00:00 (quit) mykhal: Quit: leaving
00:22 (join) dmac
00:25 (quit) jonrafkind: Ping timeout: 252 seconds
00:26 (join) jonrafkind
00:37 (quit) jonrafkind: Ping timeout: 240 seconds
00:44 (join) jonrafkind
00:56 (quit) pcavs: Quit: Leaving.
00:57 (quit) jonrafkind: Ping timeout: 252 seconds
01:05 (join) jonrafkind
01:06 (join) veer_
01:06 (quit) dmac: Ping timeout: 260 seconds
01:09 (quit) jao: Ping timeout: 276 seconds
01:14 (quit) jonrafkind: Ping timeout: 260 seconds
01:16 (join) jonrafkind
01:35 (quit) jonrafkind: Ping timeout: 260 seconds
01:41 (join) jonrafkind
01:49 RacketCommitBot: [racket] plt pushed 2 new commits to master: http://git.io/K8fvig
01:49 RacketCommitBot: [racket/master] Some racketization; rearrange and fix "client-gui.rkt" which had some - Eli Barzilay
01:49 RacketCommitBot: [racket/master] Use "Submit to" instead of "Assignment" for the field name - Eli Barzilay
01:53 (quit) jonrafkind: Ping timeout: 240 seconds
01:59 (join) jonrafkind
02:06 (quit) jonrafkind: Ping timeout: 260 seconds
02:10 mithos28: somehow I am reversing the bits of doubles when communicating with the ffi layer, anyone know what I am doing wrong?
02:11 eli: mithos28: ?
02:12 mithos28: I have an ffi function which takes a double and writes it to an flvector
02:13 mithos28: when I pull the value out of the flvector it is a much different value
02:13 eli: You didn't mix up floats and doubles?
02:13 mithos28: repeating this operation gets me the original value, so I assume somehow the bits are getting put in the wrong order
02:14 mithos28: maybe I should only be using doubles
02:14 eli: So you did use floats?
02:14 mithos28: flvectors use doubles, correct?
02:14 eli: Yes, I think so.
02:15 mithos28: I am not trying to use floats, I'll look again but I don't think I have
02:15 eli: You can try to typecast to bytes or something, so you'll be able to inspect the bits.
02:16 eli: mithos28: Weren't you doing libgit2?
02:16 mithos28: no
02:16 eli: So what is it?
02:17 mithos28: that was joan arnaldich
02:17 mithos28: I'm working on LLVM
02:17 eli: LLVM?  Why?
02:18 eli: Not trying to flame, but when we interfaced it we concluded that it's much easier to just construct text in the LLVM language and send that over.
02:18 (quit) veer_: Read error: Connection reset by peer
02:18 (quit) dnolen: Quit: dnolen
02:19 mithos28: I haven't had too much trouble, I used the first draft of my library to write a compiler for my class last spring.
02:19 mithos28: That is currently my only client of the library, so I don't have much else to go by.
02:19 eli: Trouble is not the thing that I was referring to...
02:19 (join) veer
02:20 mithos28: explain?
02:20 eli: Just seemed that going via text means that it's easier to actually get things working.
02:20 (nick) veer -> veer_
02:20 eli: IIRC, most of their API was very complex.
02:20 eli: At least at that time.
02:20 eli: Also, it was only a C++ thing, which made things more impossible.
02:21 mithos28: There are c-extern bindings now
02:21 eli: OK, so maybe those are easier to handle.
02:21 eli: Does the interface have commands to emit instructions?
02:22 mithos28: I have the whole API converted to racket already, and am currently trying to make it memory safe
02:22 mithos28: yes you can write modules out to files
02:22 mithos28: There is also a jit, which I had working and has bit rotted (that is the issue with the floats currently)
02:23 eli: So basically the difference is between things like (emit-foo 1 2) and (printf "foo(1,2)")?
02:23 mithos28: No.
02:23 mithos28: You build up a module as a data structure, and then emit that.
02:23 mithos28: Inside the basic blocks it is close to (emit-foo 1 2)
02:24 eli: So even memory-wise there's not much difference between that and going via text?
02:26 mithos28: they are probably similar
02:27 eli: BTW, you could play with the code that we had then, it had some cute ideas.
02:28 mithos28: where is it?
02:29 eli: http://svn.plt-scheme.org/plt/branches/alex/
02:29 eli: The guy that wrote it disappeared, but the nice idea there was to use a kind of a HOAS approach to syntax representation.
02:30 eli: IOW, a function was represented by code that actually binds its input.
02:31 mithos28: HOAS has always confused me
02:32 eli: At least in this context it's pretty simple to understand.
02:33 eli: For example, if you want to translate some "assembly" code with let-like meaning (asm-let ([x 1]) ...code...)
