2015-02-04T00:01:20Z profess joined #lisp
2015-02-04T00:02:03Z mishoo quit (Remote host closed the connection)
2015-02-04T00:02:25Z mishoo joined #lisp
2015-02-04T00:02:33Z eudoxia quit (Quit: Leaving)
2015-02-04T00:03:22Z a201502012 quit (Quit: Page closed)
2015-02-04T00:04:01Z t4nk072 quit (Ping timeout: 246 seconds)
2015-02-04T00:12:06Z bgs100 joined #lisp
2015-02-04T00:12:58Z zbrown_ is now known as zbrown
2015-02-04T00:13:03Z zbrown quit (Changing host)
2015-02-04T00:13:03Z zbrown joined #lisp
2015-02-04T00:14:26Z dagnachew joined #lisp
2015-02-04T00:16:50Z cadadar quit (Quit: Leaving.)
2015-02-04T00:18:23Z cdtaylor quit (Remote host closed the connection)
2015-02-04T00:21:20Z mishoo quit (Ping timeout: 252 seconds)
2015-02-04T00:22:50Z Bicyclidine quit (Quit: leaving)
2015-02-04T00:22:53Z emaczen quit (Ping timeout: 245 seconds)
2015-02-04T00:24:27Z emaczen joined #lisp
2015-02-04T00:25:25Z robot-beethoven joined #lisp
2015-02-04T00:25:37Z munksgaard quit (Ping timeout: 264 seconds)
2015-02-04T00:30:33Z MoALTz_ joined #lisp
2015-02-04T00:33:01Z fridim_ quit (Ping timeout: 256 seconds)
2015-02-04T00:33:34Z MoALTz quit (Ping timeout: 245 seconds)
2015-02-04T00:35:00Z dagnachew quit (Quit: WeeChat 1.1.1)
2015-02-04T00:35:51Z kcj joined #lisp
2015-02-04T00:37:38Z a201502012 joined #lisp
2015-02-04T00:39:24Z JokesOnYou77 joined #lisp
2015-02-04T00:39:42Z JokesOnYou77: Hello everyone
2015-02-04T00:40:01Z meiji11 joined #lisp
2015-02-04T00:41:06Z nyef: Hello JokesOnYou77.
2015-02-04T00:42:48Z dpb9cpu joined #lisp
2015-02-04T00:42:52Z taspat``` joined #lisp
2015-02-04T00:44:16Z taspat`` quit (Ping timeout: 245 seconds)
2015-02-04T00:45:48Z harish quit (Ping timeout: 245 seconds)
2015-02-04T00:49:56Z wheelsucker quit (Quit: Client Quit)
2015-02-04T00:50:11Z linux_dream quit (Quit: Leaving)
2015-02-04T00:51:50Z ruste_ quit (Read error: Connection reset by peer)
2015-02-04T00:54:05Z EvW joined #lisp
2015-02-04T00:55:43Z dkcl joined #lisp
2015-02-04T00:57:33Z manuel__ joined #lisp
2015-02-04T00:59:34Z frkout quit (Remote host closed the connection)
2015-02-04T01:00:01Z frkout joined #lisp
2015-02-04T01:00:01Z dkcl quit (Remote host closed the connection)
2015-02-04T01:00:22Z psy_ quit (Read error: Connection reset by peer)
2015-02-04T01:01:03Z psy_ joined #lisp
2015-02-04T01:04:23Z frkout quit (Ping timeout: 250 seconds)
2015-02-04T01:07:50Z ruste_ joined #lisp
2015-02-04T01:10:27Z nowhereman quit (Read error: Connection reset by peer)
2015-02-04T01:14:36Z TDog joined #lisp
2015-02-04T01:15:40Z cocosp quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
2015-02-04T01:19:25Z hiroakip quit (Ping timeout: 250 seconds)
2015-02-04T01:22:08Z urandom__ quit (Quit: Konversation terminated!)
2015-02-04T01:23:18Z k-dawg quit (Quit: This computer has gone to sleep)
2015-02-04T01:29:41Z antonv quit (Ping timeout: 245 seconds)
2015-02-04T01:31:41Z hiroakip joined #lisp
2015-02-04T01:32:39Z nyef: froggey: Have you given much though to how to handle internals documentation for Mezzano?
2015-02-04T01:34:02Z hellofunk quit (Ping timeout: 245 seconds)
2015-02-04T01:34:40Z oleo is now known as Guest10153
2015-02-04T01:34:43Z bugrum quit (Ping timeout: 250 seconds)
2015-02-04T01:36:24Z oleo__ joined #lisp
2015-02-04T01:38:13Z Guest10153 quit (Ping timeout: 264 seconds)
2015-02-04T01:40:32Z Posterdati quit (Ping timeout: 252 seconds)
2015-02-04T01:43:31Z ft quit (Ping timeout: 272 seconds)
2015-02-04T01:43:41Z EvW quit (Ping timeout: 250 seconds)
2015-02-04T01:45:06Z ft joined #lisp
2015-02-04T01:49:04Z nyef: ... Wha? If there wasn't a CONS area the GC wouldn't be able to tell the difference between a CONS and an object header? Really?!?
2015-02-04T01:50:18Z nyef: Which leads to the question, was this the point of the CONS area in the first place, or was the existence of a CONS area something that the type system took advantage of?
2015-02-04T01:50:28Z MrWoohoo quit (Quit: ["Textual IRC Client: www.textualapp.com"])
2015-02-04T01:50:29Z vanila: interesting :)
2015-02-04T01:51:36Z nikki93 joined #lisp
2015-02-04T01:51:50Z nyef: vanila: I have enough of a handle on the system at this point to be able to start in on two chapters of an internals manual, even though I've never built it past the point where it fails to find a running file-server.
2015-02-04T01:52:18Z BitPuffin joined #lisp
2015-02-04T01:52:31Z vanila: I don't know how you do it!
2015-02-04T01:52:32Z nikki93 quit (Remote host closed the connection)
2015-02-04T01:52:47Z emaczen: If I have an :initarg and an initialize-instance method how can I specify which one to use when creating an object via 'make-instance?
2015-02-04T01:53:04Z vanila: I  haven't starte reading the code seriously yet, for me building is a prereq (and im so happy that it worked)
2015-02-04T01:53:29Z enitiz joined #lisp
2015-02-04T01:53:51Z hiroakip quit (Ping timeout: 245 seconds)
2015-02-04T01:54:09Z Bike: emaczen: initialize-instance receives the initargs, so it can choose what to do with them. or did you mean something else.
2015-02-04T01:54:10Z nyef: Oh, it builds to the point that the initial REPL comes up, and that's enough for me to be able to test making changes to supervisor/ things.
2015-02-04T01:55:21Z nikki93 joined #lisp
2015-02-04T01:55:38Z emaczen: Bike: Sometimes I only want the :initarg parameter in the defclass form and other times I just want initialize-instance to be called.
2015-02-04T01:55:59Z Bike: the initialize-instance is always called
2015-02-04T01:56:10Z Bike: if you want it to use initargs only sometimes, write the method to do so
2015-02-04T01:56:37Z emaczen: Does the defclass form get evaluated before intialize-instance?
2015-02-04T01:57:26Z Bike: the defaulted initargs do, yeah
2015-02-04T01:57:37Z emaczen: If so, I could make the initform be nil and then check for nillity
2015-02-04T01:59:02Z vanila: nyef, is there a problem with building or do you just not want to wait so long?
2015-02-04T01:59:24Z nyef: Couldn't be bothered getting the file-server bit sorted out.
2015-02-04T01:59:34Z emaczen: Bike: I essentially want a copy constuctor.
2015-02-04T02:00:04Z emaczen: Bike: actually that isn't right.
2015-02-04T02:00:24Z wheelsucker joined #lisp
2015-02-04T02:00:25Z enitiz quit (Ping timeout: 264 seconds)
2015-02-04T02:00:32Z emaczen: Bike: I just have different arguments to my constructor sometimes
2015-02-04T02:02:55Z nikki93 quit (Remote host closed the connection)
2015-02-04T02:03:22Z nyef: vanila: Really, there's enough for me to do with the system even without getting the file-server going.
2015-02-04T02:03:42Z vanila: alright :)
2015-02-04T02:04:10Z vanila: the file server isn't too difficult, I managed to get it working so if you ever want to try i can give details
2015-02-04T02:05:01Z vanila: im curious how much everyhing will speed upwith a bit of register allocation in the compiler - I doubt it's a bottleneck though
2015-02-04T02:05:36Z nyef: Oh, I'm not expecting it to be DIFFICULT... But at the same time I looked at it and said "it looks like it'll be simple enough to add support for a real filesystem..."
