00:03 (quit) Senjai: Ping timeout: 245 seconds
00:41 (quit) jsnikeris: Read error: Connection reset by peer
00:41 (join) jsnikeris
01:00 (quit) mattmight: Ping timeout: 255 seconds
01:25 (quit) mithos28: Quit: mithos28
02:15 (join) Senjai
02:15 Senjai: Hey Geisteskrankh
02:19 (join) mattmight
02:42 (join) Tabemasu
03:30 Senjai: DT``:  you around
03:32 Senjai: anyone around?
03:33 Senjai: rudybot: (define (multfunc f n) (* 3 (f n)))
03:33 rudybot: Senjai: Done.
03:34 Senjai: rudybot: (define (add5 n) (+ n 5))
03:34 rudybot: Senjai: Done.
03:34 Senjai: rudybot: (multfunc (compose add5 sub1) 0)
03:34 rudybot: Senjai: ; Value: 12
03:36 Senjai: rudybot: (multfunc (compose (compose f add5) sub1) 0)
03:36 rudybot: Senjai: error: reference to undefined identifier: f
03:36 Senjai: rudybot: (multfunc (compose (compose multfunc add5) sub1) 0)
03:36 rudybot: Senjai: error: procedure multfunc: expects 2 arguments, given 1: 4
03:43 Senjai: rudybot: (define (tell-the-world n) (cond [(zero? n) "you all are dead"] [else (string-append "you all are dead" (tell-the-world (sub1 n)))]))
03:43 rudybot: Senjai: Done.
03:43 Senjai: (tell-the-word 10)
03:43 Senjai: rudybot: (tell-the-world 10)
03:43 rudybot: Senjai: ; Value: "you all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are dead"
03:44 Senjai: rudybot: (tell-the-world 100)
03:44 rudybot: Senjai: ; Value: "you all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyou all are deadyo
03:45 Senjai: rudybot: (define (tell-the-world n) (cond [(zero? n) "you all are dead"] [else (string-append "you all are dead " (tell-the-world (sub1 n)))]))
03:45 rudybot: Senjai: Done.
03:45 Senjai: rudybot: (tell-the-world 100)
03:45 rudybot: Senjai: ; Value: "you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all ar
03:45 Senjai: Better :)
03:46 Senjai: rudybot: (time (tell-the-world 100))
03:46 rudybot: Senjai: ; Value: "you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead you all ar
03:46 Senjai: rudybot: (time (tell-the-world 5))
03:46 rudybot: Senjai: ; Value: "you all are dead you all are dead you all are dead you all are dead you all are dead you all are dead"
03:47 Senjai: rudybot: (^ 5 3)
03:47 rudybot: Senjai: error: reference to undefined identifier: ^
03:47 Senjai: rudybot: (expt 5 3)
03:47 rudybot: Senjai: ; Value: 125
03:47 Senjai: rudybot: (tell-the-world (expt 10 1000000))
03:48 Senjai: :/
03:48 rudybot: Senjai: error: with-limit: out of time
03:48 Senjai: lol
03:48 Senjai: poor rudybot
03:48 Senjai: rudybot: hi
03:48 rudybot: Senjai: error: reference to undefined identifier: hi
03:49 Senjai: rudybot: (define (add1toinfinity) (+ 1 (add1toinfinity)))
03:49 rudybot: Senjai: ; stdout: "cpu time: 0 real time: 0 gc time: 0\ncpu time: 0 real time: 0 gc time: 0\n"
03:49 Senjai: rudybot: (define (add1toinfinity) (+ 1 (add1toinfinity)))
03:49 rudybot: Senjai: Done.
03:50 Senjai: rudybot: (add1toinfinity)
03:50 rudybot: Senjai: error: evaluator: terminated (out-of-memory)
03:50 Senjai: Hehe
03:50 (join) PLT_Notify
03:50 PLT_Notify: racket: master Eli Barzilay * 02f6ef0 (7 files in 4 dirs): New Racket version 5.1.0.4. - http://bit.ly/hcCRnL
03:50 (part) PLT_Notify
03:51 Senjai: rudybot: (require 2htdp/image)
03:51 rudybot: Senjai: your sandbox is ready
03:51 rudybot: Senjai: error: dynamic-require: unknown module: '#%mred-kernel; need to run in gracket instead of racket
03:52 Senjai: rudybot: (require 2htdp/batch-io)
03:52 rudybot: Senjai: Done.
