00:00 (join) jeapostrophe
00:03 (join) jesyspa
00:17 (quit) mithos28: Remote host closed the connection
00:18 (join) mithos28
00:20 (join) mithos28_
00:21 (quit) mithos28_: Remote host closed the connection
00:21 (quit) mithos28: Read error: Connection reset by peer
00:21 (join) mithos28_
00:29 (quit) blomqvist: Quit: You've got to hide your love away
00:30 (join) wahjava
00:31 (quit) veer: Quit: Leaving
00:32 (join) mithos28
00:32 (quit) mithos28_: Read error: Connection reset by peer
00:42 (quit) jeremyheiler: Quit: Computer has gone to sleep.
00:47 (join) weirdo
00:47 weirdo: hello
00:47 weirdo: i think i found a bug in r5rs mode
00:47 weirdo: racket@seaking2/ellipsis-patmatch> (pair? (vector-ref #(foo (1 2 3)) 1))
00:47 weirdo: #f
00:47 weirdo: racket@seaking2/ellipsis-patmatch> (pair? '(1 2 3))
00:47 weirdo: #t
00:51 (join) mithos28_
00:52 (quit) mithos28_: Remote host closed the connection
00:52 (quit) mithos28: Read error: Connection reset by peer
00:53 weirdo: racket@seaking2/ellipsis-patmatch> pair?
00:53 weirdo: #<procedure:mpair?>
00:53 (join) mithos28
01:12 weirdo: the reader's broken in r5rs mode :(
01:13 neilv: if you file a bug, i suspect it will be fixed promptly
01:36 (join) rbarraud
01:37 (quit) jeapostrophe: Read error: Operation timed out
01:37 (join) gf3
01:38 (join) veer
01:41 veer: can this be shortened :
01:41 veer: (inexact->exact (truncate (exact->inexact (/ width max-width))))
01:41 veer: where width > max-width and both are positive integers
01:44 veer: oh sorry please ignore my question
01:45 (quit) DrDuck1: Ping timeout: 244 seconds
02:00 (join) mceier
02:14 (join) hkBst
02:14 (quit) hkBst: Changing host
02:14 (join) hkBst
02:16 (quit) jacius: Remote host closed the connection
02:25 (quit) duomo: Quit: Linkinus - http://linkinus.com
02:45 (quit) rbarraud: Remote host closed the connection
02:47 (join) mithos28_
02:50 (quit) mithos28: Ping timeout: 264 seconds
02:50 (nick) mithos28_ -> mithos28
02:56 (join) djcoin
02:59 djcoin: Is there a way in Racket to mark some datastructure as being immutable ?
02:59 djcoin: (Hi there ! :) )
03:06 veer: djcoin: I think structures are immutable by default
03:07 jonrafkind: which datastructures are you looking at
03:08 djcoin: Well, I think I will rephrase my question because I have got no clue about how racket may handle immutability and that's what I was curious about :)
03:10 djcoin: So my question should be rather, how immutability is handled in racket, how can it be ensured ?
03:15 jonrafkind: by creating an immutable datastructure
03:15 veer: djcoin: its up to you to ensure that your data structure is immutable or you don't use any thing that mutate your data structure
03:15 jonrafkind: lists are immutable by default
03:15 jonrafkind: (hash) creates an immutable hash, if you want a mutable one you do (make-hash)
03:17 jonrafkind: structs are immutably by default unless you use the #:mutable keyword
03:20 djcoin: All right, thanks a lot. But mutable datastructure can leak from other library into your own code. Is there some good practice when designing library to yield only immutable one ? An ?immutable function to check ?
03:20 djcoin: Im just being curious here. Hope I dont bother :)
03:23 jonrafkind: theres (immutable? x) but i dont know if it works on absoltuely everything
03:23 jonrafkind: but your question is about designing a library so it sounds like you want a type system to ensure that you don't accidentally create mutable datastructures..?
03:23 jonrafkind: there is typed racket but honestly I dont know what it does in terms of mutability
03:24 jonrafkind: a quick guess is that it does not discern between mutable and immutable types
03:26 jonrafkind: it seems typed racket has (HashTable k v) but doesn't say anything about whether the hash is mutable or not
03:26 neilv: you could use contracts to keep mutable objects from coming in. or you could make macros that filter incoming objects to transparently convert them to immutable ones (which has its own problems)
03:27 (join) rbarraud
03:27 djcoin: All right, just wanted to make sure I did not miss some key concept of Racket with immutability
03:28 neilv: plt switching pairs to immutable was a big help to me. i wish they hadn't stopped at pairs
03:28 djcoin: Thanks a lot for your input
03:28 neilv: djcoin: you should ask on the email list
03:28 djcoin: Yeah
03:28 jonrafkind: yes, both (make-hash) and (make-immutable-hash) have the same type in TR
03:47 (quit) jonrafkind: Ping timeout: 246 seconds
04:09 (join) MayDaniel
04:26 weirdo: djcoin, why do you care at all?