2015-02-04T02:06:15Z nyef: Oh, good grief.
2015-02-04T02:06:31Z nyef: Hey, is there an atomic compare-and-swap operation in this thing?
2015-02-04T02:07:48Z enitiz joined #lisp
2015-02-04T02:08:49Z nyef: Yeah, I totally have a lot that I can do just within the scope of supervisor/, runtime/, and system/.
2015-02-04T02:16:18Z nell quit (Quit: WeeChat 1.2-dev)
2015-02-04T02:18:39Z taspat``` quit (Ping timeout: 264 seconds)
2015-02-04T02:24:04Z mogglebang joined #lisp
2015-02-04T02:27:01Z PinealGlandOptic quit (Ping timeout: 250 seconds)
2015-02-04T02:27:28Z nyef: Okay, there IS some sort of CAS operation, but the implementation leads me in yet another direction.
2015-02-04T02:29:06Z PaulCapestany quit
2015-02-04T02:31:07Z pillton quit (Ping timeout: 272 seconds)
2015-02-04T02:31:19Z emaczen: Bike: Thanks -- I figured it out
2015-02-04T02:31:21Z stepnem quit (Ping timeout: 250 seconds)
2015-02-04T02:32:41Z badkins quit
2015-02-04T02:34:41Z mogglebang quit (Ping timeout: 246 seconds)
2015-02-04T02:36:26Z kapil__ joined #lisp
2015-02-04T02:37:53Z nell joined #lisp
2015-02-04T02:38:28Z mogglebang joined #lisp
2015-02-04T02:38:47Z hiyosi quit (Quit: My Mac has gone to sleep. ZZZzzz…)
2015-02-04T02:42:16Z pjb` joined #lisp
2015-02-04T02:42:17Z xyh: how do I define a new host? to use it in logical pathname namestring as "host:asd;dsa.org" ?
2015-02-04T02:43:58Z PinealGlandOptic joined #lisp
2015-02-04T02:44:19Z pjb quit (Ping timeout: 272 seconds)
2015-02-04T02:46:53Z harish joined #lisp
2015-02-04T02:47:55Z cluck` joined #lisp
2015-02-04T02:48:23Z cluck quit (Ping timeout: 250 seconds)
2015-02-04T02:48:59Z Mon_Ouie quit (Quit: WeeChat 1.1.1)
2015-02-04T02:50:21Z aap quit (Read error: Connection reset by peer)
2015-02-04T02:50:28Z aap joined #lisp
2015-02-04T02:56:10Z frkout joined #lisp
2015-02-04T02:57:41Z nyef: xyh: In Mezzano? I'd take a look at mezzano.file-system.local:add-local-file-host and mezzano.file-system.remote:add-simple-file-host and their uses.
2015-02-04T02:58:12Z Takumo quit (Ping timeout: 244 seconds)
2015-02-04T03:00:53Z Takumo joined #lisp
2015-02-04T03:04:05Z frkout quit (Remote host closed the connection)
2015-02-04T03:04:31Z frkout joined #lisp
2015-02-04T03:05:56Z krfantasy joined #lisp
2015-02-04T03:06:49Z nikki93 joined #lisp
2015-02-04T03:07:16Z nikki93 quit (Remote host closed the connection)
2015-02-04T03:07:30Z vdamewood quit (Quit: ["Textual IRC Client: www.textualapp.com"])
2015-02-04T03:07:46Z nikki93 joined #lisp
2015-02-04T03:08:31Z xyh: nyef: in ccl.  and maybe I should not define new host.
2015-02-04T03:09:07Z frkout quit (Ping timeout: 252 seconds)
2015-02-04T03:09:44Z nyef: Ah.
2015-02-04T03:09:51Z nyef: If you just want the pathname translations...
2015-02-04T03:09:57Z ynniv joined #lisp
2015-02-04T03:09:58Z nyef: clhs logical-pathname-translation
2015-02-04T03:09:58Z specbot: Couldn't find anything for logical-pathname-translation.
2015-02-04T03:10:00Z nyef: clhs logical-pathname-translations
2015-02-04T03:10:00Z specbot: http://www.lispworks.com/reference/HyperSpec/Body/f_logica.htm
2015-02-04T03:10:23Z nyef: I don't know that that'll do what you want, though.
2015-02-04T03:11:13Z fragamus joined #lisp
2015-02-04T03:11:34Z ynniv: It looks like ftype doesn’t support typed lists, yes? If I want a typed collection I need to make a class or a homogenous vector?
2015-02-04T03:12:19Z nyef: Unless it's a list of known-at-compile-time length, yes.
2015-02-04T03:12:46Z nyef: If you know the length at compile-time, you can produce a suitably-nested CONS type.
2015-02-04T03:13:12Z mogglebang quit (Quit: mogglebang)
2015-02-04T03:13:35Z ynniv: Right now I’m making dumb mistakes declaring test data, so yes my lists are being declared at compile time.
2015-02-04T03:14:14Z cluck` is now known as cluck
2015-02-04T03:14:15Z cmack quit (Ping timeout: 250 seconds)
2015-02-04T03:14:56Z nyef: Well, I'll point out that (cons fixnum (cons fixnum null)) is a type specifier for a two-element list of FIXNUMs.
2015-02-04T03:15:05Z nikki93 quit (Remote host closed the connection)
2015-02-04T03:15:29Z burtons joined #lisp
2015-02-04T03:15:41Z ynniv: Ah, the cons type accepts parameters. List does not at all?
2015-02-04T03:15:54Z nyef: LIST, IIRC, is defined as (OR CONS NULL).
2015-02-04T03:16:02Z nyef: clhs list
2015-02-04T03:16:02Z specbot: http://www.lispworks.com/reference/HyperSpec/Body/a_list.htm
2015-02-04T03:16:17Z ebrasca quit (Quit: ebrasca)
2015-02-04T03:16:23Z BitPuffin quit (Ping timeout: 240 seconds)
2015-02-04T03:16:32Z ynniv: That makes sense. I think this solves my problem, thanks.
2015-02-04T03:16:45Z nyef: Compare that to the (linked) page for System Class CONS.
2015-02-04T03:17:28Z ASau quit (Read error: No route to host)
2015-02-04T03:19:58Z Qudit314159 joined #lisp
2015-02-04T03:20:27Z mogglebang joined #lisp
2015-02-04T03:20:33Z JokesOnYou77: Knowing the type means the cimpiler knows how much RAM to allocate for the stored object, right?  Does it give other advantages as well?
2015-02-04T03:21:31Z ynniv: The compiler can throw an error if you attempt to call a function with the wrong argument types.
2015-02-04T03:21:53Z Quadrescence joined #lisp
2015-02-04T03:22:31Z JokesOnYou77: So it gets me built-in type error checking as well.  But I was wondering about what makes lisp with type declarations faster.
2015-02-04T03:24:34Z mogglebang quit (Client Quit)
2015-02-04T03:24:45Z ASau joined #lisp
2015-02-04T03:26:05Z ynniv: It can avoid checking types at runtime.
2015-02-04T03:26:09Z Qudit314159 quit (Read error: Connection reset by peer)
2015-02-04T03:26:30Z Qudit314159 joined #lisp
2015-02-04T03:26:32Z nyef: Right. The compiler doesn't need to emit type-checks, which can also avoid massive runtime dispatch.
2015-02-04T03:27:36Z nyef: I had a project where the runtime was sped up from something like ten hours to less than a minute by suitable type definitions, inlining, and changing storage representation to allow storing unboxed word-wide data.
2015-02-04T03:28:17Z JokesOnYou77: Ok, to when I don't declare types, the compiler is calling CHECK0TYPE, or some equivalent, all over the place and that's a significant burden in terms of funcalls.    That makes complete sense.
2015-02-04T03:28:26Z JokesOnYou77: Wow, that's awesome.
2015-02-04T03:28:40Z JokesOnYou77: And thank you for the explanation.
2015-02-04T03:30:01Z nyef: Now, that wasn't because of the cost of the allocation or GC, that was the compiler knowing that it could use an ADD instruction instead of a full call to GENERIC-+. Things like that.
2015-02-04T03:30:51Z fragamus quit (Quit: Computer has gone to sleep.)
2015-02-04T03:31:08Z vanila quit (Quit: Leaving)
2015-02-04T03:31:13Z nyef: At the same time, any consing that my program did meant that it was doing BIGNUM allocation, which meant that it was doing something that would require a type-check down the line, which meant it was going to be slow.
2015-02-04T03:32:05Z JokesOnYou77: Bignum allocation definitely sounds like a case for type declarations.
2015-02-04T03:32:29Z nyef: Not just for type declarations.
2015-02-04T03:32:42Z nyef: I was doing something with 48-bit integers on a 32-bit machine.
