Lajla: It just happens that (x y z) is actually (x . (y . (z . ()))) in disguise
22:31 Hydrant: (+ 1 . (2)) this in particular confuses me
22:31 Lajla: THis is how lists are built up.
22:31 Lajla: Same as (+ 1 2)
22:31 Hydrant: okay
22:31 Lajla: Hydrant, are you familar with the structure of linked lists?
22:31 Hydrant: '(0 1 . 2)
22:31 Hydrant: that doens't make sense to me though
22:32 Hydrant: yes, I'm an experienced C++ guru... just new to racket ;-)
22:32 Lajla: Well, that is just a shorthant for '(0 . (1 . 2))
22:32 Lajla: Ahhh
22:32 (join) jfalcon
22:32 Lajla: Well, (1 2 3 4) is a linked list.
22:32 Lajla: So it's just a nice sugar for (1 . (2 . (3 . (4 . ()))))
22:32 Hydrant: the guide shows: (cons 0 (cons 1 2)) giving '(0 1 . 2) which I can't parse
22:32 Hydrant: it says this is not a list
22:32 Lajla: And (x y . z) is just for (x . (y . z)) (called an 'improper list')
22:32 Hydrant: right
22:33 Hydrant: which, interpreting as linked lists means no NULL terminator ?
22:33 Lajla: Yeah
22:33 Lajla: You generally want that
22:33 Hydrant: why no segfault then
22:33 Lajla: So you want to do (cons 0 (cons 1 (cons 2 '())))) basically.
22:33 Lajla: Because you can create improper lists.
22:33 Lajla: Also, racket won't really segfault on you.
22:34 Hydrant: right, but the underlying question was addressing the linked list metaphore
22:34 Hydrant: which means there is other info than just the linked lsit
22:34 Hydrant: ahh... can't type
22:34 bremner: Hydrant: (cons a b) makes a pair
22:34 bremner: Hydrant: if b is a list, so is (cons a b)
22:34 Lajla: Hydrant, well, racket has a latent type system.
22:35 Lajla: So it will only cry when you pass an improper list to a function which expects a proper one.
22:35 Hydrant: such a strange world
22:35 Lajla: rudybot, (map sqrt '(1 2 3 4 . 5))
22:35 rudybot: Lajla: error: map: expects type <proper list> as 2nd argument, given: (1 2 3 4 . 5); other arguments were: #<procedure:sqrt>
22:35 Hydrant: can I think of . as "list append" ?
22:36 Lajla: rudybot, (if #f (map sqrt '(1 2 3 4 . 5)) "everything went okay")
22:36 rudybot: Lajla: ; Value: "everything went okay"
22:36 Lajla: Hydrant, cons?
22:36 Hydrant: . is cons
22:36 Lajla: Ehhh, yes and no I suppose.
22:36 Lajla: rudybot, (1 . 2)
22:36 rudybot: Lajla: error: eval:1:0: application: bad syntax in: (1 . 2)
22:36 Lajla: rudybot, (cons 1 2)
22:36 rudybot: Lajla: ; Value: (1 . 2)
22:37 (quit) jfalcon: Quit: jfalcon
22:39 Hydrant: umm... can I just safely ignore the . until I get more advanced?
22:40 Hydrant: I seem to get more confused the deeper I try to understand things
22:41 Lajla: Depends on what you call 'advanced'
22:41 Lajla: It's not an operator by the way.
22:41 Lajla: (x . y) is a pair of two values.
22:41 Lajla: A list is just a nesting of pairs terminated with ()
22:41 Lajla: Which you can look at like null.
22:42 Hydrant: okay... so a pair of lists would look like '(1, 2, 3) . '(4, 5, 6) ?
22:42 Lajla: Now where did you get that comma.
22:43 Lajla: And it would look like ((1 2 3) . (4 5 6))
22:43 Hydrant: habit :-)
22:43 Lajla: brackets cannot be dropped or added.
22:43 Lajla: x is different from (x)
22:43 Lajla: x is just ... x, (x) is a list containing only x.
22:43 Hydrant: right... (x) executes x whereas x is the value right
22:43 Hydrant: oh
22:43 Lajla: Well (x) is a call to x with no arguments considered as code.
22:44 Lajla: Hydrant, have you seen the term 'homo-iconic' come by?
22:44 Lajla: Or 'code as data'?
22:44 Hydrant: code as data yes
22:44 Hydrant: homo-iconic no
22:44 Lajla: Well, the latter is just fancy schmancy greek for same.
22:44 Lajla: But as data (x) is a list containing a symbol x.
22:44 Lajla: As code, it's a function call to the function stored in the variable x with no arguments.
22:45 Lajla: as code (x . y) is nothing, it's an error.
22:45 Lajla: As data, it's a pair.
22:45 bremner: (let ((y empty)) (x.y))
22:46 Lajla: rudybot, (let ((y empty)) (x.y))
22:46 rudybot: Lajla: error: reference to undefined identifier: x.y
22:46 Hydrant: so racket requires runtime analysis to determine if something is data or code, and the operations differ depending upon runtime types right ?
22:46 Lajla: rudybot, (let ((y empty)) (x . y))
22:46 rudybot: Lajla: error: eval:1:17: application: bad syntax in: (x . y)
22:47 Lajla: Hydrant, no, not really.
22:47 Lajla: code is data
22:47 Lajla: It's just a data structure you give to your compiler and say 'compile this, mofo'
22:47 Lajla: You could see it like a description of a computer program encoded in some complex data structure.