04:26 weirdo: mutability has its places
04:27 djcoin: weirdo: Good question indeed. I was comparing to Clojure which only accept immutable one. And I think immutability is awesome for concurrent programming. So I was curious about how Racket handled that.
04:28 weirdo: there's no point to verify mutability on arbitrary data
04:29 weirdo: if you don't like it, don't mutate
04:29 weirdo: it's not like you can do anything with arbitrary typed data
04:30 djcoin: Sure "you" can choose not to mutate data. But what about the whole librairies, framework you may be using ? That was my point.
04:31 weirdo: so you're going to prevent the user of the library from passing mutable data to your code?
04:32 weirdo: also, nice abstractions can be build with mutable structures
04:32 weirdo: which hide away the mutability itself
04:32 weirdo: why don't create a new universe every tail-call while at that?
04:32 weirdo: eh, i'm wasting my time
04:33 djcoin: weirdo: So that was your response to immutability everywhere in clojure
04:33 djcoin: deep
04:33 weirdo: clojure has STM
04:33 weirdo: racket, hardly
04:41 (join) MightyFoo
04:42 MightyFoo: morning all
04:45 djcoin: moring
04:45 djcoin: morning * !
04:47 (nick) MightyFoo -> tim-brown
04:58 (join) bitonic
05:00 (quit) jyc__: Read error: Connection reset by peer
05:09 (join) lcc
05:13 (quit) lcc: Quit: ERC Version 5.3 (IRC client for Emacs)
05:47 (quit) bitonic: Ping timeout: 246 seconds
05:54 (quit) ivan\: Ping timeout: 245 seconds
05:56 (join) ivan\
06:03 (quit) ivan\: Ping timeout: 265 seconds
06:05 (quit) rbarraud: Ping timeout: 265 seconds
06:06 (join) antithesis
06:08 (join) ivan\
06:26 (join) masm
06:43 (join) bitonic
06:48 (join) lcc
06:56 (join) snearch
07:04 (quit) neilv: Ping timeout: 246 seconds
07:32 (quit) lcc: Quit: ERC Version 5.3 (IRC client for Emacs)
07:51 (join) gciolli
07:54 (join) lcc
07:58 (quit) Shviller: Ping timeout: 246 seconds
07:58 (join) Shviller
09:02 aidy: djcoin: there are a few functional data structures on PLaneT
09:05 djcoin: aidy: thanks
09:05 (join) anRch
09:11 (quit) hkBst: Quit: Konversation terminated!
09:12 (join) jeapostrophe
09:14 (quit) snearch: Quit: Verlassend
09:22 (quit) masm: Quit: Leaving.
09:36 (quit) lcc: Quit: ERC Version 5.3 (IRC client for Emacs)
09:47 (join) gridaphobe
09:50 (quit) Shvillr: Read error: Connection reset by peer
09:55 (join) dnolen
10:06 (quit) mceier: Quit: leaving
10:07 (quit) anRch: Quit: anRch
10:08 (quit) jeapostrophe: Ping timeout: 240 seconds
10:08 (quit) MayDaniel: Read error: Connection reset by peer
10:21 (join) Kaylin
10:25 ozzloy: is there a way to escape strings for postgresql queries?
10:38 chandler: hm. what situation are you in that you're having trouble using parameters instead of escaped strings?
10:38 offby1: ozzloy: I'd assume that the built-in db access libraries did that for you
10:38 (quit) Kaylin: Quit: Leaving.
10:39 ozzloy: chandler, what are parameters in this context?