2015-02-04T03:32:48Z JokesOnYou77: And it also sounds like the compiler can choose lower level functions (ADD) when it knows the type which makes sense
2015-02-04T03:33:00Z JokesOnYou77: ahhh, that will do the trick lol.
2015-02-04T03:33:11Z mogglebang joined #lisp
2015-02-04T03:33:12Z nyef: And there were places where I had 33-bit values as well.
2015-02-04T03:33:29Z JokesOnYou77: Is 48-bit int still a BIGNUM on 64-bit machine?  I guess it doesn't change, right?
2015-02-04T03:34:21Z Bike: it oughtabe a fixnum
2015-02-04T03:34:23Z nyef: SBCL on 64-bit machines is currently using 63-bit fixnums, but should be able to support anything from 61-bit to 63-bit. Anything larger becomes a BIGNUM.
2015-02-04T03:35:05Z JokesOnYou77: Oh, so it does make a difference.
2015-02-04T03:35:12Z JokesOnYou77: Bike, why fixnum?
2015-02-04T03:35:15Z nyef: Huge difference at times.
2015-02-04T03:35:26Z Bike: JokesOnYou77: why not?
2015-02-04T03:36:54Z JokesOnYou77: Bike, good question lol.  I had to go check the hyperspec :P
2015-02-04T03:37:08Z nikki93 joined #lisp
2015-02-04T03:37:53Z Bike: there's no major semantic differences, might as well have fixnums as large as you can
2015-02-04T03:38:55Z a201502012 quit (Ping timeout: 246 seconds)
2015-02-04T03:39:20Z JokesOnYou77: Bike, On another matter, beach's test wouldn't run on LW.  I left him a memo but I wasn't sure if I should wait for him or ask you?
2015-02-04T03:39:56Z ynniv: sbcl from homebrew on macosx: (typep (expt 2 61) 'fixnum) —> T
2015-02-04T03:40:03Z Bike: well, i didn't write it, but i can try to help, sure
2015-02-04T03:43:01Z JokesOnYou77: Bike, It segfaults :(  http://paste.lisp.org/+34CJ
2015-02-04T03:43:27Z Bike: ...I have no help for that, wow.
2015-02-04T03:43:35Z Bike: Before it even outputs anything?
2015-02-04T03:44:04Z JokesOnYou77: Bike, just to eliminate the improbable (as I can attribute one of the worst busgs I've ever had to this), I ran a memtest and it's not bad ram
2015-02-04T03:44:15Z Bike: heh, no, that's doubtful
2015-02-04T03:44:21Z JokesOnYou77: It creates the output file but it's blank.
2015-02-04T03:44:30Z nyef: minion: advice 12000?
2015-02-04T03:44:30Z minion: Looking for a compiler bug is the second-to-last resort. The last resort is blaming bad RAM. It's never the correct hypothesis.
2015-02-04T03:44:49Z Bike: yeah, i have no idea what could be causing that, sorry. you might want to ask lw support, that code should be reasonably innocuous.
2015-02-04T03:44:53Z JokesOnYou77: I left it run over night last night, no change
2015-02-04T03:45:11Z bgs100 quit (Quit: bgs100)
2015-02-04T03:45:42Z zacharias_ joined #lisp
2015-02-04T03:46:04Z Bike: What, it's not immediate?
2015-02-04T03:46:08Z krfantasy quit (Quit: krfantasy)
2015-02-04T03:46:15Z nikki93 quit (Remote host closed the connection)
2015-02-04T03:46:20Z JokesOnYou77: nyef, LMAO!!! The worst bug I've ever had in my life!  After scratching my head for two days I noticed a pattern in the addresses of the errors and ran a memtest.  Corsair sent me the refund just last week :D
2015-02-04T03:47:34Z |3b|: ynniv: sbcl can be configured so that isn't true, but it is by default
2015-02-04T03:48:04Z JokesOnYou77: Bike, frustratingly, it's slightly more complicated than that.  When I run it from the gui LW ide it just dies and prints that message to the command line where I launched it from.  When I run it from then command line or from emacs (using a LW image as a back end) it just hangs with no message.
2015-02-04T03:48:16Z nikki93 joined #lisp
2015-02-04T03:48:33Z zacharias quit (Ping timeout: 246 seconds)
2015-02-04T03:49:06Z Bike: can you configure lw to start with a larger heap?
2015-02-04T03:49:47Z nyef: JokesOnYou77: Sure. And as a compiler hacker, I can assure you that quite often I AM looking for a compiler bug.
2015-02-04T03:50:28Z theos quit (Disconnected by services)
2015-02-04T03:50:34Z JokesOnYou77: nyef, firend of a friend found a bug in ruby and they sent him a little ruby.  If I traced a bug to the compiler I don't know if I'd believe myself.
2015-02-04T03:50:46Z JokesOnYou77: Bike, I'm not sure, I'm looking at docs now
2015-02-04T03:50:58Z pillton joined #lisp
2015-02-04T03:50:59Z theos joined #lisp
2015-02-04T03:51:05Z Bike: well, you know nyef, surely that should remind you that compilers are made by humans too :p
2015-02-04T03:51:22Z nyef: Bike: As one of said humans, I find it hard to forget. d-:
2015-02-04T03:51:54Z nikki93 quit (Read error: Connection reset by peer)
2015-02-04T03:52:23Z nikki93 joined #lisp
2015-02-04T03:53:39Z Bike: i glanced at their knowledge db whatever, plenty about stack sizes but nothing about straight segfaults
2015-02-04T03:55:02Z JokesOnYou77: Yeah, it also looks like the defualt heap is pretty big
2015-02-04T03:57:00Z JokesOnYou77: I think the default heap is 4TB :P
2015-02-04T03:57:19Z ynniv: |3b|: why would I not want that to be true?
2015-02-04T03:57:49Z Bike: ...well do you actually have 4 TB of ram?
2015-02-04T03:58:00Z PaulCapestany joined #lisp
2015-02-04T03:58:30Z JokesOnYou77: Also, if it's a stack overflow I'll get a prompt for a restart.  Where are the objects that file is creating going?  Heap, right?
2015-02-04T03:58:40Z JokesOnYou77: No, only 24GB :<
2015-02-04T03:58:44Z Bike: probably, yes.
2015-02-04T03:59:51Z Bike: it really seems like something lispworks support would like to know about http://www.lispworks.com/support/bug-report.html
2015-02-04T04:02:27Z JokesOnYou77: I can do that.  I also have an e-mail for a LW consultant through work, may reach out to him.
2015-02-04T04:02:33Z ynniv: Can I deftype the arguments to a function and then use that type in the ftype for that function? The function type seems to want a list in the first position and won’t let me use my type.
2015-02-04T04:02:50Z Bike: ynniv: you can. the first part of type is a list of arguments, you see.
2015-02-04T04:04:38Z beach joined #lisp
2015-02-04T04:04:47Z beach: Good morning everyone!
2015-02-04T04:04:47Z minion: beach, memo from JokesOnYou77: After well over an hour it's still running.  I'm going to leave it over night and tomorrow while I'm at work.  But if it doesn't finish, what should I run to get some useful information out of it?
2015-02-04T04:05:07Z JokesOnYou77: beach, we were just talking about you
2015-02-04T04:05:13Z Bike: log from the last few minutes is relevant to your test, beach.
2015-02-04T04:05:17Z beach: JokesOnYou77: Oh?
2015-02-04T04:05:27Z beach: OK, let me check.
2015-02-04T04:05:31Z nyef: Hello beach.
2015-02-04T04:05:34Z cmack joined #lisp
2015-02-04T04:05:53Z nyef: ... And good night. I'm pretty much done for the day.
2015-02-04T04:05:57Z ynniv: Bike: (deftype arguments () '(cons number (cons string))) (declaim (ftype (function arguments t) functionname)) wants arguments to be a list
2015-02-04T04:06:04Z beach: 'night nyef.
2015-02-04T04:06:12Z nyef quit (Quit: G'night all)
2015-02-04T04:06:38Z JokesOnYou77: g'night nyef
2015-02-04T04:06:54Z Bike: ynniv: right, because it's a lambda list, you can use &rest and stuff
2015-02-04T04:07:18Z beach: JokesOnYou77: You can do one thing for me: Remove all the OPTIMIZE declarations and run it again.  Hopefully it will then not segfault but tell us what is wrong.
2015-02-04T04:07:33Z ynniv: sure, but my function doesn’t use &rest etc.
2015-02-04T04:08:10Z drmeister: Hi beach - I was just composing an email to you.
2015-02-04T04:08:26Z beach: drmeister: Is that good or bad?