22:48 bremner: parse trees, basically
22:48 Lajla: Like, (+ 1 2 3) this is a list of 1 symbol and 3 numbers as data
22:48 Lajla: If you give this to your implementation, it will read it and consider it a program that adds 3 numbers
22:48 Lajla: Yeah
22:48 Lajla: you basically enter your parse tree manually
22:48 Lajla: you could see it like that
22:48 Hydrant: the object I get back after compiling is callable though, right
22:49 Hydrant: so I would imagine there are different supported characteristics
22:49 Hydrant: whereas data is not callable
22:50 Lajla: Well, you basically enter a data structure in your compiler which describes a program
22:50 Lajla: and it then compiles the program it describes.
22:50 Lajla: You can also give the same data structure as data, as argument to anoher program.
22:50 Lajla: Which then analyses it and does stuff with it.
22:51 Lajla: Like, compiles it, perhaps? =P
22:53 Hydrant: how do I differentiate between the types
22:53 Hydrant: if x is callable... (x) is a call right
22:53 Hydrant: but you've also said (x) is a list if it's just data
22:53 bremner: contex
22:53 Hydrant: so this implies to me that racket 'knows' that it is data
22:53 bremner: t
22:54 Hydrant: okay
22:54 Hydrant: so I could just do ('(1 2)) and it knows "this guy is a n00b"
22:54 Hydrant: whereas (x) it has to figure out an assignment to x and it could be either
22:54 bremner: nothing so complex.
22:55 bremner: Hydrant: I suggest you go through a few simple tutorials, the philosphical things will be less puzzling then
22:55 Lajla: Hydrant, well, let's have this (let ((x (list 1 2 3)) (x))
22:56 Lajla: rudybot, (let ((x (list 1 2 3)) (x))
22:56 rudybot: Lajla: error: eval:1:0: read: expected a `)' to close `('
22:56 Hydrant: yah, I'll go back to my tutorials... the . was very unclear to me which is where I got stuck
22:56 Lajla: rudybot, (let ((x (list 1 2 3)) (x)))
22:56 rudybot: Lajla: error: eval:1:0: let: bad syntax in: (let ((x (list 1 2 3)) (x)))
22:56 Lajla: rudybot, (let ((x (list 1 2 3))) (x))
22:56 rudybot: Lajla: error: procedure application: expected procedure, given: (1 2 3) (no arguments)
22:56 Lajla: 'm usually very good at this.
22:56 Lajla: Anyway, x housed a list
22:56 Lajla: I tried to call it with no arguments
22:56 Lajla: error.
22:56 Lajla: Scheme is a 'lisp-1' as they call it, functions and values share the same namespace.
22:57 Lajla: Basically, functions are 'callable values' that are stored in variables like any other values.
22:57 bremner: rudybot, (let ((x (list 1 2 3)) x)
22:57 rudybot: bremner: (defun largest-width-in-region (start end) (interactive "_r") (save-excursion (goto-char start) (let ((column (loop while (< (point) end) maximize (- (point-at-eol) (point-at-bol)) do (forward-line 1)))) (message "Largest column is %s" column) column)))
22:57 Lajla: a variable can contain a number like 3 in one place, and a function in another place.
22:57 Lajla: wth is that.
22:57 Lajla: You missed a parentheses too.
22:58 Lajla: Hydrant, there is very little static typing going on.
22:58 Lajla: Like, you can have like (+ 3 x) where x depending on runtime info is a list or a number
22:58 Hydrant: right, I would expect it to be fully dynamic typing
22:58 Lajla: if it's a list, you've done something wrong.
23:01 Lajla: Hydrant, but basically, procedures are callable values, that's the idea, if you try to call a value that is not callable, that's a runtime error, a variable can contain a callable or uncalalble value depending on runtime information.
23:01 Lajla: You can like do (set! + 3)
23:01 Hydrant: right, Python is similar
23:01 Lajla: You've suddenly made the variable + no longer house the addition function but the number 3.
23:01 Lajla: Yeah
23:01 Lajla: But they so stole it from Scheme.
23:01 Lajla: Scheme already did this in like 1980
23:01 Hydrant: I don't think much CS isn't stolen from somewhere
23:02 Lajla: Yeah
23:02 Lajla: good thing
23:02 Lajla: like, if it was chemistry
23:02 Lajla: They would have gotten a patent on it or soemthing.
23:02 Hydrant: umm
23:02 Hydrant: all software concepts are patented now
23:03 Hydrant: even "efficiently solving NP complete problems, perhaps by approximation" is patented
23:03 Lajla: Wat
23:03 Hydrant: P=NP, P!=NP... either way, we got the patent baby
23:03 Hydrant: yah
23:03 Lajla: I'm pretty sure that you cannot patent algorithms in the EU at the least.
23:03 Lajla: You cannot patent mathematical discoveries here.
23:03 Hydrant: sure, but computers aren't math
23:03 Hydrant: not according to the local computer guy
23:03 Lajla: The point about patent is that you can only patent things you could otherwise keep a secret.
23:04 Lajla: The pointa bout a patent is that you are granted a monopoly in exchange for revealing your technique.
23:04 Lajla: Thus you cannot patent mathematical discoveries, as they are generally useless if you don't reveal them at all.
23:05 Hydrant: well, what about "single-click purchase"
23:05 Hydrant: does that deserve a patent?
23:06 Hydrant: http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&p=1&u=%2Fnetahtml%2FPTO%2Fsearch-bool.html&r=10&f=G&l=50&co1=AND&d=PTXT&s1=20101228.PD.&s2=%28PA.INST.%29&OS=ISD/12/28/2010+AND+IS/PA&RS=ISD/12/28/2010+AND+IS/PA
23:06 rudybot: http://tinyurl.com/4t9dff3
23:29 (join) Senjai
23:37 (quit) masm: Ping timeout: 240 seconds