00:12 (join) cdidd
00:13 (quit) mithos28: Quit: mithos28
00:14 (join) mithos28
00:15 (quit) anonymou1: Ping timeout: 255 seconds
00:21 (join) Kaylin
00:24 (quit) jao: Read error: Operation timed out
00:32 (join) Kaylin1
00:35 (quit) Kaylin: Ping timeout: 276 seconds
00:40 (join) rmathews
00:50 (quit) Draggor: Ping timeout: 272 seconds
00:54 (quit) mceier: Quit: leaving
00:56 (quit) tcsc: Quit: computer sleeping
01:00 (join) tcsc
01:10 (join) gridaphobe
01:12 (join) Draggor
01:30 (quit) rbarraud__: Excess Flood
01:30 (join) rbarraud__
01:55 (join) asvil
02:03 (quit) mithos28: Quit: mithos28
02:18 (join) mceier
02:43 (quit) Kaylin1: Read error: Connection reset by peer
02:46 (quit) gridaphobe: Remote host closed the connection
02:47 (join) gridaphobe
02:47 (quit) gridaphobe: Read error: Connection reset by peer
02:48 (join) Kaylin
02:50 (quit) tcsc: Quit: computer sleeping
02:56 (quit) ASau: Ping timeout: 248 seconds
03:11 (quit) jonrafkind: Ping timeout: 252 seconds
03:31 (join) rmathews_
03:34 (quit) rmathews: Ping timeout: 246 seconds
03:34 (nick) rmathews_ -> rmathews
03:56 (join) hkBst
03:58 (quit) eikonos: Quit: Leaving.
03:59 (quit) hkBst: Excess Flood
04:00 (join) hkBst
04:00 (quit) hkBst: Changing host
04:00 (join) hkBst
04:07 (quit) hkBst: Ping timeout: 246 seconds
04:07 (join) hkBst
04:07 (quit) hkBst: Changing host
04:07 (join) hkBst
04:16 (quit) Nisstyre-laptop: Ping timeout: 256 seconds
04:33 (join) hkBst_
04:33 (quit) hkBst_: Changing host
04:33 (join) hkBst_
04:34 (quit) hkBst: Ping timeout: 246 seconds
04:36 (quit) dauterive: Quit: Leaving
04:40 (join) bitonic
04:40 (quit) hkBst_: Read error: Connection reset by peer
04:45 (join) hkBst_
04:45 (quit) hkBst_: Changing host
04:45 (join) hkBst_
04:50 (quit) hkBst_: Read error: Connection reset by peer
04:51 (join) hkBst_
04:51 (quit) hkBst_: Changing host
04:51 (join) hkBst_
04:51 (join) ijp
05:12 (quit) Kaylin: Read error: Connection reset by peer
05:14 (quit) rbarraud__: Ping timeout: 255 seconds
05:48 (quit) hkBst_: Read error: Connection reset by peer
05:48 (join) hkBst_
05:48 (quit) hkBst_: Changing host
05:48 (join) hkBst_
05:52 (join) soegaard
05:55 (quit) hkBst_: Remote host closed the connection
05:56 (join) hkBst_
05:56 (quit) hkBst_: Changing host
05:56 (join) hkBst_
05:58 (join) eikonos
05:58 (quit) eikonos: Changing host
05:58 (join) eikonos
06:03 (quit) eikonos: Ping timeout: 255 seconds
06:17 (quit) ijp: Ping timeout: 245 seconds
06:25 (quit) bjz_: Quit: Bye!
06:34 (join) rmathews
06:37 (part) anttih
06:39 (quit) bitonic: Ping timeout: 264 seconds
06:47 (quit) jrslepak: Ping timeout: 252 seconds
06:48 (join) jrslepak
06:54 (join) vkz
07:06 (quit) hkBst_: Ping timeout: 246 seconds
07:06 (join) hkBst__
07:14 (quit) noam_: Read error: Connection reset by peer
07:15 (join) noam
07:32 (join) jeapostrophe
07:32 (quit) jeapostrophe: Changing host
07:32 (join) jeapostrophe
07:32 (join) samth_
07:36 (join) bitonic
07:58 (join) eikonos
07:58 (quit) eikonos: Changing host
07:58 (join) eikonos
08:02 (quit) bitonic: Ping timeout: 240 seconds
08:03 (quit) eikonos: Ping timeout: 240 seconds
08:05 (quit) vkz: Quit: vkz
08:11 (join) RacketCommitBot
08:11 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/Wz4C0w
08:11 RacketCommitBot: racket/master 2e7f4ba Jay McCarthy: Allowing github URLs from PNRs to use illegal package names
08:11 (part) RacketCommitBot
08:18 (quit) samth_: Ping timeout: 245 seconds
08:19 (join) mizu_no_oto
08:20 (join) tilde`
08:22 (quit) mizu_no_oto: Client Quit
08:41 (join) bitonic
08:46 (quit) tilde`: Ping timeout: 260 seconds
08:51 (join) mizu_no_oto
08:55 (quit) mizu_no_oto: Client Quit
08:56 (join) mizu_no_oto
09:07 (join) vkz
09:08 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
09:14 (join) mizu_no_oto
09:20 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
09:30 (join) anRch
09:30 (quit) bitonic: Ping timeout: 255 seconds
09:34 samth: gf3: you had an es6 question?
