01:50 (join) Bob__
01:50 Bob__: hey everyone
02:06 Lajla: Hey Bob__.
02:11 (quit) stamourv`: Ping timeout: 265 seconds
02:18 Bob__: I was wondering if one of you could help me?
02:19 Bob__: with a particular question
02:21 Lajla: Bob__, spit it.
02:21 Lajla: As you know, I am the expert on all things related to giving answers that seem plausibly correct to newbies but are gravely incorrect to the trained eye.
02:21 Bob__: so I need to write a function to insert the missing decades into a list of decades
02:22 Bob__: basically it consumes three things:
02:22 Bob__: start decade, end decade, list of decades
02:22 Lajla: Ahh
02:23 Bob__: basically, if there are any decades in the list of decades that are missing in the period set by start decade - end decade
02:23 Bob__: you add them in
02:23 Lajla: And these decades are just integers?
02:23 Bob__: yes
02:23 Bob__: the restriction is
02:23 Lajla: Seems easy enough
02:23 Lajla: Ahh
02:23 Lajla: restriction
02:23 Bob__: I can only use these for recursion: http://www.htdp.org/2003-09-26/Book/curriculum-Z-H-27.html#node_sec_21.2
02:23 Lajla: Ahhh
02:23 Bob__: so far, I've been able to create a predicate
02:23 Bob__: that will consume a decade and a list of decades
02:24 Bob__: and determine whether that decade is missing from the list
02:24 Bob__: but that's it
02:25 Lajla: Ehh, Bob__ one thing
02:25 Bob__: go on
02:25 Lajla: can you assume the original list is ordered?
02:25 Bob__: yes
02:25 Lajla: Like, from lower to upper decade?
02:25 Bob__: oh yeah
02:25 Bob__: forgot to mention
02:25 Bob__: that's very important
02:25 Lajla: Ahh, good.
02:25 Lajla: Yeah
02:25 Lajla: thought so
02:26 Bob__: oh yeah
02:27 Bob__: there was a hint given, which makes sense to me
02:27 Bob__: basically, it says to determine all the missing decades
02:27 Bob__: I assume that once you have the original list and the list of missing decades
02:27 Bob__: you can append them
02:27 Bob__: and then sort
02:28 Bob__: to get the completed list with all the decades correctly inserted
02:32 Lajla: Hmm
02:33 Lajla: that would be a very inefficent and borderline-imperative way to do it.
02:33 Lajla: I'm not quite sure of it either if I can only use those functions for recrusion.
02:33 Bob__: that's allright
02:43 (quit) jonrafkind: Ping timeout: 265 seconds
02:48 Bob__: wow holy crap
02:48 Bob__: I dunno what I just did
02:48 Bob__: but it worked
02:48 (join) hanDerPeder
03:06 (join) alexsuraci
03:37 (join) masm
03:43 (quit) Bob__: Ping timeout: 265 seconds
03:50 (join) lucian
04:04 (join) b4283
04:04 b4283: hi there
04:04 b4283: i have a question about the mzscheme shell
04:05 b4283: the arrow keys and libreadline keys doesn't work
04:06 b4283: there's just ^P ^[[A ^[[B, stuffs like this
04:18 lucian: b4283: try rlwrap
04:20 b4283: lucian: it worked, thanks so much
04:20 b4283: does this mean there's something wrong with my setup
04:20 lucian: b4283: you're welcome. it's a nice trick for broken consoles :)
04:20 lucian: b4283: no, there's something wrong with mzscheme. it doesn't use readline or an equivalent
04:21 b4283: i see
04:21 b4283: then it will be interesting to investigate this "rlwrap" tool
04:21 b4283: might gain some knowledge about linux system
04:21 b4283: :P
04:26 b4283: i see.. rlwrap simply faking the input for the user
04:51 (quit) offby1: Ping timeout: 245 seconds
05:02 (quit) Quetzalcoatl_: Ping timeout: 250 seconds
05:02 (join) Quetzalcoatl_
05:05 lucian: b4283: yeah, it's a readline buffer
05:16 (quit) jao: Ping timeout: 240 seconds
05:38 (quit) Quetzalcoatl_: Ping timeout: 240 seconds
05:38 (join) mceier
05:48 (join) Quetzalcoatl_
05:59 eli: b4283: http://docs.racket-lang.org/readline/
05:59 eli: lucian: And this part explains why it's not on by default: http://docs.racket-lang.org/readline/index.html#%28part._.License_.Issues%29
05:59 rudybot: http://tinyurl.com/2cbsat7
06:05 (quit) Quetzalcoatl_: Ping timeout: 245 seconds
06:11 (join) b-man_
06:27 (join) jao
06:28 lucian: eli: well, there are other implementations of readline
06:28 lucian: it's not even that complicated a library, it could be written in racket
06:30 eli: lucian: readline?  not complicated?  You're kidding, right?