02:33 mithos28: I never got how it was amenable to anything but evaluation, like inspection because everything was hidden inside the meta languages functions
02:33 eli: where in code you want to be able to mutate `x'
02:34 eli: then what you do is make `asm-let' be a piece of syntax that emits whatever is needed to allocate a temporary variable,
02:34 eli: and then the thing is that mutations of the variable need to refer to it by its address,
02:35 eli: so you make that `asm-let' also bind `x' -- at the racket level -- to whatever information is needed to refer to the llvm binding later.
02:35 mithos28: I'll look into that when I get to creating the nice racket API, currently I am stuck with the C transliteration
02:35 eli: This was even cuter with a letrec-like binder, which was implemented by a form with a letrec-like scope.
02:36 eli: The main thing that it gets you is that you don't have to deal with symbols, or gensyms, or all that stuff.
02:37 eli: The student was initially very much against doing this, then tried it just because I insisted that it will be a good idea, and then got all excited about it since it made compilation way faster.
02:37 eli: (Before that he worked with syntax values.)
02:38 mithos28: I've been mulling in my head how I envision the racket API to look and nothing has seemed good. Maybe I should take your suggestion aswell
02:38 mithos28: so my floating bits are not flipped, but the bytes are
02:40 eli: Re that API, yes, I'll be happy if that approach is revived.  (The code is probably useless otherwise.)
02:40 eli: Re the bytes -- you should try to do the same from C, perhaps there's something in LLVM that does that.
02:41 mithos28: maybe, It looks like an endianness problem
02:41 (join) masm
02:54 mithos28: It was an llvm bug, the default endianness was wrong
02:58 (quit) mithos28: Quit: mithos28
03:05 veer_: do we have any Cryptography library that works on all platform?
03:06 veer_: "we" means racket here
04:04 (part) veer_: "Leaving"
04:13 (join) noelw
04:15 (join) veer_
04:16 veer_: noelw:  how do you charge your client , per hour or in bulk?
04:17 noelw: veer_: typically fixed price by job
04:17 noelw: occasionally by hour
04:18 veer_: noelw: how large are the projects?
04:18 noelw: varies a lot. days to months
04:20 veer_: noelw: thanks for answers :)
04:20 noelw: np. why do you ask?
04:21 veer_: I was thinking of doing small time consultancy
04:22 noelw: Ah ok. Good luck with it.
04:23 veer_: thanks
05:43 (join) _danb_
05:48 (join) Burlingk
06:12 (quit) masm: Ping timeout: 260 seconds
06:13 (quit) veer_: Quit: Leaving
06:59 (join) hkBst
06:59 (quit) hkBst: Changing host
06:59 (join) hkBst
07:06 (join) mceier
07:15 (join) masm
07:18 (join) veer
07:33 (quit) mceier: Ping timeout: 240 seconds
07:41 (quit) Burlingk: Quit: Leaving
08:19 (join) ahinki
08:32 (join) dmac
08:36 (join) realitygrill_
08:37 (quit) realitygrill: Ping timeout: 260 seconds
08:37 (nick) realitygrill_ -> realitygrill
08:38 (join) pcavs
08:56 (quit) _danb_: Quit: ERC Version 5.3 (IRC client for Emacs)
08:59 (quit) dmac: Ping timeout: 260 seconds
09:03 (join) jao
09:08 (join) bluezenix
09:08 (quit) pcavs: Quit: Leaving.
09:15 (quit) jao: Ping timeout: 240 seconds
09:41 (quit) veer: Quit: Leaving
09:52 (join) pcavs
10:10 (nick) samth_away -> samth
10:21 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/aX1stA
10:21 RacketCommitBot: [racket/master] fix problems in `copy-file'; add `exists-ok?' argument - Matthew Flatt
10:32 (join) sstrickl
10:38 (quit) realitygrill: Quit: realitygrill
10:44 (quit) ahinki: Ping timeout: 258 seconds
10:45 (join) ahinki
10:47 (quit) hkBst: Remote host closed the connection
10:48 (quit) sstrickl: Quit: sstrickl
11:00 (join) jao
11:01 (join) veer
11:08 (part) ahinki
11:17 (join) samth_
11:17 (quit) samth_: Client Quit
11:19 (quit) bluezenix: Quit: Leaving.
11:22 (join) sstrickl
11:24 (quit) sstrickl: Client Quit
11:24 (join) sstrickl
11:24 (join) shofetim
11:29 veer: I am having problem with parameter and type :
11:29 veer: (: document-path (Parameterof (U Path-String False)))
11:30 veer: (define document-path (make-parameter #f))
11:30 veer: (call-with-input-file (document-path) parse-xml)
11:30 veer: does not work
11:32 noelw: Presumably you need to ensure that (document-path) returns a Path-String?