09:34 (join) bitonic
09:34 samth: gf3: perhaps you'd like to flame someone about something?
09:34 (quit) rins: Ping timeout: 246 seconds
09:34 (join) francisl
09:40 (quit) bitonic: Ping timeout: 255 seconds
09:52 (join) bitonic
09:59 (join) mithos28
10:03 (join) mizu_no_oto
10:04 (join) Shozan
10:04 (join) sizz_
10:05 (join) fasta_
10:06 (join) dca-
10:06 (join) micro___
10:06 (quit) rmathews: Ping timeout: 256 seconds
10:07 (quit) eataix: Ping timeout: 276 seconds
10:07 (quit) Cryovat: Ping timeout: 276 seconds
10:07 (quit) fasta: Ping timeout: 276 seconds
10:07 (quit) mau: Ping timeout: 276 seconds
10:07 (quit) sizz: Ping timeout: 276 seconds
10:07 (quit) dca: Ping timeout: 276 seconds
10:07 (quit) micro__: Ping timeout: 276 seconds
10:07 (quit) SHODAN: Ping timeout: 276 seconds
10:07 (join) eataix
10:07 (quit) eataix: Changing host
10:07 (join) eataix
10:07 (join) rmathews
10:07 (join) mau_
10:07 (join) Cryovat
10:11 (join) rmathews_
10:11 (quit) rmathews: Ping timeout: 245 seconds
10:11 (nick) rmathews_ -> rmathews
10:14 (quit) tobi: Remote host closed the connection
10:17 (join) rins
10:24 stamourv: mithos28: ping
10:24 mithos28: stamourv: pong
10:24 stamourv: Did you see my potential solution wrt fixnum soundness?
10:25 mithos28: no, where did you send it
10:25 stamourv: In reply to the bug.
10:25 stamourv: Friday morning.
10:25 stamourv: (PR13501.)
10:25 mithos28: ugh, somehow the thread got muted
10:26 stamourv: No-one replied after that.
10:27 (quit) bitonic: Ping timeout: 240 seconds
10:27 mithos28: I think that would still allow the program that breaks it
10:29 stamourv: How so? `z' would only be a subtype of `Integer', so no type-dead code elimination will happen.
10:29 mithos28: It will be of subtype (-val 1000000000)
10:30 (join) bitonic
10:30 mithos28: which when restricted to Fixnum is (Un)
10:30 mithos28: I think you could remove -PosIntNotFixnum entirely in my program and it would still error
10:31 stamourv: With the current subtping relation, yes. I'm proposing to change subtyping by changing the problematic static predicates.
10:31 mithos28: because it is the difference between the static predicate and runtime predicate for the -Fixnum type
10:32 mithos28: You were only proposing changes for the non fixnum types
10:32 stamourv: Right, I see your point.
10:33 stamourv: Well, would also changing the static predicates for fixnum types work?
10:33 mithos28: to what?
10:33 stamourv: (lambda (x) #f)
10:33 mithos28: no, because it returning false is the whole issue
10:33 stamourv: So basically, everything that has a sub-Integer value type, is only a subtype on integer.
10:34 (join) dyoo
10:34 stamourv: These value types would (with the clause I proposed adding to subtyping) all be subtypes of Integer.
10:34 mithos28: The issue is that returning false is not 'I cannot prove this', it is saying 'This is false'
10:35 mithos28: Integer doesn't play into my example at all
10:35 mithos28: and I dont think subtype does either
10:35 mithos28: Ah, we might be able to fix overlap
10:35 stamourv: IIRC (haven't looked at `subtype*' in some time), if the static predicate returns #f, subtyping tries something else.
10:35 mithos28: Subtype might not even get called
10:36 mithos28: restrict is the first one
10:36 stamourv: Oh, you think the problem is in restrict?
10:36 mithos28: which calls overlap
10:36 stamourv: Hadn't thought of that.
10:36 mithos28: the problem is in overlap
10:36 stamourv: I thought it was all about subtyping.
10:37 mithos28: because 10000000 does intersect -Fixnum, but overlap says it doesn't
10:38 stamourv bring up the relevant code.
10:39 mithos28: so overlap calls subtype
10:39 stamourv: Right.
10:39 mithos28: but we can add a special case before then, because overlap can be slightly conservative
10:40 stamourv: You mean adding a special case to `overlap'?
10:40 mithos28: and subtype cannot, as it must be decidable
10:40 mithos28: yes
10:41 stamourv: Not sure that's the right place. This really sounds like a subtyping issue to me.