06:30 lucian: eli: all of its features are complicated
06:30 eli: And re other implementations, the only one that I'm aware of is editline, which is incompatible with readline.
06:30 lucian: eli: but just a write buffer + history isn't all that hard
06:30 eli: Any kind of console interaction is insanely complicated, IME.
06:31 lucian: eli: depends how much you want to support
06:32 eli: I know of one project that was supposed to implmenet a readline thing in scheme, which didn't get anywhere.
06:32 lucian: eli: like http://newcenturycomputers.net/projects/readline.html
06:32 lucian: eli: and there's linenoise, which is rather small
06:32 lucian: not sure if it's still supported, though
06:33 (join) Quetzalcoatl_
06:33 eli: (That page says "... on Windows" where there isn't such a problem to begin with.)
06:34 lucian: eli: https://github.com/antirez/linenoise
06:35 lucian: eli: i might just port linenoise to scheme to learn it better :)
06:37 lucian: eli: it'd be much better if racket shipped with a minimal readline, that you can replace with gnu readline if you need its features
06:40 eli: lucian: That code looks extremely naive, and all hard-wired.
06:41 lucian: eli: it is, both
06:41 lucian: but it works quite well in most cases
06:41 lucian: so it's a significant improvement over nothing
06:41 b4283: so i had to recompile racket with readline
06:42 eli: lucian: Racket is big enough that you can't get by with just a few random terminals that you happen to have around,
06:42 eli: but it's certainly a small piece of code and converting it to racket can make it even smaller --
06:42 lucian: eli: yeah, that's what i meant
06:42 lucian: sure, it may not cover all cases
06:43 eli: if you do that then it would be nice to see if it's useful to everyone that needs it, and if so (given some time to use it), we can certainly switch.
06:43 eli: b4283: You don't need to recompile anything for readline to work.
06:44 b4283: ah
06:44 lucian: eli: yeah, sure. planet is great
06:44 eli: lucian: But doing this involves an intimidating maintenance -- the nightmarish kind where things fail for other people with no clue why or how.
06:44 b4283: i found the docs, thanks
06:44 lucian: eli: i'll have a look at porting it when i have time
06:45 lucian: eli: well, it can have a whitelist of platform&terminal combinations where it obviously works
06:45 eli: BTW, we also use readline's completion, and it also has paren matching etc.
06:45 lucian: eli: yeah, i know
06:45 lucian: readline's way better
06:45 lucian: but not even having history is silly
06:45 eli: ?
06:45 eli: It certainly does...
07:02 (join) neilv
07:27 (join) martinhex
07:44 (quit) Quetzalcoatl_: Ping timeout: 276 seconds
07:48 (join) Quetzalcoatl_
07:49 (quit) mceier: Ping timeout: 255 seconds
07:51 (join) mceier
07:55 (join) dzhus
08:06 (quit) Quetzalcoatl_: Ping timeout: 264 seconds
08:12 (join) Quetzalcoatl_
08:15 (quit) neilv: Quit: Leaving
08:17 (quit) masm: Quit: Leaving.