03:52 Senjai: oooo
03:52 Senjai: This could be fun
03:57 Senjai: (find-system-path 'home-dir)
03:57 Senjai: rudybot (find-system-path 'home-dir)
03:57 Senjai: rudybot: (find-system-path 'home-dir)
03:57 rudybot: Senjai: ; Value: #<path:/home/erich/>
04:01 Senjai: (directory-exists? "/home/erich/")
04:01 Senjai: (directory-exists? '/home/erich/)
04:02 Senjai: (directory-exists? #<path:/home/erich/>)
04:02 Senjai: (+ 1 2)
04:02 (part) Senjai
04:04 (join) Senjai
04:33 (part) Tabemasu
04:49 (quit) misterm: Ping timeout: 255 seconds
05:11 (join) lucian
06:19 (join) masm
06:36 (join) MayDaniel
07:27 askhader: Oh, oh!
07:27 askhader: I know this one!
07:37 Senjai: hm,'
07:57 DT``: rudybot, (directory-exists? "/home/erich")
07:57 rudybot: DT``: error: directory-exists?: `exists' access denied for /home/erich
07:58 (join) tauntaun
08:03 (join) Fisherman
08:04 Fisherman: Hi. Is there a smart way to ignore hidden files/directories with directory-list?
08:14 tauntaun: (filter (curry regexp-match "^\\.") (map path->string (directory-list)))
08:18 DT``: (filter-map (compose (curry regexp-match "^\\.") path->string) (directory-list))
08:19 DT``: wait, no.
08:19 DT``: we are both wrong.
08:19 tauntaun: Oh?
08:20 DT``: yours matches only the hidden files/directories.
08:20 DT``: mine doesn't work.
08:21 DT``: (filter-map (λ (x) (and (not (regexp-match "^\\." (path->string x))) x)) (directory-list "."))
08:21 tauntaun: right
08:22 tauntaun: Actually I'd meant to use "^[^\\.\" for the regexp pattern
08:22 tauntaun: I mean: "^[^\\.]"
08:23 tauntaun: So why do you need "." as arg to directory-list?
08:24 DT``: I was testing and I didn't look it up in the docs.
08:26 DT``: But, isn't faster to just (not (char=? (string-ref (path->string x) 0) #\.))?
08:28 (join) Fisherma1
08:30 Fisherma1: thanks
08:31 (quit) Fisherman: Ping timeout: 246 seconds
08:39 (quit) MayDaniel: 
08:45 DT``: rudybot, (letrec ((code data) (data code)) (eq? code data))
08:45 rudybot: DT``: ; Value: #t
09:03 (quit) Fisherma1: Quit: leaving
09:30 (join) anRch
09:45 cpach: hi folks! i'm trying to download a file and saving it to a local copy. i've built a function based on copy-port. it seems to work fine with plain http, but it borks on https.
09:45 cpach: here's my function: http://slexy.org/view/s2Y5IYyHSw
09:45 (join) mceier
09:46 cpach: first, maybe i should ask if this is a sane approach at all? :)
10:44 (quit) anRch: Quit: anRch
10:48 (join) mithos28
10:54 (join) dnolen
10:54 (quit) dnolen: Excess Flood
10:56 (join) dnolen
11:10 offby1: cpach: that code looks fine.  But SSL support, if I recall correctly, is sort of half-backed
11:10 offby1: baked
11:10 offby1: lemme see how I've done it
11:13 offby1: cpach: http://ix.io/1xu  <-- I'm not sure if this will work "as is", but this is the general idea for doing SSL
11:18 Geisteskrankh: Senjai, still there?
11:19 offby1: cpach: here are some style changes to your function: http://ix.io/1xv
11:19 offby1: I inverted the two "call-with" things, since I figured there's no point opening the output file unless we've successfully connected to the web server
11:19 offby1: also I used "curry" since I like it :)
11:20 bremner: yum
11:20 offby1: with rice.
11:21 (quit) Senjai: Ping timeout: 245 seconds
11:26 cpach: offby1: thank you for the examples! i'll have a look at them later.
11:28 offby1: cpach: pester eli for help; he wrote what I pasted (frankly, I don't understand it)
11:46 (join) studentUK1
11:46 studentUK1: Hi, is anyone here?
11:49 bremner: studentUK1: go ahead and ask your question
11:50 studentUK1: thanks, I've just started looking at the little schemer book and I cant find a lot of the formula to practice with
11:52 studentUK1: one of them I'm stuck on is Is it true that this is an S-expression?      xyz
11:53 bremner: unless I miss something subtle, yes it is.