10:41 stamourv: Why don't you think adding the special case to subtyping would work?
10:42 mithos28: what does (subtype 100000000 Fixnum) return?
10:42 mithos28: If it returns true, then I break soundness on 32 bit machines
10:42 mithos28: If it returns false, I break soundness on 64 bit machines
10:43 stamourv: Ok, I think I see what you mean.
10:43 stamourv: Now, why doesn't Integer have that problem.
10:44 (join) RacketCommitBot
10:44 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/o4ryag
10:44 RacketCommitBot: racket/master b9d1194 Danny Yoo: Correct hyperlinks to 2htdp/image.
10:44 (part) RacketCommitBot
10:44 stamourv is trying to remember how that part of the code works,
10:44 mithos28: Index also has the same issue I think
10:44 stamourv: So basically, IIUC, if we make TR handle (-val 100000000) and co like Integer, we win. Right?
10:45 mithos28: If you cannot make the type (-val 100000000), we win
10:45 mithos28: The bug remains though, just we close up one way to exploit it
10:46 stamourv: You don't think we can fix it and still allow singleton types?
10:47 mithos28: we can, just there are issues with types in the > 2^32 < 2^64 range
10:47 mithos28: I'm not sure what you mean by handling them like integer
10:50 (quit) asvil: Ping timeout: 255 seconds
10:50 mithos28: The core issue is that subtyping is supposed to be decidable, but if we don't know what kind of machine we are running on, then such value types cannot be a subtype of -Fixnum or -PosIntNotFixnum, but only of -Integer
10:52 (quit) jeapostrophe: Ping timeout: 256 seconds
10:54 carleastlund: mithos28, is the problem that you can compile (and typecheck) at one word size and run the resulting code on a different word size?
10:55 stamourv: carleastlund: Yes.
10:56 mithos28: carleastlund: well its that at compile time, I don't know which size will be at runtime
10:56 mithos28: crosscompiling should be possible, but you need to know upfront
10:56 (quit) anRch: Quit: anRch
10:56 stamourv: mithos28: Re Integer: I mean, for subtyping, overlap, etc. we consider that value types for integers >255 are just the Integer type.
10:56 carleastlund: Yeah, TR should have a runtime assertion about word size when compile time has depended on it.
10:56 stamourv: Dunno if that's possible,
10:58 mithos28: stamourv: Subtyping returns a yes/no answer. Returning a no answer, when it should return yes is what is the problem here
10:58 stamourv: Right. So if that type gets to subtyping at all, we lose.
10:58 mithos28: stamourv: Correct
10:58 stamourv: Ok, now I think I understand better.
10:59 mithos28: unless both of the -fixnum and -posintnotfixnum are in the right hand side
10:59 mithos28: Its only if exactly one of them is there
10:59 stamourv: So basically, we have 2 options: (1) Disallow singleton types between 2^30-1 and 2^62-1 or (2) make sure that they can't reach subtyping.
11:00 stamourv: Not sure if (2) is doable, and even if it is, not sure how robust and non-kludgey we can make it.
11:01 mithos28: I think 1 is the better answer to actually implement
11:01 stamourv: I'm starting to lean this way too.
11:01 mithos28: It still leaves the original hole open, just closes an attack vector
11:02 mithos28: Especially since I don't think anyone actual wants types like that
11:02 stamourv: Would be a weird limitation, but I agree, it probably wouldn't be a big loss.
11:03 stamourv: And what do you mean by it would leave the original hole open?
11:03 mithos28: the other option is to attempt 2, and add a case in subtyping to make it an internal error if it did happen
11:03 mithos28: that the static predicates for -Fixnum and -PosIntNotFixnum are incorrect
11:03 mithos28: and cannot possibly be correct
11:04 mithos28: and there may be other ways to use that to break soundness
11:05 stamourv: AFAIK, subtyping is the only place that uses static predicates. But yeah, if another place started using them, that would be a hole.
11:05 (quit) dnolen: Ping timeout: 255 seconds
11:05 stamourv: IIRC, the only reason we even have static predicates is for subtyping of value types.
11:06 stamourv: Re internal error in subtyping: That sounds good too. (I'd rather raise an actual typechecking error, though.)
11:07 mithos28: That looks correct, so it would close all possible attack vectors at this time
11:07 (join) didi
11:08 mithos28: stamourv: That may be possible aswell. The issue is that subtype is below typechecking, in the structure of the code
11:08 mithos28: and so I'm not sure if it makes sense to have it raise typechecking errors
11:08 stamourv: Make sense from a code organisation perspectice, or from a user perspective?
11:09 (quit) em: Ping timeout: 276 seconds
11:09 (quit) mceier: Quit: leaving
11:12 mithos28: from a code organization
11:13 stamourv: Right. I believe some parts of error reporting depend on subtyping. That could be an issue.
11:14 stamourv: Actually, `utils/tc-utils.rkt' doesn't depend on subtyping, so we may be good.
11:14 mithos28: what syntax would you report to the user?