10:39 (quit) jrslepak: Quit: This computer has gone to sleep
10:40 chandler: ozzloy: parameters are things which get substituted for placeholders in your query
10:40 ozzloy: offby1, i'm thinking of the case (query-rows db (format "select * from users where id = ~a" id))
10:40 chandler: e.g. (query-rows db "select foo from table where blah = $1" "this-is-a-parameter"))
10:41 chandler: ok, you definitely should be using parameters for that case, using the syntax I just gave
10:41 ozzloy: and id could contain something like "; drop everything"
10:41 ozzloy: ic
10:41 chandler: http://docs.racket-lang.org/db/query-api.html has a bunch of examples of this
10:41 ozzloy: i will have to look back at the docs
10:41 ozzloy: cool
10:42 chandler: see section 3.1 in particular: "Use parameters instead of Racket string interpolation (eg, format or string-append) to avoid SQL injection, where a string intended to represent a SQL scalar value is interpreted as—possibly malicious—SQL code instead."
10:43 ozzloy: cool
11:02 (quit) dnolen: Ping timeout: 246 seconds
11:04 (join) Shvillr
11:08 (join) jrslepak
11:13 asumu: djcoin: most data structures in Racket are built out of structures, so if the base structure is immutable there is no way to mutate it anyway.
11:14 asumu: And even if you implement something with mutation internally, you can export a pure interface (by simply not exporting mutating functions).
11:15 (join) nilyaK
11:15 asumu: The more interesting aspect of Clojure's data structures is that they are often persistent.
11:16 asumu: Which is not true of most Racket data structures, but that's not a fundamental limitation. They just haven't been designed to be.
11:18 (join) jeapostrophe
11:18 (quit) jeapostrophe: Changing host
11:18 (join) jeapostrophe
11:21 (join) mceier
11:28 djcoin: Thanks asumu. I'm not an expert of immutability/persistent/pure functional data structure. How do you define persistent ? Does it mean you can reuse any part of a persistent datastructure to make other immutable/persistent datastructure without worrying ?
11:37 (join) ssbr__
11:43 asumu: djcoin: persistent means it preserves the old version of itself. In extreme cases, you can go through the entire history.
11:44 asumu: (and modified)
11:44 asumu: In Racket's case, they are persistent usually in the trivial sense that they are copied.
11:45 djcoin: asumu: Whats the difference with immutable ? It aims at the composable aspect of an immutable datastructure I guess ?
11:45 asumu: But Clojure's data structures usually implement persistence via sharing portions of the data structure between versions.
11:45 djcoin: Ok
11:46 asumu: Immutable and persistent mean different things. But persistent data structures are often immutable.
11:46 asumu: And immutable data structures are all persistent in a trivial sense.
11:47 (join) MayDaniel
11:47 asumu: (so when I say persistent, I really mean they are implemented with efficient sharing)
11:48 asumu: Anyway, it would be nice to more data structures like that in Racket. I think it's just a matter of people spending the time to implement them.
11:48 asumu: *to have more
11:49 djcoin: Yeah
11:50 djcoin: I will document myself more on this (even If I read a few stuff around I still have problem figuring out the difference with persitence/immutability etc.)
11:51 djcoin: Thanks a lot for your explanations
11:53 asumu: djcoin: here's an example. In Racket, vectors can be immutable, but I think they are still implemented the same or in a similar way as mutable ones. And they just copy on operations.
11:53 asumu: In Clojure, vectors are really trees: http://blog.higher-order.net/2009/02/01/understanding-clojures-persistentvector-implementation/
11:53 rudybot: http://tinyurl.com/ap5chf
11:55 asumu: And if you just change part of it, you don't have to copy the whole tree. Just the relevant sub-trees.
11:56 djcoin: asumu: And what are your thoughts about forcing the whole language to have such immutable datastructure ? Seems good for concurrent application (if you need such application of course)
11:56 djcoin: I for one does not have such need in fact, but I find it interesting
12:00 djcoin: But I won't bother you more - I guess I will come back later when I will have more clue about all this
12:01 asumu: I think I'd prefer to provide both, but I don't really know either.
12:08 (quit) nilyaK: Quit: Leaving.