11:54 studentUK1: but how would I go about putting this on DrRacket?
11:55 bremner: one way is (read)<return>xyz<return>
11:56 (quit) dnolen: Quit: dnolen
11:56 studentUK1: I tried that but I get "expand: unbound identifier in module"
11:57 bremner: what language (#lang) do you use?
11:57 studentUK1: racket
11:58 bremner: hmm. It works fine for me. What happens when you type (read) ?
11:58 studentUK1: I get some big box bellow with eof
11:59 bremner: right. so type in that box
11:59 studentUK1: I see 'xyz comes up
11:59 bremner: there yah go.
12:01 bremner: now try something your pretty sure isn't an S-expression
12:01 studentUK1: thanks, thats sorted out that one. Do you know of a website that lists in detail for racket as I've got more than S-expressions to look at.
12:01 bremner: although, that might be tough ;)
12:01 bremner: studentUK1: racket-lang.org has tons of documentation
12:02 bremner: there is a "guide" for getting started.
12:02 (quit) studentUK1: Quit: Page closed
12:40 (join) MayDaniel_
12:58 (quit) masm: Ping timeout: 255 seconds
13:05 (join) masm
13:18 tauntaun: When is the strictly functional style not appropriate?
13:19 (quit) jsnikeris: Remote host closed the connection
13:19 tauntaun: I mean functional updates, etc.
13:20 tauntaun: I'm trying to write an RPG, which of course has agents, objects, and locations.  And it seems to me that certain things might be better done with set! and maybe even with an object-oriented approach.
13:25 DT``: tauntaun, go with functional style until you really can't make it look good.
13:25 DT``: I usually set! only on hash tables.
13:26 tauntaun: Yes, obviously, but the question is where to draw the line.
13:26 tauntaun: OK, consider the following scenario.
13:26 tauntaun: An agent picks up an object.
13:26 tauntaun: Game objects have locations.
13:26 tauntaun: The location must change when the agent picks it up, from the room to the agent.
13:26 tauntaun: Following me so far?
13:26 DT``: yeah.
13:27 tauntaun: Several things need to change because of this action:
13:27 tauntaun: 1.  The object's location must now refer to the agent.
13:27 tauntaun: 2. The agent's list of possessions must now include the object.
13:27 tauntaun: 3. The location's list of contents must lose the object.
13:27 DT``: I'd use imperative style here.
13:27 tauntaun: Can these three updates really be done in the functional style without awkwardness?
13:28 tauntaun: ok.
13:28 tauntaun: Anyone else have any opinions?
13:34 offby1: you might look at the "Universe" stuff ... not sure where it lives, but it's mostly or entirely functional
13:34 offby1: and yet it's designed, I think, for writing the sort of games for which you'd think you'd need state.
13:41 tauntaun: offby1: thanks, although it looks like it requires the 2htdp language...
13:43 (quit) masm: Ping timeout: 255 seconds
14:16 (join) dnolen
14:42 (join) masm
15:15 (quit) mithos28: Quit: mithos28
15:20 Geisteskrankh: Who invented syntax-case?
15:20 Geisteskrankh: And when
15:20 Geisteskrankh: and how
15:22 (join) mithos28
15:31 (quit) dnolen: Read error: Connection reset by peer
15:36 (join) dnolen
15:36 (quit) dnolen: Excess Flood
15:37 (join) dnolen
15:37 (quit) mithos28: Quit: mithos28
15:59 (quit) dnolen: Quit: dnolen
16:17 (join) dnolen
16:17 (quit) dnolen: Excess Flood
16:19 (join) dnolen
16:23 (quit) lucian: Remote host closed the connection
17:15 offby1: me, yesterday, I was inspired by some tea leaves
17:16 (quit) MayDaniel_: Read error: Connection reset by peer
17:17 bremner: lies! those were not tea leaves.
17:40 (join) shofetim`
17:42 (quit) shofetim: Ping timeout: 255 seconds
17:44 Geisteskrankh: offby1, you troll me.
17:50 offby1: I do, it's true.
17:56 Geisteskrankh: offby1, what do you have to say about yourself young man?
17:56 offby1: I say: I am not a young man
17:57 offby1: but that's OK
17:57 (quit) DT``: Ping timeout: 240 seconds
17:58 (join) DT``
18:06 Geisteskrankh: offby1, how old are you?