11:15 stamourv: So, which solution do you prefer? Disallowing the problematic value types altogether, or raising an error in subtyping?
11:15 stamourv: Re syntax: Not sure.
11:15 mithos28: raising an error in subtyping
11:15 stamourv: So yeah, would be barely nicer than an internal error. But still a bit nicer.
11:16 stamourv: Alright, I'm on it.
11:16 mithos28: But I believe unless you are trying to break it, you won't run into the error
11:16 stamourv: Agreed.
11:16 stamourv: Thanks a lot for your help! (and your patience :) )
11:16 mithos28: awesome, thanks. Ping me when you have it working and I'll take a look
11:17 mithos28: Unless I'm at work
11:18 stamourv: Will do.
11:24 samth: stamourv, mithos28: i prefer disallowing the singleton types altogether
11:25 mithos28: ok, I'm fine with that
11:25 stamourv: samth: Any reason?
11:26 samth: stamourv: it seems like a valuable invariant that all types are comparable with subtyping
11:26 samth: also, can't we just outlaw 32bit machines?
11:27 stamourv: mithos28: Re sorting for union: ISTR there was a semi-high-level union constructor, that did sorting, but assumed its elements were non-overlapping. Unless it fell prey to refactoring, the numeric tower probably uses it.
11:27 mithos28: well I think TR should support 32 bit machines, as long as racket proper does. If you can convince Matthew to outlaw them, I'm all for moving to the future.
11:27 stamourv: samth: Re invariant: agreed. But whichever solution we pick, the limitation will feel arbitrary.
11:28 stamourv: Re 32 bit machines: I'm using one right now.
11:28 mithos28: stamourv: Yeah, I'm looking for it but found another bug in my search
11:28 mithos28: stamourv: Move to the future
11:28 samth: stamourv: mithos28's patch is a too low a level for that fn
11:28 samth: mithos28: i'm not thinking about not supporting them
11:28 mithos28: samth: But I may be able to move it to a common file.
11:28 samth: or about banning them from racket
11:28 stamourv: mithos28: Re future: That would involve getting a new machine, which would involve spending time to set it up, etc.
11:28 samth: more wishing that the future was here already
11:28 mithos28: samth: I know, sarcasm is hard to express over text
11:29 samth: stamourv: i agree that the limitation feels arbitrary, but i think the maxim here is "make illegal states unrepresentable", as Yaron Minsky put it
11:29 (quit) hkBst__: Quit: Konversation terminated!
11:31 stamourv: samth: I don't think it's that clear cut. These singleton types are perfectly fine, as long as you don't try to check whether they're subtypes of range-limited types.
11:32 samth: stamourv: but the only types that exist are range-limited types and their unions
11:32 stamourv: Hmm, right.
11:32 samth: (plus Any)
11:32 samth: so I don't think there's much use in allowing them
11:32 stamourv: And staying clear of subtyping is pretty hard.
11:32 stamourv: Yeah, I see your point.
11:33 stamourv: Ok, changing gears and implementing that instead.
11:37 asumu: stamourv: what, 32-bit machine? How are you going to use more than 64GB of RAM then!?
11:37 mithos28: asumu: you mean 4GB
11:37 stamourv: I don't. It has 2.
11:37 mithos28: stamourv: And you run racket on it?
11:38 stamourv is happy with his 2007-era mac mini hand-me-down running Debian.
11:38 (quit) otterdam: Ping timeout: 246 seconds
11:38 mithos28: my 2007 macbook with upgraded ram couldn't handle racket and a browser
11:38 asumu: I believe Linux with PAE can do up to 64GB.
11:38 stamourv: mithos28: Until about a year ago, my main laptop was an OLPC XO. The experience taught me a lot about computing frugality.
11:39 stamourv: I could run Racket (not DrRacket, though) + a browser + emacs + minor things just fine in 256mb of RAM.
11:39 stamourv: Everything being 32 bits helped.
11:39 (join) otterdam
11:40 stamourv: Building Racket, though, could take ~12 hours (mostly due to swapping).
11:40 mithos28: which browser? Chrome wants more than that for some of my tabs
11:41 stamourv: Firefox 3.5x. Conkeror worked fine too.
11:41 stamourv: I tend to disallow most js, which helps.
11:43 asumu: But then how are you going to run all those newfangled sites using EmberJS. :p
11:43 (join) anonymous
11:43 didi: It totally helps. It made my, now defunct, laptop go from _burning_ to just working.
11:43 (nick) anonymous -> Guest96181
11:43 Guest96181: Hello everyone!
11:43 asumu: Hallo.
11:46 anonymou2: (##grammar said you're wrong about the word) Tell me, please > (andmap string? '())
11:46 anonymou2: #t
11:46 anonymou2: rudybot: (andmap string? '())
11:46 rudybot: anonymou2: your sandbox is ready
11:46 rudybot: anonymou2: ; Value: #t
11:46 anonymou2: is it ok?