08:19 (join) MayDaniel
08:29 (quit) MayDaniel: 
08:52 (quit) martinhex: Remote host closed the connection
08:54 (join) kevin01123
09:01 (quit) b-man_: Read error: Connection reset by peer
09:01 (part) kevin01123: "ERC Version 5.3 (IRC client for Emacs)"
09:21 (quit) alexsuraci: Read error: Connection reset by peer
09:23 (quit) hanDerPeder: Quit: hanDerPeder
09:27 (join) alexsuraci
09:28 (join) hanDerPeder
09:39 (join) kevin01123
09:40 kevin01123: If I wanted to compile racket version 5.0.2 from the git repo, would I just clone it and the issue a "git checkout v5.0.2" and compile it?
09:42 (join) b-man_
10:07 (quit) kevin01123: Remote host closed the connection
10:21 (quit) rudybot: *.net *.split
10:21 (quit) Lajla: *.net *.split
10:21 (quit) Guest1821: *.net *.split
10:23 (quit) Ogion: Ping timeout: 276 seconds
10:24 (join) MayDaniel
10:24 (join) rudybot
10:24 (join) Lajla
10:24 (join) Guest1821
10:24 (join) Ogion
10:41 (join) rmitt__
11:05 (quit) alexsuraci: Quit: Lost terminal
11:15 (quit) rmitt__: Ping timeout: 265 seconds
11:32 (quit) b4283: Ping timeout: 272 seconds
11:40 (join) anRch
11:55 (join) offby1
11:56 (quit) jao: Ping timeout: 240 seconds
11:56 (join) samth_
11:57 (quit) samth_: Remote host closed the connection
11:58 (quit) b-man_: Remote host closed the connection
11:58 (join) alexsuraci
11:59 lucian: i can't find any tty libs for racket
12:00 (join) stamourv
12:01 offby1: it is often the case that one cannot find any X libs for Racket, for various values of X.
12:02 offby1: by "tty libs" to you mean "libraries that manipulate the cursor position, background color, etc of old-fashioned terminal emulators like 'xterm'"?
12:03 lucian: offby1: something like termios.h i guess
12:03 offby1: remember that Racket has a decent C FFI, so if there's a _c_ library that does what you want, then you'll be able to use it (in theory anyway)
12:03 lucian: offby1: yeah, that's what I'll probably have to do
12:03 lucian: i was hoping for something nicer
12:04 offby1: welcome to the world of minority languages :-|
12:04 lucian: offby1: it sucks, man :)
12:04 lucian is primarily a python guy
12:05 offby1: ooh
12:05 offby1: no batteries included here :)
12:05 lucian: offby1: so the racket homepage lies? :P
12:06 offby1: probably not
12:06 offby1: what are you looking at?
12:08 samth: offby1, he is refererring to this "Racket includes both batteries ..."  on the home page
12:08 lucian: offby1: "Draw more pictures or build a web server from scratch. Racket includes both batteries and a programming environment, so get started!"
12:08 samth: lucian, Racket certainly comes with batteries, in the form of extensive libraries
12:08 (quit) MayDaniel: 
12:09 offby1: lucian: ah, well, compared to other Scheme implementations, racket has lots of nice libraries.
12:09 samth: it doesn't happen to have one to bind to termio
12:09 offby1: But compared to, say, Python ... :-|
12:09 offby1: or C# or Java ...
12:09 offby1: lucian: so this is your chance to contribute!
12:10 lucian: yeah, i realise racket is one of the richer schemes
12:10 lucian: offby1: my goal was implementing something like linenoise https://github.com/antirez/linenoise/
12:11 lucian: i was hoping there's a termios or similar binding bundled with racket, so that my racknoise would work out of the box
12:11 offby1: ha, antirez!