11:33 noelw: Oh, and if you want advice about the consulting feel free to email (FWIW)
11:33 veer: I am making sure (documet-path) is not false using cond
11:34 veer: sure
11:34 noelw: Try binding the value of (document-path) to a name, and performing the conditional on that name
11:35 noelw: In a multi-threaded program the value of (document-path) could change in the time you inspect it in the conditional and use it in the call-with-input-file expression
11:36 noelw: Thus I don't think Typed Racket should allow that.
11:36 veer: noelw: thanks binding to a name worked
11:36 noelw: \o/
11:46 veer: good point on multi-thread , so much to learn
11:54 (join) jonrafkind
11:58 veer: I am writing xpath implementation ,initial version is at  https://github.com/veer-public/xpath ,  a experiment with typed racket
11:58 veer: xml is pervasive
11:58 jonrafkind: thats excellent, i wanted xpath just last week
12:00 (quit) noelw: Quit: noelw
12:02 jonrafkind: veer, 'eval-xpath' doesn't seem like the greatest name
12:03 (join) superjudge
12:03 veer: jonrafkind : ok , I'll think of new name , any suggestion :)
12:04 jonrafkind: hm.. xpath-query?
12:04 jonrafkind: do you compile the strings into some internal "xpath" structure?
12:05 veer: ok , so it will be xpath-query
12:05 veer: yes
12:05 jonrafkind: ok, probably xpath-query should take an extra argument that is the document, which is (current-document) if not specified so you can do the parameterized thing
12:06 veer: ok , i can do that
12:07 (quit) superjudge: Client Quit
12:07 veer: btw at present it only support child axis , will add other's soon
12:17 (join) superjudge
12:22 (join) mceier
12:24 (quit) jao: Ping timeout: 260 seconds
12:34 (quit) veer: Quit: Leaving
12:55 (join) ahinki
13:02 (join) anRch
13:03 (quit) sstrickl: Read error: Connection reset by peer
13:03 (join) sstrickl_
13:04 (join) jao
13:11 (join) blake_johnson
13:19 (join) bluezenix
13:19 (quit) bluezenix: Client Quit
13:19 (join) MayDaniel
13:21 (join) veer
13:21 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/IluUHw
13:21 RacketCommitBot: [racket/master] Add evaluation contexts for by-need letrec calculus as an example - Casey Klein
13:22 veer: jonrafkind: I am not compiling string into structure at compile time , but at runtime
13:22 veer: I got confused , sorry
13:23 jonrafkind: ok yea thats what i thought was happening anyway
13:23 jonrafkind: the point is you are not just passing strings around, you pass some internal representation
13:24 veer: right
13:26 veer: now I goto sleep , bye :)
13:26 (quit) veer: Quit: Leaving
13:46 (quit) jao: Ping timeout: 258 seconds
14:00 (quit) anRch: Ping timeout: 243 seconds
14:08 (join) jao
14:30 (quit) jao: Ping timeout: 260 seconds
14:34 jonrafkind: how do I fix this warning? WARNING: no declared exporting libraries for definition in: honu-read
14:36 stamourv: Are you requiring `honu-read' for label?
14:36 jonrafkind: (require (for-label honu/core/read))
14:37 stamourv: Then I don't know.
14:39 (join) jao
14:39 jonrafkind: stamourv, do you know where eli is?
14:42 stamourv: Hold on, lemme see if his door is open.
14:42 stamourv: Nope, and no lights.
14:43 (nick) sstrickl_ -> sstrickl
14:43 jonrafkind: ok the solution for my scribble error was to put @defmodulelang before @section
14:50 (quit) jao: Ping timeout: 260 seconds
14:51 (quit) asumu: Read error: Operation timed out
14:57 (join) ChibaPet
14:58 (join) asumu
15:23 (quit) superjudge: Quit: superjudge
15:31 (part) janne
15:32 (quit) MayDaniel: Read error: Connection reset by peer
15:44 (quit) ahinki: Remote host closed the connection
15:44 (quit) blake_johnson: Quit: blake_johnson
15:48 (join) MayDaniel
15:56 (join) blake_johnson
16:03 (quit) blake_johnson: Quit: blake_johnson
16:19 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/DXQgDA
16:19 RacketCommitBot: [racket/master] use new copy-file argument that Matthew provided; this is a better fix than - Robby Findler
16:26 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/3e1qzg
16:26 RacketCommitBot: [racket/master] windows: fix `copy-file' changes - Matthew Flatt
16:27 (quit) ChibaPet: Quit: Leaving.
16:31 (join) ChibaPet
16:32 (join) jao
16:32 (quit) jao: Changing host
16:32 (join) jao
16:58 (quit) sstrickl: Quit: sstrickl
17:01 (quit) ChibaPet: Quit: Leaving.