11:46 anonymou2: is everything works great?
11:47 mithos28: anonymou2: All elements in that list are strings, so it returns true
11:47 anonymou2: sorry... If the lsts are empty, then #t is returned.
11:47 anonymou2: i had to RTM...  I am sorry.
11:47 anonymou2: http://docs.racket-lang.org/reference/pairs.html#%28def._%28%28lib._racket%2Fprivate%2Fmap..rkt%29._andmap%29%29
11:47 rudybot: http://tinyurl.com/aorvyeg
11:48 asumu: Also it's a general logic thing.
11:48 asumu: The negation of (andmap string? '()) is (ormap (lambda (x) (not (string? x))) '())
11:49 asumu: Which is #f, so this must be #t.
11:49 (join) jonrafkind
11:49 (quit) jonrafkind: Changing host
11:49 (join) jonrafkind
11:50 anonymou2: rudybot: (ormap string? '())
11:50 rudybot: anonymou2: ; Value: #f
11:50 anonymou2: rudybot: (ormap (lambda (x) (not (string? x))) '())
11:50 rudybot: anonymou2: ; Value: #f
11:51 anonymou2: (ormap (lambda (x) (not (an-owner-of-universe? x))) '())
11:51 anonymou2: (andmap an-owner-of-universe? '())
11:51 anonymou2: nevermind
11:53 (quit) Demosthenex: Quit: leaving
11:57 mithos28: stamourv: You make need to go down to 2^28 because of Indexes
11:57 dyoo: anonymou2: for similar reasons, (*) is 1, (+) is zero, (and) is #t, (or) is false.  It's the identity of the particular operation.
11:58 stamourv: mithos28: Yep, I remembered. ;)
11:58 (join) Kaylin
11:58 (join) tcsc
12:01 didi didn't know about (and) and (or)
12:02 soegaard: http://stackoverflow.com/questions/11979683/why-python-built-in-all-function-returns-true-for-empty-iterables
12:02 rudybot: http://tinyurl.com/936se3a
12:02 asumu should s/#t/(and)/ s/#f/(or)/ in all his code
12:02 (join) jeapostrophe
12:03 soegaard: Also: http://en.wikipedia.org/wiki/Vacuous_truth
12:03 stamourv: asumu: Similarly, you can use `list*' as the identity function.
12:03 stamourv: rudybot: (map list* '(1 2 3))
12:03 rudybot: stamourv: your sandbox is ready
12:03 rudybot: stamourv: ; Value: (1 2 3)
12:03 (quit) mizu_no_oto: Quit: ["Textual IRC Client: www.textualapp.com"]
12:03 asumu: Also this:
12:03 stamourv: That one's kind of evil, though.
12:04 asumu: rudybot: (require (for-meta 0+0i unstable/contract))
12:04 rudybot: asumu: your sandbox is ready
12:04 rudybot: asumu: Done.
12:04 asumu: rudybot: maybe/c
12:04 rudybot: asumu: error: maybe/c: undefined; <NEWLINE>  cannot reference an identifier before its definition <NEWLINE>   in module: 'program
12:04 asumu: rudybot: option/c
12:04 rudybot: asumu: ; Value: #<procedure:option/c>
12:04 (join) anonus
12:05 (part) anonus
12:06 stamourv: rudybot: (require (for-meta 27/3 racket/base))
12:06 rudybot: stamourv: Done.
12:06 soegaard: !?!
12:06 stamourv: Now, to write something that uses phase 27/3.
12:06 soegaard: (require (for-meta 335/113 racket/base))
12:06 stamourv: soegaard: It's not as insane as it looks. That's really just phase 9.
12:07 stamourv: (which, agreed, is somewhat insane)
12:07 stamourv: soegaard: That one won't work.
12:07 soegaard: rudybot: (require (for-meta 335/113 racket/base))
12:07 rudybot: soegaard: your sandbox is ready
12:07 rudybot: soegaard: error: #<undefined>:1:19: for-meta: phase level must be #f or an exact integer <NEWLINE>   at: 335/113 <NEWLINE>   in: (for-meta 335/113 racket/base)
12:07 mithos28: hey, phase 9 is not isane
12:07 mithos28: racket uses phase 16
12:08 soegaard: It is *very* rare I need phase 2.
12:08 stamourv: Not directly, though. It does it via chains of requires.
12:08 mithos28: stamourv: right
12:08 stamourv: I think the highest in the codebase is `for-meta' 3.
12:08 stamourv: IIRC, it's somewhere in Honu. I forget what it was for.
12:10 (join) anRch
12:11 soegaard: I needed phase two to implement var parameters (for Pascal): https://github.com/soegaard/minipascal/blob/master/minipascal/experiments/define-function.rkt
12:11 rudybot: http://tinyurl.com/bc6a5gv
12:11 (join) eikonos
12:11 (quit) eikonos: Changing host
12:11 (join) eikonos
12:11 soegaard: Maybe to implement Honu macros?