18:10 (quit) tauntaun: Quit: Ex-Chat
18:37 (quit) dnolen: Quit: dnolen
18:41 (part) shofetim`: "ERC Version 5.3 (IRC client for Emacs)"
18:49 (quit) cow-orker: Ping timeout: 248 seconds
18:50 (join) jonrafkind
18:57 (join) cow-orker
19:00 (join) lucian
19:03 (quit) lucian: Remote host closed the connection
19:33 (nick) coyotama|2 -> lakkris
19:33 lakkris waves hi
19:36 jonrafkind makes a snow angel
19:47 (join) tauntaun
19:59 (join) misterm
20:16 (join) flop
20:17 (nick) flop -> mzlost
20:18 mzlost: howdy, connected from racket website ;o)     Examples i'm trying from net only work with "#lang mzscheme"  and always give errors i can't solve with "#lang racket"...  is mzscheme an old unsupported dialect or something okay to use?
20:21 DT``: racket should be preferred
20:21 DT``: *.
20:21 jonrafkind: what errors do you get?
20:23 mzlost: jonra, have had a few diff. erros...   when trying a selection of different macros i got an error:   unbound identifier for    datum->syntax-object
20:23 jonrafkind: where did you find these examples?
20:24 jonrafkind: datum->syntax-object is named 'datum->syntax' in #lang racket
20:24 Geisteskrankh: You tell 'em.
20:24 mzlost: varried websites... ah, i see.. also found   defmacro becomes define-macro
20:24 Geisteskrankh: defmacro and define-macro are baaaad
20:24 Geisteskrankh: I tell 'em.
20:24 mzlost: lol
20:25 Geisteskrankh: mzlost, you should use define-syntax
20:26 mzlost: thanks... i'm really just learning and got sidetracked on these examples.   It feels like there are a lot of dialects and a lot of (baaaad advice) out there int he world ;o)
20:26 mzlost: thanks for the tip jonrafkind, and the warning geisteskrankh
20:27 Geisteskrankh: This is no mere warning my padawan.
20:27 Geisteskrankh: This is an order.
20:29 mzlost: lol.. okay.. i'll take it to heart.    here is one of the web sites i was getting some examples from:  http://is.gd/8qxsKY
20:30 jonrafkind: omfg the first example makes my eyes bleed
20:30 jonrafkind: i want to kill something now
20:31 jonrafkind: he should have just used cut if he wanted a better syntax for lambda instead of making up some lame shorthand
20:31 mzlost: it  was the first example i tried, and it didn't work of course... i guess because it made #lang racket bleed too
20:31 Geisteskrankh: mzlost, don't they all use define-syntax there
20:31 Geisteskrankh: jonrafkind, no
20:31 Geisteskrankh: it rocks.
20:32 jonrafkind: what, : ?
20:32 (quit) mceier: Quit: leaving
20:32 Geisteskrankh: jonrafkind, the first example
20:32 Geisteskrankh: The [: / x 2] macro.
20:32 DT``: wait, why is he using syntax-case instead of with-syntax?
20:32 Geisteskrankh: Eh
20:32 Geisteskrankh: [: / _ 2]
20:32 jonrafkind: (define-syntax-rule (i x ...) (if x ...)) wow! what a great macro!!!!!!!!!!1111
20:32 jonrafkind: i saved a whole 1 letter!
20:32 Geisteskrankh: jonrafkind, oh
20:32 Geisteskrankh: no
20:32 Geisteskrankh: that's silly
20:32 Geisteskrankh: [: / _ 2]  ; <-- much better than (lambda (x) (/ x 2))
20:32 jonrafkind: saving 1 letter is silly but saving 4 is not?
20:32 Geisteskrankh: I meant that one.
20:32 Geisteskrankh: No
20:33 Geisteskrankh: it's about not naming a variable when you only use one anyway.
20:33 Geisteskrankh: Like
20:33 Geisteskrankh: I also have this macro
20:33 jonrafkind: yes but theres cut fort hat
20:33 Geisteskrankh: (& ...)
20:33 jonrafkind: where you can do (cut (/ $1 $2 $3)) or something, i forget exactly
20:33 Geisteskrankh: Which is basically (lambda ((X void) (Y void) (Z void)) ...)
20:33 Geisteskrankh: Yeah
20:34 Geisteskrankh: that's cool
20:34 Geisteskrankh: But I have X Y and Z
20:34 Geisteskrankh: don't ask me why
20:34 Geisteskrankh: Oh wait
20:34 Geisteskrankh: ask me why
20:34 Geisteskrankh: it has a reason
20:34 Geisteskrankh: gogogoog, ask me.