2015-02-04T04:08:33Z Bike: ynniv: it's part of the syntax of the 'function' type, i'm saying.
2015-02-04T04:08:43Z drmeister: Regarding the F->M instruction and the V datum.
2015-02-04T04:08:56Z JokesOnYou77: HAHA! Stack overflow!  Extending the stack by 3x
2015-02-04T04:08:57Z drmeister: Good - I'm just trying to understand how to implement them.
2015-02-04T04:08:59Z Bike: the first argument to the type is not a type specifier, it's a lambda list type specifier gizmo
2015-02-04T04:09:08Z drmeister: What is the "V" datum?
2015-02-04T04:09:13Z beach: JokesOnYou77: That's good.  Thanks.
2015-02-04T04:09:20Z beach: drmeister: Multiple values.
2015-02-04T04:09:21Z ynniv: Right. I’m asking if there’s a way around that.
2015-02-04T04:09:38Z Bike: you'd have to define your own function type
2015-02-04T04:09:47Z JokesOnYou77: hmmm....I may have to close some browser tabs to get this to finish lol
2015-02-04T04:09:54Z Bike: like (deftype ynniv-function () '(ftype (number string) t))
2015-02-04T04:10:09Z drmeister: beach: Now I have an array of for multiple values and a count of the number of multiple values.  Would "V" represent all of that?
2015-02-04T04:10:09Z cmack quit (Ping timeout: 250 seconds)
2015-02-04T04:10:21Z beach: drmeister: Sure, that would be fine.
2015-02-04T04:10:54Z beach: drmeister: What do you return from a function if only one value is returned?
2015-02-04T04:11:10Z drmeister: What does the F->M instruction do?   Take a single object and set it as  the first and only multiple-value ?
2015-02-04T04:11:34Z beach: It takes a fixed number of objects and turns it into multiple values.
2015-02-04T04:11:42Z beach: F is for "fixed"
2015-02-04T04:12:22Z drmeister: When I compile:   (cleavir-compile nil '(values 1 2 3))
2015-02-04T04:12:35Z drmeister: It doesn't use the F->M instruction - it calls the VALUES function.
2015-02-04T04:12:53Z beach: That's because we don't do inlining.
2015-02-04T04:12:57Z drmeister: I thought it should use the F->M instruction so I was surprised.
2015-02-04T04:13:26Z drmeister: Oh, if the values function were inlined it would inline an F->M instruction?
2015-02-04T04:13:26Z ynniv: Bike: Oh, “function” is only there to make a list out of a lambda list
2015-02-04T04:13:39Z beach: drmeister: You are allowed to treat VALUES as a special operator.  Then you can do what you want.
2015-02-04T04:14:14Z Bike: ynniv: what?
2015-02-04T04:14:26Z drmeister: That's ok, I'd rather have inlining take care of it. Additional special operators give me gas.
2015-02-04T04:14:40Z ynniv: You used ftype without function because the list was not a literal lambda list
2015-02-04T04:15:07Z drmeister: I think I see what to do with F->M and V
2015-02-04T04:15:14Z Bike: oh, i mistyped. it should be (defype ynniv () '(function (number string) t)), my mistake.
2015-02-04T04:15:17Z drmeister: I'll give it a whirl
2015-02-04T04:15:57Z JokesOnYou77: beach, it finished!  Must have been a stack overflow all along, very odd.
2015-02-04T04:16:06Z beach: JokesOnYou77: Yes, it can happen.
2015-02-04T04:16:10Z beach: Thanks for doing that.
2015-02-04T04:16:37Z drmeister: beach: Oh, I know what confused me.  The (defmethod translate-branch-instruction ((instruction cleavir-ir:return-instruction) inputs output)
2015-02-04T04:17:08Z beach: drmeister: Because RETURN is not a branch instruction?
2015-02-04T04:17:14Z qubitnerd joined #lisp
2015-02-04T04:17:15Z drmeister: The inputs are a list of values - is that correct?
2015-02-04T04:17:21Z JokesOnYou77: np, it's supposed to have *STACK-OVERFLOW-BEHAVIOR* set to :ERROR though so I should get a restart.  I didn't realize the compiler settings could change that.  I'll email you the output presently.
2015-02-04T04:18:02Z beach: drmeister: Not values.  Either of type DATUM or already translated to Common Lisp variables.  I forget which.
2015-02-04T04:18:17Z drmeister: No - I'm ok with it being a branch-instruction - although, yeah - when I think about it shouldn't it be a simple-instruction?
2015-02-04T04:18:18Z beach: JokesOnYou77: Thank yo very much.
2015-02-04T04:18:22Z ynniv: Bike: Bah. So I can’t wrap up the function arguments into a single type that I can reuse later
2015-02-04T04:18:41Z beach: drmeister: I think because simple instruction is defined to have a single successor, whereas RETURN has 0.
2015-02-04T04:18:57Z drmeister: Got it.
2015-02-04T04:19:07Z Bike: ynniv: you could have (defvar *arguments* '(number string)) (deftype ynniv () `(function ,*arguments* t)) or such
2015-02-04T04:19:11Z beach: drmeister: What do return from a function that returns a single value?
2015-02-04T04:19:14Z Bike: ynniv: i'm not sure i understand why you'd want to reuse this
2015-02-04T04:19:18Z beach: drmeister: Do you still return an array?
2015-02-04T04:20:16Z drmeister: beach: No, every function returns a structure that contains a single value and a count of how many values there are.  The second, third, and so on values are in the multiple value array.
2015-02-04T04:20:23Z MrWoohoo joined #lisp
2015-02-04T04:20:53Z beach: drmeister: Wow, that is not going to be fast.
2015-02-04T04:20:54Z drmeister: I could return three or five values and a count of the number of values.
2015-02-04T04:21:20Z beach: drmeister: Where do you allocate the structure and the array?
2015-02-04T04:21:34Z drmeister: beach: It's a little more involved than that.
2015-02-04T04:21:43Z beach: Oh, OK.
2015-02-04T04:21:55Z drmeister: In the simplest implementation the structure is on the stack and it's returned by value.
2015-02-04T04:22:09Z drmeister: The LLVM backend may optimize structure return and put it into registers.
2015-02-04T04:22:16Z ynniv: Bike: I know, it’s a weird case. I want to use (vector arguments *) as well as define the lambda list of the function to be arguments so that they don’t diverge.
2015-02-04T04:22:20Z beach: drmeister: I see.
2015-02-04T04:22:24Z drmeister: I could return five values and a count and LLVM could put them all in registers.
2015-02-04T04:22:43Z drmeister: Yes, it's similar to passing a count and the first five values in registers.
2015-02-04T04:22:44Z Bike: that... is extraordinarily bizarre
2015-02-04T04:23:04Z Bike: ynniv: you know that the vector will almost certainly be upgraded to T, right?
2015-02-04T04:23:19Z drmeister: I'm not sure that LLVM does this optimization of multiple return values in registers on X86-64
2015-02-04T04:23:19Z ynniv: Not yet
2015-02-04T04:23:30Z ynniv: what rules guide that?
2015-02-04T04:23:43Z drmeister: beach: It's something I plan to investigate once I get things up and running.
2015-02-04T04:23:44Z Bike: try (upgraded-array-element-type '(cons number (cons string null)))
2015-02-04T04:23:56Z beach: drmeister: Surely, there must be a fixed way of doing it.  This is part of the call/return protocol after all.  I don't see how LLVM can choose on a per-function basis how to return values.
2015-02-04T04:24:17Z Bike: (btw: (deftype list (&rest elems) (if (null elems) 'null `(cons ,(first elems) (list ,@(rest elems))))))
2015-02-04T04:24:30Z drmeister: Right - there is a return convention just like there is a calling convention.
2015-02-04T04:24:52Z beach: drmeister: That's what I am asking.  What is that convention?
2015-02-04T04:25:34Z drmeister: I don't know.   LLVM handles the details.   I'll summarize.
2015-02-04T04:25:40Z drmeister: Here's my calling convention:
2015-02-04T04:26:01Z drmeister: define internal void @REPL({ {}*, i32 }* sret %result-ptr, {}* %closed-af-ptr, i32 %nargs, {}* %farg0, {}* %farg1, {}* %farg2, {}* %farg3, {}* %farg4) {
2015-02-04T04:26:07Z ynniv: Bike: Yeah, comes back T. Guess I’m too far in the weeds on this one.
2015-02-04T04:26:33Z beach: drmeister: OK, thanks.  All Greek to me.
2015-02-04T04:26:47Z beach: drmeister: Forget I asked! :)
2015-02-04T04:26:59Z drmeister: The first argument %result-ptr has the attribute "sret" - this means "structure return".  The simplest implementations just treat it like a pointer to a struct on the stack and fill it - slow.