12:09 (join) nilyaK
12:15 (quit) veer: Quit: Leaving
12:35 (join) mithos28
12:36 (join) anRch
12:42 (quit) djcoin: Ping timeout: 246 seconds
12:57 (quit) mithos28: Quit: mithos28
13:05 (quit) anRch: Quit: anRch
13:27 (join) jacius
13:27 (join) jonrafkind
13:27 (quit) jonrafkind: Changing host
13:27 (join) jonrafkind
13:31 (join) lcc
14:03 (join) carleastlund
14:11 (quit) lcc: Quit: fuzzy logic
14:11 (quit) bitonic: Quit: ERC Version 5.3 (IRC client for Emacs)
14:12 (join) lcc
14:17 (join) anRch
14:22 (quit) gciolli: Ping timeout: 248 seconds
14:52 (join) RacketCommitBot
14:52 RacketCommitBot: [racket] plt pushed 2 new commits to master: http://git.io/IzqV2Q
14:52 RacketCommitBot: [racket/master] Add examples for most class clauses. - Asumu Takikawa
14:52 RacketCommitBot: [racket/master] Fix class init grammar in docs - Asumu Takikawa
14:52 (part) RacketCommitBot
15:00 (join) soegaard
15:00 (quit) antithesis: Remote host closed the connection
15:09 (join) antithesis
15:12 (quit) antithesis: Remote host closed the connection
15:14 (quit) anRch: Quit: anRch
15:24 (join) bitonic
15:24 (quit) bitonic: Remote host closed the connection
15:26 (join) bitonic
15:38 (quit) bitonic: Remote host closed the connection
15:40 (join) jyc
15:46 (join) bitonic
15:54 (join) adu
15:58 (join) jeremyheiler
15:59 (part) ChibaPet
16:14 (quit) jacius: Remote host closed the connection
16:24 (join) rbarraud
16:29 (join) RacketCommitBot
16:29 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/q0K2JQ
16:29 RacketCommitBot: [racket/master] Fix Performance Report on unsaved editors. - Vincent St-Amour
16:29 (part) RacketCommitBot
16:30 (quit) nilyaK: Quit: Leaving.
16:32 (quit) rbarraud: Read error: Connection reset by peer
16:38 (quit) jrslepak: Quit: This computer has gone to sleep
16:40 (quit) MayDaniel: Read error: Connection reset by peer
16:43 (join) antithesis
16:53 (quit) antithesis: Remote host closed the connection
16:54 (quit) jesyspa: Quit: leaving
17:08 (join) Fare
17:10 (join) rbarraud
17:11 (quit) gridaphobe: Remote host closed the connection
17:14 (quit) rbarraud: Read error: Connection reset by peer
17:19 (quit) cdidd: Ping timeout: 240 seconds
17:19 (join) gridaphobe
17:22 (quit) stamourv: Remote host closed the connection
17:26 (join) stamourv
17:26 (quit) stamourv: Changing host
17:26 (join) stamourv
17:35 (join) cdidd
17:41 (quit) soegaard: Quit: soegaard
17:43 (join) rbarraud
17:44 (join) Kaylin
17:50 (quit) rbarraud: Read error: Connection reset by peer
17:50 (join) rbarraud
17:51 (join) DrDuck
18:06 (join) jacius
18:06 (quit) jeremyheiler: Quit: Computer has gone to sleep.
18:07 (quit) mceier: Quit: leaving
18:10 (quit) jeapostrophe: Ping timeout: 250 seconds
18:13 (join) Urm3l
18:20 (quit) rbarraud: Read error: Connection reset by peer
18:22 (join) rbarraud
18:35 (join) djcoin
18:51 (quit) adu: Quit: adu
19:02 (quit) Urm3l: Quit: Urm3l
19:04 (quit) rbarraud: Read error: Connection reset by peer
19:12 (join) jrslepak
19:22 (quit) jyc: Ping timeout: 244 seconds
19:35 (quit) djcoin: Quit: WeeChat 0.3.2
19:45 (join) jeremyheiler
19:47 (quit) jonrafkind: Ping timeout: 248 seconds
19:52 (join) dnolen
19:57 (quit) carleastlund: Quit: carleastlund
20:15 (quit) stchang: Read error: Operation timed out
20:18 (join) stchang
20:19 (quit) bitonic: Ping timeout: 246 seconds
20:32 (quit) dnolen: Ping timeout: 248 seconds
20:42 (join) bitonic
21:36 (join) dalaing
21:47 (join) jeapostrophe
21:49 (join) jyc
22:04 (quit) gridaphobe: Remote host closed the connection
22:05 (quit) bitonic: Ping timeout: 246 seconds
22:07 (quit) jeapostrophe: Ping timeout: 250 seconds
22:07 (join) jeapostrophe
22:26 (join) mmalorni
22:37 (quit) jeapostrophe: Read error: Operation timed out
22:38 (quit) jeremyheiler: Quit: Computer has gone to sleep.