20:34 jonrafkind: and aif is just dumb, imo
20:34 jonrafkind: another 5 letters saved
20:34 Geisteskrankh: What does aif do?
20:34 jonrafkind: people who like aif probably like to write C code like if (x++ > 5){ ... }
20:35 DT``: everyone has his personal dirty macros.
20:35 jonrafkind: it binds the result of the condition in the body of the then, I think
20:35 Geisteskrankh: jonrafkind, I do that all the time.
20:35 Geisteskrankh: Isn't that just or?
20:35 DT``: Geisteskrankh, (aif pred true false) is (let ((it pred)) (if it true false))
20:35 DT``: (with ``it'' being unhygienic)
20:36 Geisteskrankh: What is the use?
20:36 Geisteskrankh: It's just if
20:36 jonrafkind: so you dont have to write the let
20:36 jonrafkind: its a time saving macro
20:36 DT``: Geisteskrankh, you can reference the `pred' part using `it'.
20:36 jonrafkind: it saves exactly 0.00000000000000000000000001 seconds
20:36 Geisteskrankh: Ohhh
20:36 Geisteskrankh: yeah
20:36 tauntaun: does hygiene really make a difference in that example?
20:36 Geisteskrankh: I guess
20:36 Geisteskrankh: tauntaun, that
20:36 jonrafkind: which you can use to look at more lolcat pictures on the internet
20:36 Geisteskrankh: you can reference it.
20:36 Geisteskrankh: That's actually pretty cool.
20:36 Geisteskrankh: Yeah, I like that.
20:36 Geisteskrankh: The less code duplication the better
20:37 Geisteskrankh: If it's a common pattern you see, better abstract it out.
20:37 Geisteskrankh: jonrafkind, also
20:37 tauntaun: Lajla: I'm afraid your answer was a non-answer again.
20:37 Geisteskrankh: this is awesome in javascript: while (obj = obj.nextSibling()) { ....}
20:37 jonrafkind: what a failure
20:37 tauntaun: what a commonplace
20:38 Geisteskrankh: or while(obj = array[i++]) { ... }
20:38 Geisteskrankh: truly so short and elegant.
20:38 jonrafkind barfs
20:38 Geisteskrankh: Baby
20:38 Geisteskrankh: give it to me.
20:38 Geisteskrankh: rawr.
20:38 tauntaun: But someone please confirm, for the record, that hygiene makes no difference to (let ((it pred)) (if it true false)).
20:38 jonrafkind: do you know how many stupid bugs have resulted in such idioms
20:38 DT``: tauntaun, no.
20:38 DT``: you can't reference `it' if it is hygienic.
20:39 tauntaun: jonrafkind: you mean *from* such idioms?
20:39 jonrafkind: yea
20:39 Geisteskrankh: tauntaun, hygiene makes all the difference.
20:39 Geisteskrankh: jonrafkind, no, they remove it if any
20:39 Geisteskrankh: because you take away code duplication
20:39 tauntaun: Then I'm still confused about hygiene.
20:39 tauntaun: It's the macro that introduces the let and the binding to 'it', right?
20:39 jonrafkind: no, it just moves the code to a place where semantics are unclear
20:39 Geisteskrankh: jonrafkind, for you maybe.
20:39 DT``: tauntaun, yes.
20:39 jonrafkind: while (obj = array[i]){ i += 1; }
20:39 Geisteskrankh: But this is lajla you're speaking with.
20:40 DT``: and that *breaks* hygiene.
20:40 tauntaun: So, suppose, the hygiene algorithm renames 'it' to blah.
20:40 Geisteskrankh: jonrafkind, to be honest
20:40 DT``: the hygienic version of aif is (aif var pred true false)
20:40 Geisteskrankh: I'd rather abstract that as
20:40 tauntaun: The let becomes (let ((blah pred)) (if blah true false))
20:40 DT``: you have to name `it' yourself.
20:40 Geisteskrankh: foreach (obj, array) { ....}
20:40 tauntaun: which has the same semantics.
20:40 jonrafkind: yes of course it needs abstractions, but lacking the proper abstractions I would not recommend using the built-in features to create an ad-hoc version
20:40 jonrafkind: doing so is dangerous
20:40 Geisteskrankh: jonrafkind, hmm, I can see where you get there yeah
20:41 Geisteskrankh: I mean, if the array contains a false value.
20:41 Geisteskrankh: But going over all the elements of an array is a common task
20:41 Geisteskrankh: So basically you want to abstract as much as possible to ensure that if you don't fail in the abstraction, you fail nowhere.