12:11 offby1 big redis fan
12:11 lucian: but if i have to bind to C, i might as well bind to linenoise directly
12:11 samth: then you'd have to ship linenoise
12:12 samth: also, there are some benefits from readline
12:12 samth: for example, it gives you tab-completion based on bound identifiers in Racket
12:14 samth: also, isn't there a bsd-licensed reimplementation of readline?
12:14 (quit) alexsuraci: Read error: Connection reset by peer
12:14 lucian: samth: i haven't found one that was maintained
12:14 lucian: linenoise is tiny by comparison, too
12:15 samth: ah, it's called editline
12:15 samth: Racket includes both batteries
12:15 samth: http://www.thrysoee.dk/editline/
12:16 lucian: samth: libedit isn't compatible with readline afaik
12:19 samth: in what sense?
12:20 lucian: samth: different API, so it needs a binding
12:20 samth: ah
12:23 lucian: samth: so I was thinking porting linenoise to scheme would be a good exercise, since i'm new to scheme
12:24 samth: i agree
12:24 samth: it should be easy to write a binding to the necessary parts of termio
12:25 lucian: samth: i guess. i was hoping i wouldn't have to touch the ffi yet :)
12:25 samth: the ffi is really easy
12:26 (quit) Ogion: Ping timeout: 240 seconds
12:27 (join) Ogion
12:35 (quit) anRch: Quit: anRch
12:44 (join) asumu
12:45 (join) MayDaniel
12:46 (quit) tv|z: Ping timeout: 245 seconds
12:48 (join) tv|z
12:48 (join) jao
12:57 (join) alexsuraci
13:11 (join) jonrafkind
13:22 lucian: samth: apparently termios is always compiled statically, so the ffi can't use it unless i make a .so/.dylib
13:34 (join) rmitt__
13:35 (quit) rmitt__: Client Quit
13:41 (quit) MayDaniel: 
14:08 samth: lucian, that's strange
14:09 lucian: samth: there's no .so for termios, not so strange i guess
14:10 samth: i have /usr/lib/python2.6/lib-dynload/termios.so on my system, so some people do it
14:15 lucian: samth: that's python's binding
14:15 samth: ah, ok
14:17 lucian: samth: bah, terminals suck all around apparently
14:18 lucian: samth: so I can either make a C module for racket or a dummy .so for ffi to load
14:19 lucian: damn, this is harder than I'd have liked
14:19 lucian: i was hoping to learn some scheme, instead i'm fighting C ...
14:20 (join) _mo_
14:22 (join) b-man_
14:28 samth: lucian, it seems that languages just typically ship a termios.so with their bindings
14:43 (join) masm
14:54 (join) anRch
14:54 (quit) dzhus: Ping timeout: 255 seconds
14:55 (quit) masm: Quit: Leaving.
14:56 (join) masm
15:00 (join) MayDaniel
15:07 eli: termios.h looks like a kernel header, so it's not something to statically link, there's just not code to link to.
15:08 (quit) MayDaniel: Read error: Connection reset by peer
15:17 lucian: eli: depends on platform, it seems
15:17 eli: Even more fun then.
15:18 eli: (I get the feeling that this will not help you in having fun hacking...)
15:18 lucian: eli: it won't, indeed
15:19 lucian: eli: although it seems it is a kernel API on interesting platforms
15:21 lucian: eli: so how do I do system calls with the ffi?
15:23 eli: lucian: probably like other functions, but seriously -- if you're relatively new, it's not a good direction.
15:23 eli: (As much as I'd like to have an in-racket readline thin...)
15:40 clklein: Is disabling the stdout diff the only effect of the drdr:random prop?
15:42 (quit) anRch: Quit: anRch
15:42 samth: clklein, it also doesn't tell the committer when they "break" it
15:43 clklein: samth: Oh, we should decouple those.