22:39 (nick) Aethaeryn -> nyreahtea
22:42 (quit) lcc: Quit: ERC Version 5.3 (IRC client for Emacs)
23:01 (join) lcc
23:06 (join) rbarraud
23:11 (join) jonrafkind
23:15 eli: Fare: So, something like this? -- http://tmp.barzilay.org/x
23:16 Fare: a bit like this, yes
23:17 Fare: except that the listof can be recursively used, and for each use, a new function is defined, though structurally equal expressions have the same underlying function.
23:17 eli: (Actually, slightly revise it to say that (sum 12) could throw a syntax error otherwise it could be done with plain functions.)
23:18 Fare: yes
23:20 (join) jeapostrophe
23:20 (quit) jeapostrophe: Changing host
23:20 (join) jeapostrophe
23:21 asumu: rudybot: (define foo (listof integer?))
23:21 rudybot: asumu: your sandbox is ready
23:21 rudybot: asumu: Done.
23:21 asumu: rudybot: (foo 12)
23:21 rudybot: asumu: ; Value: #f
23:21 asumu: rudybot: (foo '(1 2))
23:21 rudybot: asumu: ; Value: #t
23:21 asumu: (I have no context, so I have no idea what that example was about though)
23:22 (join) BeLucid_
23:22 asumu: (just thought it was cute that the contract works here)
23:22 (quit) BeLucid: Ping timeout: 246 seconds
23:22 eli: asumu: That's not coincidental, given the person who wrote that example...
23:23 asumu: eli: you?
23:24 eli: Yes.
23:24 Fare: I suppose that one could trust the compiler to do some common expression elimination at the module and/or global level to eliminate duplicates of identical contracts.
23:25 asumu: eli: FYI, Google has now adopted your meta description in search. Sitelinks are still weird.
23:25 eli: Fare: That would be easy with some memoization of things like `listof'
23:25 eli: asumu: It did??
23:26 Fare: eli: is that memoization user-controllable?
23:26 eli: asumu: Whoa -- Yes...  I actually checked it earlier today, and noticed only that the bogus sitelinks are there...
23:26 Fare: also, normalizing the things that are to be memoized can be "interesting" in presence of scoping.
23:27 eli: Fare: Well, the point is that whatever you do with that memoization is independent of how you'd implement it.
23:28 Fare: also, what about memoizing the stuff at runtime as well as at syntax time?
23:29 eli: No problem with that, you'd have two different memoization tables, one for each level.
23:29 Fare: at what point is the runtime one filled?
23:29 Fare: since I'm not at a code point, I can't expand into code that fills the runtime table.
23:29 eli: When the module is invoked.
23:30 Fare: so I need to expand into code that will be run at module init time or such
23:31 Fare: and/or somehow emit annotations that are collected by whichever program creates the module expansion.
23:31 eli: Module-toplevel code gets to run when you invoke the module, so that part is easy.
23:34 Fare: any macro can register module-toplevel code in addition to expanding into something fit for the current scope?
23:34 (join) siduction
23:35 Fare: is probably even more important for syntax-parse kind of macros -- you might want more levels than just module-toplevel and current point.
23:35 eli: There are "lift" macros that expand into code that gets lifted into the module's toplevel.
23:35 eli: But I suspect that you don't need those.
23:36 Fare: i.e. for call-by-exploded-structure kind of lambda-parameter macros, you'd want to expand into both several parameters in the lambda-list and code that wraps around the body of the lambda to do funky things with bindings
23:37 (join) vu3rdd
23:37 (quit) vu3rdd: Changing host
23:37 (join) vu3rdd
23:41 eli: Fare: I don't know what you mean by that...
23:41 (quit) dalaing: Quit: Leaving.
23:41 eli: Which is why I suggested an incremental thing towards what you really want...
23:41 eli: In any case, I got delayed due to a stupid mistake, and due to munching.
23:41 eli: How about this:
23:42 eli: http://tmp.barzilay.org/x
23:42 eli: (reload)
23:42 (join) dalaing
23:42 (part) dalaing
23:44 Fare: in the CL example, the list-of need not appear in a deftype, but in any expression in the type sublanguage.
23:45 eli: It's possible to write a compiler kind of thing that would start from, say, (Listof Integer) and "compile" it to a runtime expression like (listof integer?)
23:47 Fare: of course - but whichever processors accepts the sublanguage then needs to be somehow extended to account for the macro.
23:48 eli: You mean make the type expression be a normal expression?
23:51 (quit) rbarraud: Read error: Connection reset by peer
23:52 (join) rbarraud
23:53 (join) veer
23:56 Fare: no, it's never gonna be a normal expression
23:56 Fare: it's processed by a different processor