17:11 (quit) MayDaniel: Read error: Connection reset by peer
17:12 (quit) jao: Ping timeout: 258 seconds
17:51 (join) anRch
17:54 (nick) Nanakhiel -> Lajla
18:11 (join) jobar
18:12 jobar: hi, how to configure racket to build with clang?
18:13 jobar: there is an error in libffi
18:14 jobar: 'invalid variant 'rel''
18:14 jonrafkind: hm I think I got clang to work at some point
18:14 jonrafkind: how did you set it up? CC=clang ?
18:14 jobar: yes
18:15 jobar: I'm using debian
18:16 jonrafkind: can you paste a the last few lines of output (in pastebin)
18:16 jobar: ok
18:17 jobar: I have to sign up
18:17 samth: try paste.lisp.org
18:18 samth: also, you don't have to sign up -- just past the text and hit submit
18:18 jonrafkind: you dont have to sign up for pastebin
18:19 jobar: http://pastebin.com/43PKkjF7
18:19 jobar: ok
18:19 samth: jobar: do you have a system libffi?
18:19 jobar: can you see?
18:20 jobar: yes, debian
18:20 samth: then configure should have used that
18:21 jobar: I'm installing that now
18:21 jamessan: libffi-dev is the package you need
18:21 jobar: yes
18:21 jobar: its done
18:22 jobar: I'm configuring now
18:24 jobar: tanks
18:25 jamessan: that gets past configure, but then clang quickly bails out from too many errors :)
18:25 (nick) samth -> samth_away
18:25 samth_away: you might need to enable -horrible-gnu-extensions, or something like that
18:25 jamessan: yeah, -fheinous-gnu-extensions
18:25 jobar: ok, its compilling
18:27 jonrafkind: oh whats this now?
18:27 jonrafkind: you have to add that to CFLAGS ?
18:27 jamessan: src/racket/src/gmp/gmp.c:3486:8: error: invalid use of a cast in a inline asm context requiring an l-value: remove the cast or build with -fheinous-gnu-extensions
18:27 jonrafkind: so did you add it to CFLAGS ?
18:28 jamessan: yeah
18:28 jobar: yes thats it
18:29 jamessan: actually, CPPFLAGS
18:29 jonrafkind: er, really? i thought all -f flags were stricly compiler related (as opposed to pre-processor)
18:30 jonrafkind: unless you meant CXXFLAGS
18:32 jamessan: CXXFLAGS was probably better. CFLAGS didn't work.
18:33 jamessan: I didn't try CXXFLAGS since I didn't the a c++ compiler was being used
18:35 jobar: configure CPPFLAGS=-fheinous-gnu-extensions
18:36 (quit) anRch: Quit: anRch
18:49 (join) tunes
18:50 (nick) tunes -> Fare
18:51 (join) jao
18:55 (join) blake_johnson
18:57 (quit) blake_johnson: Client Quit
19:03 (quit) tauntaun: Read error: Connection reset by peer
19:16 (join) realitygrill
19:19 (quit) masm: Quit: Leaving.
19:30 (part) jobar
19:37 (quit) Fare: Quit: Leaving
19:46 (quit) mceier: Quit: leaving
20:28 (quit) jao: Ping timeout: 260 seconds
21:04 (quit) wtetzner: Read error: Connection reset by peer
21:22 (quit) jonrafkind: Ping timeout: 260 seconds
21:23 (join) jao
21:49 (join) jntrnr13
21:51 jntrnr13: is there a way to match a symbol defined outside of a pattern inside the pattern?  say I want to match the contents of a variable in the pattern
21:56 jntrnr13: (there may be long ways to do it, but I was curious if there were canonical ways)
22:00 (join) Demosthenes
22:02 Demosthenes: anyone have suggestions on using a lexer to parse paragraph style output, like from ifconfig?
22:03 (quit) pcavs: Quit: Leaving.
22:25 (part) shofetim: "ERC Version 5.3 (IRC client for Emacs)"
22:41 (quit) jntrnr13: Quit: jntrnr13
22:50 (quit) GeneralMaximus: Ping timeout: 258 seconds
22:50 (join) GeneralMaximus
22:57 (join) jntrnr13
22:58 (join) mithos28
23:21 (join) pcavs
23:28 (join) superjudge
23:29 (quit) superjudge: Client Quit
23:29 (quit) jntrnr13: Quit: jntrnr13
23:32 eli: samth_away, stamourv: any chance that one of you is up?
23:42 (join) dmac
23:44 (join) jonrafkind
23:49 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/1MLAIA
23:49 RacketCommitBot: [racket/master] add menus and keybindings for jumping around to the errors in the defs - Robby Findler