12:15 (join) mceier
12:16 jonrafkind: yea theres a for-meta 3 in honu. a honu macro (phase 1) is generated by a racket macro (phase 2) which needs phase 3 racket to operate
12:16 jonrafkind: actually its for pattern classes, not macros. but same difference
12:21 anonymou2: I am sorry, ##english told me that you were wrong and I was right!
12:24 anonymou2: or, if speak more correctly, a man said I am right, anothet one - it all is some different. But that the meaning is incorrect - no-one told it clearly. One asked to define, what is correct.
12:24 anonymou2 feels cool
12:25 anonymou2: ...but said that you're vrong very clearly
12:25 anonymou2 stopped offtopic
12:32 stamourv: mithos28, samth: Got it working. Will push when I come back from lunch. (And if tests pass, of course.)
12:32 mithos28: stamourv: Is it on github?
12:33 stamourv: No. I can send you the patch, if you want?
12:33 mithos28: sure
12:34 stamourv: Sent.
12:34 mithos28: got it thanks
12:34 stamourv: All the tests pass, so if that looks good to you, I'll push when I come back from lunch.
12:36 (quit) bitonic: Ping timeout: 255 seconds
12:37 mithos28: LGTM
12:42 (join) dauterive
12:48 (quit) Kaylin: Ping timeout: 244 seconds
12:51 (quit) Shviller: Read error: Connection reset by peer
12:53 (join) gridaphobe
12:53 (join) Shviller
12:54 (quit) anRch: Quit: anRch
13:23 (quit) jeapostrophe: Ping timeout: 276 seconds
13:26 (join) mizu_no_oto
13:30 (join) ASau
13:34 (join) mye
13:38 mye: How do I run raco setup on all raco link'ed collection again? I'm not sure if I  knew this
13:38 (join) sysopfb
13:39 mye: Or alternatively, what command do you folks use after a recompilation to bring your compiled/ folder up to date?
13:40 samth: mye: there's no way to run it on *just* collections created by raco link
13:40 samth: but you can just run `raco setup` to run it on all collections
13:42 mye: samth: yes but that's what I was trying to avoid. But I just did and it actually didn't rebuild all the built-in collections like it it before which I think was a bug. So I'm good to go, I was just under the wrong impression it would rebuild everything :-)
13:42 samth: mye: it should only rebuild things that need rebuilding
13:43 (join) trep
13:45 (join) Kaylin
13:46 (quit) sysopfb: Ping timeout: 260 seconds
13:49 (quit) dyoo: Quit: dyoo
13:50 (join) dyoo
13:50 (quit) dyoo: Client Quit
13:52 (quit) gridaphobe: Remote host closed the connection
13:54 (join) gridaphobe
14:04 (join) bitonic
14:04 (join) rbarraud__
14:22 (join) jao
14:22 (quit) jao: Changing host
14:22 (join) jao
14:22 (join) anonymous
14:23 (nick) anonymous -> Guest10037
14:29 (quit) bitonic: Ping timeout: 240 seconds
14:31 (join) RacketCommitBot
14:31 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/mQcQbw
14:31 RacketCommitBot: racket/master 0bccf6b Vincent St-Amour: Forbid singleton types for integers whose type is platform-dependent....
14:31 (part) RacketCommitBot
14:39 (join) bitonic
14:42 (quit) bitonic: Remote host closed the connection
14:42 (join) bitonic
14:46 (quit) Kaylin: Ping timeout: 260 seconds
14:54 (quit) sirdancealot: Ping timeout: 276 seconds
14:54 (quit) eataix: Quit: ZNC - http://znc.in
14:56 (join) ASau`
14:58 (quit) ASau: Ping timeout: 240 seconds
14:59 (quit) gridaphobe: Remote host closed the connection
15:00 (join) gridaphobe
15:04 (quit) gridaphobe: Ping timeout: 255 seconds
15:06 (join) sirdancealot
15:06 (nick) ASau` -> ASau
15:09 anonymou2: ls
15:09 anonymou2: sorry
15:25 (quit) mizu_no_oto: Quit: ["Textual IRC Client: www.textualapp.com"]
15:27 (join) jeapostrophe
15:27 (quit) jeapostrophe: Changing host
15:27 (join) jeapostrophe
15:30 mye: creating a custom web server configuration is hard, man :-/
15:31 mye: all those units and config signatures and server functions confuse me
15:38 (join) Nisstyre
15:38 (join) gridaphobe
15:39 (quit) cdidd: Ping timeout: 276 seconds
15:39 (quit) jonrafkind: Ping timeout: 252 seconds
15:44 (join) acarrico
15:45 (join) Kaylin
15:47 (quit) gridaphobe: Ping timeout: 248 seconds
15:47 (quit) Kaylin: Client Quit
15:48 (join) bjz
15:51 (join) Kaylin
15:54 (quit) vkz: Quit: vkz
16:10 (quit) bitonic: Ping timeout: 240 seconds
16:12 (quit) merijn: Quit: leaving
16:14 (join) dyoo
16:15 (join) merijn
16:24 gf3: samth: No flames ❤
16:24 gf3: samth: Just curiosity
16:24 gf3: samth: And it doesn't even have to do with modules!