2015-02-04T04:28:00Z drmeister: Better backend implementations turn it into returning multiple components of the struct in registers.   Exactly which registers - that's described in the calling convention for X86-64 - I don't know what it is.
2015-02-04T04:28:06Z drmeister: Yeah - it's messy.
2015-02-04T04:28:42Z beach: But there is a guarantee that it will do the same thing for each function?
2015-02-04T04:28:51Z drmeister: Yes.
2015-02-04T04:29:22Z k-dawg joined #lisp
2015-02-04T04:29:24Z drmeister: Whatever it does if you return the same struct and give it the sret attribute, it will do the same for each function.
2015-02-04T04:30:09Z beach: drmeister: At some point, you might want to check what it does, because the performance of such a thing is crucial to good overall performance.
2015-02-04T04:30:14Z drmeister: The way I understand it - using that approach I can pass multiple arguments into a function through registers and return multiple values from a function through registers - fast.
2015-02-04T04:30:42Z beach: OK, that's good.
2015-02-04T04:30:46Z drmeister: beach: Right - but it's out of my hands to some degree.  Since I must interoperate with C++ I have to live with what the compiler does.
2015-02-04T04:31:01Z drmeister: C functions only return one value.
2015-02-04T04:31:07Z ynniv quit (Quit: ynniv)
2015-02-04T04:31:47Z nikki93 quit (Remote host closed the connection)
2015-02-04T04:31:47Z beach: drmeister: You could have different calling conventions, and then adapt when Common Lisp calls C++ or the other way around.
2015-02-04T04:31:48Z drmeister: C++ functions can return structs by value and they can separate the parts of the struct into separate registers.   This last part I'm still trying to figure out.  But it could give me multiple return values in registers.
2015-02-04T04:31:53Z brucem: drmeister: there's a way to deal with that in LLVM.
2015-02-04T04:32:10Z drmeister: brucem: A way to deal with what in LLVM?
2015-02-04T04:32:50Z brucem: drmeister: MV
2015-02-04T04:33:14Z drmeister: What is it?  I think I know but I'm curious about how you deal with it
2015-02-04T04:33:16Z stux|RC quit (Ping timeout: 255 seconds)
2015-02-04T04:33:50Z brucem: drmeister: I think we do what beach suggested ... we have different calling conventions for Dylan code vs C / C++ ... I'd have to look again though as that may have changed at some point.
2015-02-04T04:35:05Z nikki93 joined #lisp
2015-02-04T04:35:30Z stux|RC-only joined #lisp
2015-02-04T04:36:17Z JokesOnYou77: Is there a way to get the lambda list (var names) of a function?
2015-02-04T04:36:32Z JokesOnYou77: *parameter names
2015-02-04T04:36:48Z brucem: drmeister: ahh ... it did change some since I last looked at how our LLVM back-end does this ... I'll have to read more to know better what exactly we do now.
2015-02-04T04:37:19Z Harag joined #lisp
2015-02-04T04:37:47Z Bike: JokesOnYou77: function-lambda-list, sometimes
2015-02-04T04:38:00Z Bike: JokesOnYou77: function-lambda-expression rather
2015-02-04T04:38:06Z beach: Yesterday, for kicks, I types "lock-free hash table" to the Google search engine, and I found a video of a talk by Cliff Click at Azul.  Apparently, his implementation is excellent.
2015-02-04T04:38:44Z brucem: beach: he has some good sources available as well.
2015-02-04T04:38:57Z beach: brucem: Java is it?
2015-02-04T04:39:02Z brucem: beach: for example: https://github.com/boundary/high-scale-lib
2015-02-04T04:39:15Z brucem: beach: Sure. But the ideas are extractable and understandable.
2015-02-04T04:39:31Z beach: Yeah, not complicated at all, at least according to the talk.
2015-02-04T04:40:23Z JokesOnYou77: Bike, thank you! function-lambda-list was exactly what I needed.
2015-02-04T04:40:23Z brucem: beach: to some extent, when you *also* get to control the compiler ... life gets easier. :)
2015-02-04T04:42:28Z beach: JokesOnYou77: Got your mail.  Thanks.
2015-02-04T04:42:42Z beach: brucem: Yes, I see.
2015-02-04T04:43:11Z brucem: beach: An issue though is that some of the trade-offs are not as fun for some use cases ... it is interesting and difficult.
2015-02-04T04:43:34Z beach: brucem: Do you remember the issues?
2015-02-04T04:44:16Z brucem: beach: Not sure with Click's work as it has been a while ... but some things gave up single core performance to scale better, so it isn't something you can just replace your core implementation with.
2015-02-04T04:44:34Z beach: brucem: I see.
2015-02-04T04:44:51Z beach: I would be willing to try his technique in SICL.
2015-02-04T04:45:04Z brucem: beach: Other interesting work in that area has come from Gil Tene.
2015-02-04T04:45:48Z beach: Oh, a colleague of Cliff's.
2015-02-04T04:45:59Z brucem: beach: Azul (where Click and Tene did this work) has a 768 core machine ... so this stuff was important to them for their product.
2015-02-04T04:46:13Z beach: Right.
2015-02-04T04:46:21Z brucem: but in our land of 2, 4, 8 and 16 and 32 core machines ... maybe less so at times. :)
2015-02-04T04:46:55Z beach: Maybe I can write an implementation of his technique in SBCL and compare it to the SBCL native hash tables.
2015-02-04T04:49:01Z brucem: beach: You might enjoy https://groups.google.com/forum/#!topic/mechanical-sympathy/7Xyxm95CXxE too
2015-02-04T04:49:17Z fragamus joined #lisp
2015-02-04T04:49:44Z beach: Ah, skiplists.  I like them.
2015-02-04T04:50:26Z Quadrescence quit (Quit: This computer has gone to sleep)
2015-02-04T04:51:48Z beach: The funny thing is, I have been putting off implementing hash tables for SICL, because they seemed messy.  Now I am highly motivated. :)
2015-02-04T04:53:22Z JokesOnYou77 would LOVE a 32-core machine
2015-02-04T04:53:34Z brucem: beach: I'll be interesting in hearing how that goes ... I have it on my list to replace our <table> one day.
2015-02-04T04:53:38Z k-dawg quit (Quit: This computer has gone to sleep)
2015-02-04T04:53:47Z beach: brucem: I'll keep you informed.
2015-02-04T04:54:43Z mogglebang quit (Quit: mogglebang)
2015-02-04T04:55:03Z fragamus quit (Quit: Computer has gone to sleep.)
2015-02-04T04:55:16Z beach: In SICL, I am thinking that in the old-generation of the heap, objects shouldn't move.  So rather than re-hashing after a GC, I am thinking I should promote the keys before sticking them in a hash table.  Then I don't have to re-hash.
2015-02-04T04:55:19Z nell quit (Ping timeout: 255 seconds)
2015-02-04T05:00:41Z JokesOnYou77: Is there a built-in to get the position of multiple items in a list?  It's only a line to implement but I was wondering if it's already there
2015-02-04T05:00:55Z nikki93 quit (Remote host closed the connection)
2015-02-04T05:01:58Z beach: I don't think there is anything.
2015-02-04T05:02:05Z nikki93 joined #lisp
2015-02-04T05:02:27Z frkout joined #lisp
2015-02-04T05:03:19Z JokesOnYou77: cool.  Adding to my helpers file :)
2015-02-04T05:03:32Z JokesOnYou77: MULTI-POSN
2015-02-04T05:06:40Z dmiles joined #lisp
2015-02-04T05:06:49Z gko__ joined #lisp
2015-02-04T05:07:25Z Pyridrym quit (Quit: leaving)
2015-02-04T05:07:27Z dmiles_afk quit (Ping timeout: 272 seconds)
2015-02-04T05:08:09Z innertracks quit (Quit: innertracks)
2015-02-04T05:13:42Z cluck quit (Remote host closed the connection)
2015-02-04T05:14:00Z drmeister: beach: What do you mean "promote the keys"?
2015-02-04T05:23:28Z BitPuffin joined #lisp
2015-02-04T05:23:38Z JokesOnYou77 quit (Ping timeout: 265 seconds)
2015-02-04T05:23:53Z gingerale joined #lisp
2015-02-04T05:24:07Z innertracks joined #lisp
2015-02-04T05:26:28Z eigenlicht quit (Quit: WeeChat 0.3.2)
2015-02-04T05:28:16Z innertracks quit (Client Quit)
2015-02-04T05:28:26Z BitPuffin quit (Ping timeout: 245 seconds)
2015-02-04T05:28:34Z nikki93 quit (Remote host closed the connection)
2015-02-04T05:30:59Z nikki93 joined #lisp
2015-02-04T05:31:11Z beach: drmeister: For SICL, I have planned to have 2 generations in the heap...