20:41 Geisteskrankh: Of course foreach is better.
20:41 tauntaun: Sorry guys, but could someone please cure my confusion?
20:42 Geisteskrankh: tauntaun, hygiene?
20:42 tauntaun: I made a specific point that went unremarked.
20:42 Geisteskrankh: tauntaun, well, let's take or as an example.
20:42 Geisteskrankh: Yeah, let me get there
20:42 tauntaun: no, lajla.
20:42 tauntaun: let's please take the example I gave above.
20:42 Geisteskrankh: Okay
20:42 tauntaun: First things first.
20:42 tauntaun: Is it not true that...
20:42 tauntaun: hygiene renames (let ((it pred)) (if it true false)) to...
20:43 Geisteskrankh: tauntaun, so, it becomes (let ((it pred)) (if it true false))
20:43 tauntaun: (let ((something-else pred)) (if something-else true false))
20:43 DT``: true.
20:43 Geisteskrankh: Yeah
20:43 Geisteskrankh: provided that something-else is not not in the expression itself.
20:43 tauntaun: Then how are the semantics any different?
20:43 Geisteskrankh: Well, consider this example
20:43 tauntaun: (Lajla: yes of course.)
20:44 DT``: tauntaun, the difference is:
20:44 tauntaun: Ah, wait.
20:44 DT``: you don't know the name of `something-else'.
20:44 tauntaun: OK, go ahead.
20:44 tauntaun: Yes, but the macro uses 'something-else' twice.
20:44 Geisteskrankh: (aif (member #\c some-list) (car it) "it is not there!")
20:44 Geisteskrankh: With hygiene, the 'it' in (car it) would be unbound
20:44 tauntaun: Are you saying that the hygiene algorithm will rename only the first 'something-else'?
20:44 Geisteskrankh: without hygiene, it actually gets it.
20:45 Geisteskrankh: tauntaun, the macro doesn't use something-else
20:45 Geisteskrankh: The expander *may*
20:45 Geisteskrankh: Or use whatever trick to get the same semantics.
20:45 Geisteskrankh: the macro uses 'it'
20:46 tauntaun: Look, let's start from the beginning. Hygiene is a renaming of identifiers introduced by the macro that may be captured by bindings in the larger scope, right?
20:46 DT``: right.
20:46 tauntaun: So let's suppose there's a global variable named 'hi'.
20:46 tauntaun: Then I write...
20:46 tauntaun: (let ((hi pred)) (if hi true false))
20:46 tauntaun: The macro then rewrites this to...
20:46 tauntaun: (let ((bye pred)) (if bye true false))...so as not to conflict with the global hi.  Right?
20:47 DT``: if it is the result of a macro expansion, yes.
20:47 tauntaun: What do you mean?
20:47 tauntaun: We're talking about macros, right?
20:47 DT``: if it is generated by a macro.
20:47 DT``: yea.
20:47 DT``: so, yes.
20:47 tauntaun: Why are you restating assumptions?
20:47 tauntaun: Anyway...
20:47 DT``: just to be sure.
20:48 Geisteskrankh: This is actually not needed to do by the expander
20:48 tauntaun: What??!
20:48 Geisteskrankh: As the let shadows the global variable.
20:48 Geisteskrankh: In this case
20:48 Geisteskrankh: could just as well be the same.
20:48 tauntaun: That's exactly my point.
20:48 tauntaun: OMG.
20:48 Geisteskrankh: But that's not what happens here.
20:48 tauntaun: That's what I've been saying the whole time.
20:48 Geisteskrankh: Yeah, and that's not relevant here.
20:49 tauntaun: Lajla: yes or no, does the hygiene algorithm rename the let as I have written it at 20:46?
20:49 Geisteskrankh: See my example: (aif (member #\c some-list) (car it) "it is not there!")) succeeds in a non hygienic example.
20:49 tauntaun: NO!!!!
20:49 Geisteskrankh: It doesn't.
20:49 tauntaun: We were talking about something specific.
20:49 Geisteskrankh: Okay.
20:49 Geisteskrankh: Well, then it doesn't.
20:49 Geisteskrankh: Also, I'm not in your timezone.
20:49 tauntaun: Then  why did everyone insist that hygiene "makes all the difference"?
20:49 Geisteskrankh: Minutes suffice.
20:49 Geisteskrankh: Because it does in this example.
20:49 tauntaun: (Fine: 3 minutes ago.)
20:49 Geisteskrankh: the aif
20:49 Geisteskrankh: of which we spoke.