16:24 dyoo: hmmm?  I missed the conversation
16:25 dyoo: gf3: what's the subject?
16:25 gf3: dyoo: Comprehension/generator syntax changes in ES6
16:30 bremner: hrm. It seems (case ...) on a union type confuses the TR typechecker when an equivalent cond does not.
16:31 dyoo: bremner: yeah, I ran into that problem in Whalesong.
16:31 dyoo: needed to switch over to cond instead.
16:34 (quit) Kaylin: Quit: Leaving.
16:45 soegaard: gf3: Just found http://wiki.ecmascript.org/doku.php?id=harmony:array_comprehensions   I wish they would use "in" instead of "of".
16:45 (join) anRch
16:54 gf3: soegaard: They're slightly different now
16:54 soegaard: Where should I look?
16:57 gf3: soegaard: One sec, I'll grab the notes for you
16:57 (join) gridaphobe
16:57 gf3: soegaard: https://github.com/rwldrn/tc39-notes/blob/master/es6/2013-01/jan-31.md#comprehensionsgenerator-syntax
16:57 rudybot: http://tinyurl.com/bhufm5n
17:01 soegaard: thanks
17:02 gf3: I have no issues with the new syntax, just curious about the why part, samth
17:03 samth: gf3: the why as to why we switched the syntax?
17:03 dyoo: gf3: it looks like the overview in http://wiki.ecmascript.org/doku.php?id=harmony:iterators says why "of"
17:04 samth: gf3: basically, b/c [for (x in [1,2,3]) x] looks better than [x for x in [1,2,3]]
17:06 gf3: samth: As in more "javascript-y"?
17:06 samth: gf3: just nicer
17:06 gf3: samth: Cool
17:06 gf3: samth: Also great work
17:07 gf3: samth: I'm sure I've said this before, but if it were up to me I'd lock you, Dave Herman, and Brendan Eich in a room and let you have your way with the spec
17:08 samth: gf3: :)
17:43 (quit) anRch: Quit: anRch
17:46 (join) jonrafkind
18:00 (quit) rins: Ping timeout: 256 seconds
18:04 (quit) gridaphobe: Remote host closed the connection
18:05 (quit) snorble_: Ping timeout: 245 seconds
18:10 (join) gridaphobe
18:10 (join) zeom
18:21 (join) mizu_no_oto
18:22 (quit) jao: Ping timeout: 245 seconds
18:29 (quit) didi: Ping timeout: 255 seconds
18:30 (quit) gridaphobe: Remote host closed the connection
18:31 (join) gridaphobe
18:31 (join) bitonic
18:36 (quit) gridaphobe: Ping timeout: 248 seconds
18:38 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
18:38 (join) gridaphobe
18:40 (quit) tcsc: Quit: bye!
18:48 (quit) sirdancealot: Ping timeout: 255 seconds
18:50 (quit) francisl: Quit: francisl
18:55 (join) tcsc
18:56 (quit) gridaphobe: Read error: Connection reset by peer
18:57 (join) gridaphobe
19:10 (nick) trep -> sysop_fb
19:20 (join) dnolen
19:37 (quit) soegaard: Quit: soegaard
19:44 (quit) bitonic: Remote host closed the connection
19:58 asumu: Hmm, anyone involve with Bootstrap know if it has a Twitter feed?
19:58 asumu: (and if not, it should! :))
19:59 dyoo: it's got a google+
19:59 dyoo: asumu: https://plus.google.com/115036800423679915320/posts
20:00 asumu: dyoo: Thanks, I also know they're on FB.
20:00 asumu: But it'd be nice if the Racket twitter feed could re-tweet them.
20:00 dyoo: I'll send a note to emmanuel about it right now; give me a sec...
20:01 dyoo: asumu: ok, email sent with you in CC
20:02 (join) netrino
20:02 dyoo: asumu: wait, I just realized I made an assumption.  Are you responsible for the Racket twitter feed, or is that someone else?
20:03 asumu: dyoo: Stevie and I are, yes.
20:03 dyoo: ah, ok.  :)
20:03 asumu: So no problem. :)
20:08 dyoo: I really want the wiki site to be linked up from http://racket-lang.org/community.html
20:08 dyoo: Anything I can do to make that happen?
20:09 samth: dyoo: do you have access to pltgit:iplt?
20:10 dyoo: ah, it's in iplt then?  I think I do, but I've never touched it.
20:10 asumu: samth: is the pkg.racket-lang.org stuff in there too? I wanted to make it fit the site scheme.
20:11 samth: asumu: i believe that that's either in collects/meta or nowhere
20:12 asumu: I hope the former. *checks*
20:13 dyoo: so if I push on iplt, does it automatically deploy to the web site then?