2015-02-04T05:31:19Z ben_vulpes quit (Ping timeout: 255 seconds)
2015-02-04T05:31:26Z beach: One generation is per-thread.  The other one is common to all threads.
2015-02-04T05:31:47Z beach: The per-thread generation uses a mark-and-compact collector.
2015-02-04T05:31:50Z yenda joined #lisp
2015-02-04T05:31:58Z beach: So the objects move after GC.
2015-02-04T05:32:15Z holomorph quit (Quit: holomorph)
2015-02-04T05:32:39Z beach: In the old generation, I can use a mark-and-sweep on the headers because they are all 2 words long.
2015-02-04T05:32:50Z beach: So in the old generation, objects don't move.
2015-02-04T05:33:27Z beach: In an EQ hash table, the key could be the address of the object.  But if the address changes, you have to re-hash.
2015-02-04T05:33:49Z beach: By promoting the key to the old generation, I know it will never again move, so I don't have to re-hash.
2015-02-04T05:34:13Z beach: [end of rant]
2015-02-04T05:34:18Z nikki93 quit (Remote host closed the connection)
2015-02-04T05:34:55Z yenda: hello everybody
2015-02-04T05:35:04Z beach: Hello yenda.
2015-02-04T05:37:46Z yenda: is there a way to add stepping to format iterations ? : (format "~[~a~]" -only-want-to-print-1-element-out-of-3-list) ?
2015-02-04T05:38:19Z Bike: (format "~a" (first list))?
2015-02-04T05:39:13Z yenda: no the list is a list of values like (1 2 3 4 5 6 7 8 9) and I want 1 4 7
2015-02-04T05:39:22Z metaf5 quit (Quit: WeeChat 0.4.2)
2015-02-04T05:39:31Z metaf5 joined #lisp
2015-02-04T05:40:01Z Bike: well, i'd extract the values you want, and use those?
2015-02-04T05:42:32Z |3b|: clhs ~*
2015-02-04T05:42:32Z specbot: http://www.lispworks.com/reference/HyperSpec/Body/22_cga.htm
2015-02-04T05:42:38Z emaczen: Is there a mapcdr? using (describe symbol) has not helped me find the answer
2015-02-04T05:43:09Z |3b|: (format nil "~{~a~2*~}" '(1 2 3 4 5 6 7 8 9)) -> "147"
2015-02-04T05:43:13Z Bike: What would mapcdr do?
2015-02-04T05:43:15Z beach: emaczen: What do you want to do?
2015-02-04T05:43:29Z yenda: thanks |3b| I was reading your link it's what I need
2015-02-04T05:43:43Z |3b|: clhs maplist
2015-02-04T05:43:43Z specbot: http://www.lispworks.com/reference/HyperSpec/Body/f_mapc_.htm
2015-02-04T05:43:43Z emaczen: beach: I just answered my own question by rethinking about the structure of cons cells
2015-02-04T05:43:58Z emaczen: I was thinking about mapping over pairs
2015-02-04T05:44:05Z JokesOnYou77 joined #lisp
2015-02-04T05:44:08Z ruste_ quit (Ping timeout: 245 seconds)
2015-02-04T05:44:08Z yenda: Bike: I just want it quick and dirty it's to save stuff in a file after lots of computations to display some stats
2015-02-04T05:44:14Z ben_vulpes joined #lisp
2015-02-04T05:44:16Z egp__ quit (Remote host closed the connection)
2015-02-04T05:44:17Z egp_ quit (Remote host closed the connection)
2015-02-04T05:44:25Z yenda: Bike: the whole format is "{\"~a\": {\"learning-rate\": ~a, \"momentum\": ~a, \"quadratic-limit\": ~a, \"quadratic-evolution\": [~{~$~^,~}], \"sqrt-mqe\": [~{~$~^,~}],  \"quadratic-error\": [~{~$~^,~}]}}"
2015-02-04T05:45:11Z yenda: the thing is that I use javascript to display charts and firefox dies because there is too much points
2015-02-04T05:45:32Z yenda: s/too much/too many
2015-02-04T05:45:52Z Bike: so you just want to print every third point or something?
2015-02-04T05:46:23Z yenda: probably 1 out of 100 to start with
2015-02-04T05:46:33Z egp_ joined #lisp
2015-02-04T05:46:53Z Bike: i'd just write a function to take every 1 of 100 of a list and use that.
2015-02-04T05:47:15Z Bike: oh, wait, tilde has relative positions. nvm
2015-02-04T05:47:26Z yenda: yes I'm going to use ~* first to see if the chart.js works properly
2015-02-04T05:48:19Z yenda: but I'll make a function later to make it better because I'm not sure I can take in account the fact that the number of points can be between 100 and at least 1 000 000
2015-02-04T05:48:45Z yenda: so the stepping should be a variable depending on that
2015-02-04T05:53:49Z JokesOnYou77: Write a moving average function with a variable period depending on the number of points and give that output to JS.
2015-02-04T05:55:31Z yenda: yes that's the plan. ~* doesn't seem to work
2015-02-04T05:56:57Z Jesin quit (Ping timeout: 245 seconds)
2015-02-04T05:56:58Z burtons quit (Ping timeout: 255 seconds)
2015-02-04T06:00:17Z beach: I guess I could use SXHASH to implement a portable hash table in SBCL, just so that I can test the performance.
2015-02-04T06:01:30Z yenda: I used hashtable yesterday for the first time in lisp, I really how sbcl allows you to explore it in details
2015-02-04T06:02:36Z beach: You really what?
2015-02-04T06:04:03Z JokesOnYou77: yenda, also, congratulations as it seems like ou've gotten really far with your project.
2015-02-04T06:11:13Z kobain quit (Quit: KVIrc 4.1.3 Equilibrium http://www.kvirc.net/)
2015-02-04T06:12:19Z yenda: JokesOnYou77: thanks I'm quite happy with the results, I had the presentation yesterday but I can't help to keep working on it. I want to make the code nice and implement what I learn. Last night I completely reworked it to use structures and make everything more readable
2015-02-04T06:13:01Z JokesOnYou77: yenda, Wow! that's awesome.
2015-02-04T06:14:23Z beach: yenda: What is your project again?
2015-02-04T06:15:31Z yenda: beach: it is suppose to be a project to compare the efficiency of single and multi-layered perceptron for digit recognition
2015-02-04T06:16:08Z beach: I see.  Is this work?
2015-02-04T06:17:07Z yenda: I didn't find the right parameters to have really good results on multi layered perceptron but you actually recognize the digits when you print the weights of the input layer
2015-02-04T06:17:55Z yenda: single layer does about 10% error
2015-02-04T06:19:00Z JokesOnYou77: yenda, what is F1?
2015-02-04T06:20:25Z yenda: beach: http://postimg.org/image/imdo0mvj7/ (note that it is only 1 neuron of the first layer so parts of the concept are not present)
2015-02-04T06:20:42Z yenda: JokesOnYou77: F1 ?
2015-02-04T06:21:13Z JokesOnYou77: F1 score: (/ (* 2 true-positive) (+ (* 2 true-positive) fale-positive false-negative))
2015-02-04T06:21:20Z JokesOnYou77: http://en.wikipedia.org/wiki/Sensitivity_and_specificity
2015-02-04T06:22:02Z pranavrc joined #lisp
2015-02-04T06:22:12Z JokesOnYou77: It's a good performance summary for AI systems (at least in NLP systems)
2015-02-04T06:22:16Z yenda: oh good point I'm gonna add it to the stats, I only display precision and recall (and store TP FP FN TN)
2015-02-04T06:22:52Z JokesOnYou77: Yeah, definitely add F1, it's a good way to combine precision and recall.
2015-02-04T06:23:12Z nikki93 joined #lisp
2015-02-04T06:26:25Z oleo__ quit (Quit: Verlassend)
2015-02-04T06:27:16Z JokesOnYou77: Success! I built an image of this darn project now I can test the full stack...or I can at leas see if the other code will match up :P
2015-02-04T06:27:48Z ehu joined #lisp
2015-02-04T06:28:18Z yenda: I tried with concept "0" F1 is 92%
2015-02-04T06:30:24Z JokesOnYou77: yenda, that's pretty darn good.  Is the layout just one input node for each pixel?
2015-02-04T06:32:39Z Jesin joined #lisp
2015-02-04T06:33:25Z yenda: JokesOnYou77: yes 784 inputs values for 28x28px bitmaps but it doesn't matter what you give it to learn because you define the network you want for the concept
2015-02-04T06:33:39Z emaczen quit (Quit: ERC Version 5.3 (IRC client for Emacs))
2015-02-04T06:34:06Z nikki93 quit (Remote host closed the connection)
2015-02-04T06:34:23Z JokesOnYou77: oh cool. So what is learning-rate?  The library I'm currently working with has that as well.