15:44 samth: i think random failure and random output probably go together
15:44 clklein: Here's my situation.
15:44 clklein: I'm using random testing to compare the delimited continuations model to Racket.
15:44 (quit) asumu: Read error: Operation timed out
15:44 clklein: Sometimes a random program will segfault Racket.
15:45 lucian: eli: meh, I'd like to see how io works in a lisp
15:45 lucian: eli: i mean play with it
15:45 clklein: I could eval it in a separate Racket process and then notice the non-zero exit status, but it's easier to jsut print the program then run it in-process
15:46 clklein: (also much faster than starting new VMs)
15:46 eli: lucian: You're getting into an a awfully heavy case of "IO"...
15:46 lucian: eli: i can go back to simpler things if I can't dig myself out of it
15:46 lucian: eli: now I'm trying to figure out how to make system calls
15:46 lucian: (termios.h is kinda horrible)
15:46 clklein: samth: Nevermind, I get it teh connection now.
15:48 samth: clklein, you should print the random seed and make it take the seed as a command line argument
15:49 clklein: How is that better than printing the program?
15:49 samth: that's good to do too, but it's really easy to rerun if you provide a command line that always behaves that way
15:52 lucian: eli: well, I know how to do it in Python, easily
15:53 lucian: eli: and i can write scheme code decently. so i might as well try this
15:53 lucian: eli: but it seems system calls are beyond the ffi's capabilities
15:53 eli: lucian: Something like this: (get-ffi-obj 'puts #f (_fun _string -> _int))
15:53 eli: ?
15:53 lucian: eli: hmm, i'll try
15:54 lucian: eli: it's not documented, though
15:54 eli: lucian: You'll need to (require ffi/unsafe)
15:54 lucian: eli: that's there
15:55 eli: And if you do that and then (help get-ffi-obj) on the repl you should see the documentation.
15:59 lucian: eli: heh, seems to work. I just called setuid
15:59 eli: ThereYoGo.
16:00 eli: If you have any experience with ctypes in python, the racket ffi stuff is similar, except that it's more advanced in how it deals with callbacks.
16:01 eli: And except that it's less advanced in the libffi stuff -- there's no array types and no unions.
16:01 eli: But samth is working on adding those.
16:01 samth: eli, !
16:01 eli: He's also from canada.
16:01 samth: eli, the canadians are protesting
16:01 eli: :)
16:03 lucian: eli: and not having termios.h will be kinda horrible
16:03 lucian: bah
16:03 samth: lucian, check out http://planet.racket-lang.org/display.ss?package=c.plt&owner=dherman
16:04 eli: I think that Jay's thing was similar but better, no?
16:04 samth: eli, i think it's rather different
16:05 samth: jay'
16:05 samth: jay's package just runs a c program
16:05 samth: dave's is for ffi generation
16:06 lucian: samth: ah, looks nice. thanks
16:06 eli: Well, IIRC, Dave's code creates a C stub file to runs once and get the information,
16:06 eli: wheras Jay's just allows any random C code,
16:07 samth: yes
16:07 eli: So there's not that much difference -- and if you're getting down to C you can just as well write some simple code that is easier on the glue side.
16:08 lucian: eli: yeah, if I write any C, i'll write a wrapper for termios.h
16:09 eli: lucian: Yeah, it (the other package I'm talking about) *might* be a saner approach for you if you do such a low-level library.
16:09 (quit) askhader: Quit: leaving
16:09 lucian: eli: do you have a name/link?
16:09 (join) askhader
16:09 eli: For example, it sounded like some "functions" are actually going to be macros in some platforms, and some obscure __functions on others.
16:10 samth: i think they always have to be function prototypes, or at least that's what the man page says
16:10 lucian: eli: oh yeah, that's such a horrible feature of C
16:10 eli: lucian: So if samth is right and there's always function prototypes, and they're relatively easy to interface, then you can go with that.