20:50 tauntaun: Please refere to DT``'s utterance at minute 35.
20:50 tauntaun: (*refer)
20:50 Geisteskrankh: CAn't you quote him, I don't have timestamps on.
20:51 tauntaun: <DT``> Geisteskrankh, (aif pred true false) is (let ((it pred)) (if it true false))
20:51 tauntaun: <DT``> (with ``it'' being unhygienic)
20:51 Geisteskrankh: That is true
20:51 tauntaun: What does it mean, "unhygienic", when we've agreed that the let is the binding that wins?
20:51 Geisteskrankh: It expands to that.
20:52 Geisteskrankh: Depends on the body of the let if it matters
20:52 Geisteskrankh: In your body it doesn't.
20:52 tauntaun: The bodies are the same.
20:52 Geisteskrankh: Nope, let me write the difference up:
20:52 tauntaun: Are you saying that true and/or false are expressions that may contain 'it'?
20:53 DT``: tauntaun,
20:53 DT``: rudybot, (define-syntax-rule (aif p t f) (let ((it p)) (if it t f)))
20:53 rudybot: DT``: your sandbox is ready
20:53 rudybot: DT``: Done.
20:53 DT``: this aif is *hygienic*.
20:54 DT``: rudybot, (aif (+ 2 3) it 'false)
20:54 rudybot: DT``: error: reference to undefined identifier: it
20:54 DT``: you can't refer to the `it' declared inside the macro body.
20:54 DT``: rudybot, (define-syntax (aif2 stx) (syntax-case stx () ((aif2 p t f) (with-syntax ((it (datum->syntax stx 'it))) #'(let ((it p)) (if it t f))))))
20:54 rudybot: DT``: Done.
20:54 DT``: this aif is unhygienic.
20:54 DT``: rudybot, (aif2 (+ 2 3) it 'false)
20:54 rudybot: DT``: ; Value: 5
20:55 DT``: in this aif, I *can* refer to it.
20:55 DT``: this means that if there was some other variable called ``it'', it would be shadowed.
20:55 mzlost: hey.. rudybot is cool.
20:55 DT``: rudybot, (define-syntax-rule (aif3 it p t f) (let ((it p)) (if it t f)))
20:55 rudybot: DT``: Done.
20:56 Geisteskrankh: tauntaun: http://codepad.org/o1MQhcff
20:56 mzlost: the first example on that linked page doesn't work even after subbing datum->syntax...   (map [: * _ 2]  mylist)  gives _: wildcard not allowed as an expression
20:56 DT``: in this hygienic aif, *we* name it.
20:56 tauntaun: excuse me, DT``, while I copy and paste your code into Emacs so I can indent it.
20:56 Geisteskrankh: tauntaun, my link explains it all.
20:59 mzlost: noob question.   in your link Geisteskrankh, how do you decide whether aif will be expanded hygienically or not?
20:59 Geisteskrankh: mzlost, I did it manually actually. =P
20:59 Geisteskrankh: How you write the macro determins this
20:59 tauntaun: Lajla: you did not give the macro definition.
20:59 tauntaun: In the codepad.
20:59 Geisteskrankh: (define-syntax-rules (aif p t f) (let ((x p)) (if x t f))) is hygienic.
21:00 Geisteskrankh: tauntaun, I didn't.
21:00 DT``: mzlost, if you write my first aif, it's hygienic.
21:00 DT``: my second example, aif2, it's not.
21:00 Geisteskrankh: I second DT``
21:00 Geisteskrankh: he's as smart as I am.
21:00 mzlost: lol
21:00 Geisteskrankh: *HUGGLES DT``
21:00 DT``: lol
21:00 mzlost: thanks DT... i'll scroll back and look
21:00 tauntaun: And now I see that you are using 'it' in the 'true' expression.  If you had confirmed that from the beginning, I wouldn't have suffered all this confusion. But thanks anyway.
21:01 Geisteskrankh: I did.
21:01 Geisteskrankh: DT``, didn't
21:01 Geisteskrankh: because he's totally stupid
21:01 Geisteskrankh: oh wait.
21:01 tauntaun: At minute 52: <tauntaun> Are you saying that true and/or false are expressions that may contain 'it'?
21:01 DT``: lol.
21:02 DT``: and that is why dirty macros are bad.
21:02 tauntaun: No, that is why imprecise writing is bad.
21:03 DT``: unhygienic macros were shadowing the real answer.
21:03 Geisteskrankh: My IRC senses detect some annoyance.