20:14 asumu: (ah, I see something in collects/meta that looks promising)
20:14 samth: dyoo: i don't think so -- you may need to contact eli
20:14 asumu: (but I have no idea how to test it really...)
20:15 dyoo: I'm still feeling pretty stupid.  I'm looking through iplt, and see nothing about community.html
20:16 asumu: I think you might want collects/meta/web
20:16 asumu: www/community.rkt
20:17 dyoo: am I even looking at the right thing then?  I did: git clone pltgit:iplt
20:17 samth: asumu: ah yes
20:17 samth: dyoo: see asumu
20:17 dyoo: and the only directories I see are: dns, release, and web
20:17 dyoo: none of which have collects/meta/web anywhere in the tree
20:18 asumu: And given that there's no README explaining this stuff, no need to feel stupid. Good thing I don't grade the Racket codebase, software dev style. :p
20:18 asumu: dyoo: in normal plt, not iplt.
20:18 asumu: (sorry for being unclear)
20:19 dyoo: not being unclear.  I'm simply getting conflicting information.  :)
20:19 (join) jao
20:19 (quit) jao: Changing host
20:19 (join) jao
20:22 samth: dyoo: i was wrong, and asumu was right, and it's in plt/collects/meta
20:23 dyoo: samth: ok, found it.  Thanks.  Will put something right between discussion channel and racketcon entries
20:28 (join) anonymou2
20:30 (quit) jeapostrophe: Ping timeout: 252 seconds
20:38 dyoo: samth: ok, pull request sent.  Want to make sure it gets reviewed before merge upstream.  https://github.com/plt/racket/pull/238
20:39 samth: dyoo: jfdi
20:39 samth: dyoo: r+=m
20:39 dyoo: You're making me google these abbreviations.
20:40 jonrafkind: lol is that the l33t version of rtfm
20:40 samth: jonrafkind: no
20:41 asumu: JFDI = Just Fudging Do It. ;)
20:41 samth: that should have been r+=me
20:41 samth: it's saying that i give it an r+
20:41 samth: which is a positive review
20:41 asumu: r is something and you're +=ing it?
20:43 samth: no
20:43 samth: r+ is something
20:43 samth: as in, as opposed to r-
20:43 dyoo: it's like append, right?
20:43 samth: no
20:43 samth: never mind
20:43 (quit) gridaphobe: Remote host closed the connection
20:44 dyoo: oh anyway.  Pushed.  :)
20:44 (join) gridaphobe
20:48 (quit) gridaphobe: Ping timeout: 260 seconds
20:51 (nick) samth -> samth_away
20:51 (quit) dauterive: Quit: Leaving
21:07 (quit) netrino: Remote host closed the connection
21:13 dyoo: whoa!  Who changed the splash page for drracket?  http://www.drracket.org/
21:15 asumu: Oh, I think it's been like that for a while.
21:15 asumu: And I think the culprit might be eli.
21:16 dyoo: ah, ok.  Not complaining: I like it when links don't break.  Just didn't expect the redirect at first.
21:17 (quit) dyoo: Quit: dyoo
21:23 asumu: rudybot: eval (require srfi/19) (date->time-tai (make-date 0 0 0 0 1 1 1970 0))
21:23 rudybot: asumu: your sandbox is ready
21:23 rudybot: asumu: ; Value: #(struct:tm:time time-tai 0 0)
21:23 asumu: rudybot: eval (require srfi/19) (date->time-tai (make-date 0 0 0 0 1 1 1958 0))
21:23 rudybot: asumu: ; Value: #(struct:tm:time time-tai 0 -378691200)
21:31 (quit) dnolen: Ping timeout: 256 seconds
21:43 (join) sysopfb
22:05 (quit) bjz: Quit: Leaving...
22:24 (nick) sysopfb -> sysop_fb
22:25 (join) didi
22:34 (join) chandler_
22:34 (quit) chandler_: Changing host
22:34 (join) chandler_
22:34 (quit) chandler: Disconnected by services
22:34 (nick) chandler_ -> chandler
22:35 (join) cdidd
22:55 (quit) anonymou2: Ping timeout: 245 seconds
22:57 (quit) mye: Quit: mye
23:02 (join) anonymous
23:02 (nick) anonymous -> Guest98550
23:05 (quit) tcsc: Quit: computer sleeping
23:08 (join) rmathews
23:25 (quit) CoverSlide: Ping timeout: 256 seconds
23:28 (quit) eikonos: Quit: Leaving.
23:30 (join) yixiza
23:31 (join) tcsc
23:32 (join) CoverSlide
23:33 (part) yixiza
23:42 (join) mizu_no_oto
23:51 (quit) didi: Ping timeout: 252 seconds
23:54 (quit) mizu_no_oto: Quit: Computer has gone to sleep.
23:55 (join) mizu_no_oto
23:59 (join) eikonos
23:59 (quit) eikonos: Changing host
23:59 (join) eikonos