16:11 lucian: eli: yeah, but I'm not sure that's true. i guess I can test it after
16:12 eli: Even if the interface is complicated -- for example, if it makes you run down obscure chains of pointers or whatever horrible things you can do in C -- in this case the C-wrapper approach can be overall easier.
16:12 eli: For example, you write some C code that deals with such a mess then exposes some simple int->int interface, and that's extremely simple on the racket side.
16:12 eli: In any case, that oter package is: http://planet.racket-lang.org/display.ss?package=superc.plt&owner=jaymccarthy
16:12 rudybot: http://tinyurl.com/2g4hoy3
16:15 lucian: eli: yeah, but ideally i'd not have any C code
16:15 lucian: i'll see
16:15 lucian: eli: oh, superC only works on osx
16:16 eli: It is?  That's a surprise.
16:17 lucian: eli: so I guess what i'd like is something that can parse termios.h and spit out some gen-ffi-obj
16:18 eli: lucian: That passage on the documentation is probably an attempt to say that on OSX it requires the developer tools.
16:18 eli: I'd be very surprised if it doesn't work on linux or any other system with a well behaved C compiler.
16:18 eli: (Even windows should be mostly fine.)
16:19 lucian: eli: hmm, now that I've re-read it I agree
16:19 eli: You should bug him (Jay) about that confusing bit.
16:20 eli disappears
16:21 lucian: eli: samth: this looks like what I want http://planet.racket-lang.org/package-source/dherman/c.plt/3/2/planet-docs/c/header.html
16:21 rudybot: http://tinyurl.com/33lm6cw
16:21 samth: lucian, yes, that's exactly what i was thinking of
16:26 (quit) lucian: Remote host closed the connection
16:27 (join) lucian
16:35 (join) timmcd
16:35 timmcd: hello
16:35 timmcd: I am using some of the OO/class features, and I was wondering if there were any sort of shortcuts for field and method access on objects?
16:36 timmcd: Ie, (my-object field-or-method-to-access), vs. (get-field field my-object)
16:39 (part) timmcd
16:40 (nick) samth -> samth_tofurkey
16:41 (quit) samth_tofurkey: Quit: Ex-Chat
16:44 (quit) lisppaste: Ping timeout: 240 seconds
16:44 (join) stevwonder
16:46 stevwonder: hey, can anyone help me with racket on linux?
16:47 (quit) stevwonder: Client Quit
16:58 (quit) lucian: Remote host closed the connection
17:00 (quit) b-man_: Ping timeout: 252 seconds
17:03 (quit) _mo_: Quit: Leaving
17:30 bremner: stevwonder: you'll have better luck asking a specific question. It sometimes takes a while for answers in this channel.
17:34 askhader: Did they d/c?
17:34 (quit) hanDerPeder: Quit: hanDerPeder
17:35 bremner: dunno, they tab completed.
17:35 bremner: hmm. irssi completion seems not so good at noticing people disconnect.
17:36 bremner: blame the tools, that's my motto.
17:40 askhader: Well i've rigged my irc client to filter out parts and joins for this particular channel, so tab completion is all I have to determine if someone is around short of actually doing a /name
17:54 (quit) mceier: Quit: Lost terminal
17:56 (join) mceier
18:10 (quit) mceier: Quit: leaving
19:00 (quit) Quetzalcoatl_: Ping timeout: 245 seconds
19:28 (join) b-man_
19:39 (quit) jonrafkind: Ping timeout: 255 seconds
19:55 (join) lisppaste
20:22 (quit) b-man_: Ping timeout: 260 seconds
20:40 (join) Quetzalcoatl_
21:06 (quit) Quetzalcoatl_: Ping timeout: 252 seconds
21:25 (part) masm
21:33 (quit) mapcadr_: Ping timeout: 272 seconds
21:38 (join) mapcadr
21:55 (join) Quetzalcoatl_