21:03 tauntaun: Okay, see?
21:03 tauntaun gets grouchy about sloppy writing.
21:04 mzlost: DT``: is the crucial difference the use of define-syntax-rule  vs  define-syntax  ?
21:05 Geisteskrankh: mzlost, nope.
21:05 Geisteskrankh: Best not yet worry about this.
21:05 DT``: mzlost, the crucial difference is using (with-syntax ((it (datum->syntax stx 'it))) ...)
21:05 mzlost: yeah.. i know i'm fixated on stuff beyond my very limited depth
21:05 Geisteskrankh: understanding syntax-case is hard, can only come after you understood women.
21:05 mzlost: Geisteskrankh: then i'll never understand
21:05 DT``: Geisteskrankh, syntax-parse is hard.
21:05 Geisteskrankh: mzlost, let me formalize women for you:
21:06 Geisteskrankh: forall x :  woman(x) -> forall y : diamondring (y) -> loves(x,y)
21:06 DT``: formalizing women = proving that the universe is deterministic.
21:06 mzlost: heh
21:07 Geisteskrankh: I heard that a man that can succesfully parse women emotions provably decides the haling problem.
21:09 mzlost: okay, going back to Harry and Gerry lisp videos...  thanks for the sage advice and fun.. cheers
21:11 (part) mzlost
21:13 DT``: It's 3:13 AM here, I guess I'll go to sleep.
21:13 DT``: 'night.
21:14 tauntaun: Buona notte.
21:17 Geisteskrankh: hah
21:17 Geisteskrankh: now I am the smartest
21:17 (quit) vdrab: Quit: vdrab
21:18 tauntaun: Well, since you're so smart...
21:19 tauntaun: Where can I read up about hygienic macros?
21:19 tauntaun: I read Dybvig's paper some time ago.
21:19 tauntaun: But I need to see more examples.
21:20 Geisteskrankh: tauntaun, I guess dybvig's on syntax case is one of the most important
21:20 Geisteskrankh: THere's not much to explain about hygienic macros as a general concept
21:20 Geisteskrankh: all bindings refer to what they refer to in the environment of the definition, not the expansion.
21:21 tauntaun: Yes, I knew that Lajla.  But...
21:22 tauntaun: I didn't understand how DT`` used with-syntax and datum->syntax to turn off the hygiene algorithm.
21:23 Geisteskrankh: Ahhh
21:23 Geisteskrankh: that's a scheme specific thing
21:24 Geisteskrankh: not about hygienic macros in general.
21:24 Geisteskrankh: But yeah, he basically used datum->syntax to introduce a new binding into the environment.
21:24 Geisteskrankh: Basically
21:25 Geisteskrankh: what (datum->syntax stx obj) does is that it returns a syntax object corresponding to the datum in the environment that it takes from the syntax object stx.
21:25 tauntaun: already understood
21:25 Geisteskrankh: So, what part you don't then?
21:26 tauntaun: No, I understood just that part of the call to datum->syntax
21:26 tauntaun: I guess he's sneaking in the context of stx
21:26 tauntaun: and binding it to 'it in that context?
21:26 tauntaun: (I mean: binding "it" to "'it")
21:27 tauntaun: Anyway, there has to be some book/article/source that sysematically covers hygienic macros.
21:28 Geisteskrankh: tauntaun, well, there is that Dybvig article yeah
21:28 Geisteskrankh: Or jsut the R6RS standard
21:28 Geisteskrankh: But yeah
21:28 tauntaun: They are not tutorials.
21:28 tauntaun: Even Dybvig's is too brief.
21:29 Geisteskrankh: (datum->syntax x y) does use its first argument to determine the context of the resulting syntaxobjecyt.
21:31 Geisteskrankh: You don't technically need with-syntax by the way.
21:31 Geisteskrankh: you can use quasisyntax
21:33 (join) masm1
21:34 (quit) masm: Ping timeout: 250 seconds
21:38 tauntaun: I guess I'll just have to find time to pass over the Racket documentation with a fine-toothed comb.
21:44 (join) vdrab
22:12 (quit) jonrafkind: Ping timeout: 248 seconds
22:39 (quit) tauntaun: Quit: Ex-Chat
22:52 (quit) rudybot: Ping timeout: 276 seconds
22:52 (join) rudybot_
22:54 (nick) rudybot_ -> rudybot
22:55 (join) mithos28
23:08 (quit) offby1: Read error: Connection reset by peer
23:22 (quit) vdrab: Quit: vdrab