2015-02-04T06:34:38Z JokesOnYou77: (not a library I'm writing, one I'm using)
2015-02-04T06:35:01Z yenda: learning-rate defines the pace at which the networks learns
2015-02-04T06:35:14Z yenda: when it's low the corrections made are smaller
2015-02-04T06:36:04Z quazimodo quit (Ping timeout: 252 seconds)
2015-02-04T06:36:30Z mishoo joined #lisp
2015-02-04T06:37:08Z yenda: its effects are really problem dependants and that's why I'm trying to make it easy to visualize learning statistics of generic networks
2015-02-04T06:37:25Z JokesOnYou77: yenda, the weight adjustments made during back-propagation?
2015-02-04T06:37:35Z nikki93 joined #lisp
2015-02-04T06:37:40Z JokesOnYou77: hmmm...interesting
2015-02-04T06:37:43Z pt1 joined #lisp
2015-02-04T06:38:51Z yenda: yes the formula is (* learning-rate gradient input)
2015-02-04T06:39:04Z JokesOnYou77: I feel like I've seen an experiment with setting the learning-rate using something like simulated annealing in order to avoid local minima
2015-02-04T06:39:26Z JokesOnYou77: ok, that makes sense.
2015-02-04T06:39:50Z yenda: I added a momentum to the learning-rate it has interesting effects
2015-02-04T06:40:46Z JokesOnYou77: Cool.  How does it work?
2015-02-04T06:41:03Z smokeink joined #lisp
2015-02-04T06:42:08Z gingerale quit (Ping timeout: 264 seconds)
2015-02-04T06:42:16Z yenda: I'm gonna be honest with you it's magic
2015-02-04T06:43:24Z JokesOnYou77: :D
2015-02-04T06:43:29Z JokesOnYou77 loves magic
2015-02-04T06:43:32Z yenda: I coded that with only 15hours of sleep over 4 days so I have to look at what effects it has exactly
2015-02-04T06:43:40Z JokesOnYou77: lol ok
2015-02-04T06:43:51Z hvxgr quit (Ping timeout: 264 seconds)
2015-02-04T06:44:02Z wheelsucker quit (Ping timeout: 245 seconds)
2015-02-04T06:44:52Z quazimodo joined #lisp
2015-02-04T06:46:09Z JokesOnYou77: yenda, when you do, I would be interested in how momentum compares to adagrad
2015-02-04T06:46:15Z beach left #lisp
2015-02-04T06:46:26Z yenda: what is adagrad ?
2015-02-04T06:47:29Z manuel__ quit (Quit: manuel__)
2015-02-04T06:48:35Z yenda: (googling)
2015-02-04T06:48:49Z JokesOnYou77: Adaptive subgradient optimization.  I've never implemented it but it's supposed to be a standard for adjusting learning rates
2015-02-04T06:48:51Z pt1 quit (Remote host closed the connection)
2015-02-04T06:48:52Z JokesOnYou77: http://www.magicbroom.info/Papers/DuchiHaSi10.pdf
2015-02-04T06:48:59Z JokesOnYou77: http://www.quora.com/What-is-the-purpose-of-AdaGrad-for-stochastic-gradient-decent-neural-network-training
2015-02-04T06:51:18Z yenda: that's interesting because I was also thinking about dropping the learning rate and use per-neuron stats to make the network more elastic
2015-02-04T06:52:55Z manuel__ joined #lisp
2015-02-04T06:53:17Z JokesOnYou77: Yeah, I don't know how to actually implement it, but it may be an interesting idea for a perceptron.
2015-02-04T06:53:31Z JokesOnYou77: I assume there is literature on it somewhere as well
2015-02-04T06:53:59Z yenda: it's easy I think, I'll just forget about the maths and do it based on intuition
2015-02-04T06:54:03Z zacharias_ quit (Ping timeout: 264 seconds)
2015-02-04T06:54:11Z JokesOnYou77: rofl, ok...
2015-02-04T06:54:43Z BitPuffin joined #lisp
2015-02-04T06:55:21Z theseb quit (Quit: Leaving)
2015-02-04T06:55:56Z Harag quit (Ping timeout: 264 seconds)
2015-02-04T06:56:36Z yenda: I find it easier to work with symbols and concepts and put the math as downstream as possible
2015-02-04T06:57:38Z psy_ quit (Remote host closed the connection)
2015-02-04T06:57:49Z Cymew joined #lisp
2015-02-04T06:58:24Z JokesOnYou77: As the math gets incresingly complicated I'm finding that as well, but I really like to try and understand first if possible
2015-02-04T06:59:38Z yenda: that's what I did to implement the gradient descent, I had to understand the formulas. But for a single layer I actually had the algorithm right and working already. It was doing stuff similar to the maths but not exactly
2015-02-04T06:59:49Z BitPuffin quit (Ping timeout: 264 seconds)
2015-02-04T07:00:29Z yenda: but the first versions are really unuasable it was my first lines of lisp, I was using stuff like imbricated let because I didnt know about let* for instance :D
2015-02-04T07:00:59Z munksgaard joined #lisp
2015-02-04T07:01:10Z yenda: I still don't do macros because I don't get it yet
2015-02-04T07:01:21Z JokesOnYou77: Yeah, let* is amazing.  I just wish they had it in Mathematica.
2015-02-04T07:02:14Z xificurC joined #lisp
2015-02-04T07:03:04Z nikki93 quit (Remote host closed the connection)
2015-02-04T07:03:38Z nikki93 joined #lisp
2015-02-04T07:05:07Z JokesOnYou77: yenda, I've been working in Lisp off and on for 3 years and I only just started understanding macros last weekend.  You've got plenty of time. Although, knowing what I know now, sooner may be better.
2015-02-04T07:05:08Z yenda: I wish I had more time for this neural net I have so many ideas to improve it
2015-02-04T07:05:44Z JokesOnYou77: But maybe the next project will be just as cool?  I don't know where ou're working, but this sounds liek a prtty cool project
2015-02-04T07:06:26Z yenda: JokesOnYou77: I understand what they are and what they do but I didn't learn how to actually write them and I haven't encountered a "I need a macro for that" moment. Or maybe a few but I was like "and I also need to fix 20 bugs so lets fix the bugs first"
2015-02-04T07:06:50Z JokesOnYou77: lol
2015-02-04T07:07:21Z nikki93 quit (Remote host closed the connection)
2015-02-04T07:07:29Z JokesOnYou77: I think the rule of thumb typically is, If you think you need a macro, you should probably look more carefully.
2015-02-04T07:08:44Z cocosp joined #lisp
2015-02-04T07:08:54Z Karl_Dscc joined #lisp
2015-02-04T07:11:47Z yenda: the loop macro is already breaking my head
2015-02-04T07:13:06Z JokesOnYou77: Honestly, when I first used it I was really proud of myself and I uesd it everywhere, but I have to look up the syntax every time I use it so I mostly just use DOLIST or DOTIMES most of the time.
2015-02-04T07:13:45Z meiji11 quit (Remote host closed the connection)
2015-02-04T07:13:49Z hvxgr joined #lisp
2015-02-04T07:15:55Z JokesOnYou77: yenda, I'm also experimenting with this for passing some information: http://sexpdata.readthedocs.org/en/latest/ :)
2015-02-04T07:18:44Z munksgaard quit (Ping timeout: 264 seconds)
2015-02-04T07:19:25Z nikki93 joined #lisp
2015-02-04T07:21:11Z JokesOnYou77: Well, that's it for me.  Good night all.
2015-02-04T07:23:08Z yenda: good night ! here it's 8am
2015-02-04T07:25:45Z nikki93 quit (Remote host closed the connection)
2015-02-04T07:26:09Z JokesOnYou77 quit (Quit: Ex-Chat)
2015-02-04T07:29:25Z Shinmera joined #lisp
2015-02-04T07:29:41Z pt1 joined #lisp
2015-02-04T07:31:12Z jasom: Is there a way to delete a method?
2015-02-04T07:31:38Z Karl_Dscc quit (Remote host closed the connection)
2015-02-04T07:33:20Z zacharias joined #lisp
2015-02-04T07:34:43Z Shinmera: clhs remove-method
2015-02-04T07:34:43Z specbot: http://www.lispworks.com/reference/HyperSpec/Body/f_rm_met.htm
2015-02-04T07:35:00Z Shinmera: And if you want convenience http://plaster.tymoon.eu/view/3N
2015-02-04T07:35:30Z Shinmera: Alternatively you can also use M-x slime-inspect on the generic function and remove the method in the menu.
2015-02-04T07:35:33Z manuel__ quit (Quit: manuel__)
2015-02-04T07:35:48Z Beetny joined #lisp
2015-02-04T07:35:55Z jasom: Shinmera: slime inspector ftw
2015-02-04T07:36:08Z jasom: not sure how I missed remove-method; I think that's a signal that it's bedtime for me
2015-02-04T07:36:17Z Shinmera: I just woke up!
2015-02-04T07:36:40Z Shinmera: And I missed beach, darn.
2015-02-04T07:37:37Z cocosp quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
2015-02-04T07:37:43Z fragamus joined #lisp
2015-02-04T07:38:03Z agumonkey quit (Quit: ZNC - http://znc.in)
2015-02-04T07:39:14Z agumonkey joined #lisp
2015-02-04T07:41:31Z stepnem joined #lisp
2015-02-04T07:45:32Z cocosp joined #lisp
2015-02-04T07:47:10Z MutSbeta quit (Quit: Leaving.)
2015-02-04T07:50:31Z Posterdati joined #lisp
2015-02-04T07:51:36Z nikki93 joined #lisp
2015-02-04T07:52:05Z Posterdati: froggey: hi
2015-02-04T07:52:48Z fridim_ joined #lisp
2015-02-04T07:55:14Z cadadar joined #lisp
2015-02-04T07:56:27Z cdtaylor joined #lisp
2015-02-04T07:58:57Z munksgaard joined #lisp
2015-02-04T07:59:42Z lispyone_ joined #lisp
2015-02-04T08:01:29Z d4ryus___ joined #lisp
2015-02-04T08:02:33Z mrSpec joined #lisp
2015-02-04T08:02:47Z lispyone_ quit (Client Quit)
2015-02-04T08:03:10Z fantazo joined #lisp
2015-02-04T08:03:31Z kushal quit (Ping timeout: 256 seconds)
2015-02-04T08:04:48Z d4ryus_ quit (Ping timeout: 252 seconds)
2015-02-04T08:04:55Z lispyone joined #lisp
2015-02-04T08:06:49Z Karl_Dscc joined #lisp
2015-02-04T08:08:52Z stepnem quit (Ping timeout: 240 seconds)
2015-02-04T08:09:22Z cdtaylor quit (Remote host closed the connection)
2015-02-04T08:09:39Z Anarch quit (Ping timeout: 264 seconds)
2015-02-04T08:09:40Z zadock joined #lisp
2015-02-04T08:09:49Z Harag joined #lisp
2015-02-04T08:09:57Z cdtaylor joined #lisp
2015-02-04T08:09:58Z antonv joined #lisp
2015-02-04T08:15:18Z moei quit (Quit: Leaving...)
2015-02-04T08:16:34Z stepnem joined #lisp
2015-02-04T08:17:07Z kushal joined #lisp
2015-02-04T08:17:55Z kuzy000_ joined #lisp
2015-02-04T08:18:08Z moei joined #lisp
2015-02-04T08:20:58Z qubitnerd quit (Read error: Connection reset by peer)
2015-02-04T08:22:29Z salv0 quit (Quit: Leaving.)
2015-02-04T08:23:09Z fragamus quit (Quit: Computer has gone to sleep.)
2015-02-04T08:24:09Z krrrcks_ left #lisp
2015-02-04T08:27:39Z resttime quit (Quit: resttime)
2015-02-04T08:27:44Z munksgaard quit (Ping timeout: 245 seconds)
2015-02-04T08:31:55Z angavrilov joined #lisp
2015-02-04T08:34:23Z cocosp quit (Quit: Textual IRC Client: www.textualapp.com)
2015-02-04T08:36:11Z hellofunk joined #lisp
2015-02-04T08:36:23Z harish quit (Ping timeout: 256 seconds)
2015-02-04T08:36:55Z Harag quit (Ping timeout: 252 seconds)
2015-02-04T08:37:26Z radioninja quit (Ping timeout: 252 seconds)
2015-02-04T08:39:47Z Harag joined #lisp
2015-02-04T08:40:27Z BitPuffin joined #lisp
2015-02-04T08:44:06Z mvilleneuve joined #lisp
2015-02-04T08:45:11Z BitPuffin quit (Ping timeout: 265 seconds)
2015-02-04T08:45:57Z hiroakip joined #lisp
2015-02-04T08:46:33Z kushal quit (Quit: Leaving)
2015-02-04T08:47:44Z grungier quit (Ping timeout: 245 seconds)
2015-02-04T08:48:09Z sshirokov quit (Ping timeout: 245 seconds)
2015-02-04T08:49:18Z arenz joined #lisp
2015-02-04T08:49:55Z lieven joined #lisp
2015-02-04T08:52:03Z fantazo quit (Ping timeout: 245 seconds)
2015-02-04T08:53:34Z fantazo joined #lisp
2015-02-04T08:57:18Z sshirokov joined #lisp
2015-02-04T08:57:58Z frkout_ joined #lisp
2015-02-04T08:59:33Z redeemed joined #lisp
2015-02-04T09:00:00Z moore33 joined #lisp
2015-02-04T09:01:19Z fmu quit (Read error: Connection reset by peer)
2015-02-04T09:01:50Z fmu joined #lisp
2015-02-04T09:01:53Z frkout quit (Ping timeout: 256 seconds)
2015-02-04T09:02:33Z salv0 joined #lisp
2015-02-04T09:02:38Z mutley89 joined #lisp
2015-02-04T09:03:12Z grungier joined #lisp
2015-02-04T09:03:44Z fmu quit (Read error: Connection reset by peer)
2015-02-04T09:04:36Z fmu joined #lisp
2015-02-04T09:05:37Z munksgaard joined #lisp
2015-02-04T09:10:39Z pegu quit (Read error: Connection reset by peer)
2015-02-04T09:10:54Z pegu joined #lisp
2015-02-04T09:11:36Z Lokathor quit (Ping timeout: 246 seconds)
2015-02-04T09:12:16Z Harag quit (Ping timeout: 252 seconds)
2015-02-04T09:12:39Z psy_ joined #lisp
2015-02-04T09:20:51Z yeticry quit (Ping timeout: 250 seconds)
2015-02-04T09:21:53Z yeticry joined #lisp
2015-02-04T09:23:30Z nikki93 quit (Remote host closed the connection)
2015-02-04T09:33:37Z egp_ quit (Ping timeout: 245 seconds)
2015-02-04T09:36:12Z egp_ joined #lisp
2015-02-04T09:40:55Z fantazo quit (Ping timeout: 250 seconds)
2015-02-04T09:42:59Z fantazo joined #lisp
2015-02-04T09:43:01Z nikki93 joined #lisp
2015-02-04T09:43:39Z Karl_Dscc quit (Remote host closed the connection)
2015-02-04T09:45:17Z defaultxr quit (Quit: gnight)
2015-02-04T09:46:15Z profess quit (Ping timeout: 264 seconds)
2015-02-04T09:50:31Z munksgaard quit (Ping timeout: 245 seconds)
2015-02-04T09:50:32Z egp_ quit (Ping timeout: 264 seconds)
2015-02-04T09:52:06Z scymtym_ joined #lisp
2015-02-04T09:52:31Z munksgaard joined #lisp
2015-02-04T09:55:27Z Karl_Dscc joined #lisp
2015-02-04T10:04:51Z Harag joined #lisp
2015-02-04T10:06:10Z robot-beethoven quit (Quit: ERC Version 5.3 (IRC client for Emacs))
2015-02-04T10:06:27Z nikki93 quit (Remote host closed the connection)
2015-02-04T10:07:00Z nikki93 joined #lisp
2015-02-04T10:08:37Z nikki93 quit (Read error: Connection reset by peer)
2015-02-04T10:08:58Z nikki93 joined #lisp
2015-02-04T10:10:19Z nikki93 quit (Remote host closed the connection)
2015-02-04T10:16:56Z pacon joined #lisp
2015-02-04T10:20:28Z Ralt: Shinmera: did you see my ping?
2015-02-04T10:20:39Z alchemis7 quit (Ping timeout: 256 seconds)
2015-02-04T10:21:32Z tharugrim quit (Ping timeout: 245 seconds)
2015-02-04T10:24:24Z lieven quit (Ping timeout: 246 seconds)
2015-02-04T10:24:37Z nikki93 joined #lisp
2015-02-04T10:27:43Z zadock quit (Quit: Leaving)
2015-02-04T10:28:02Z nikki93 quit (Remote host closed the connection)
2015-02-04T10:28:23Z radioninja joined #lisp
2015-02-04T10:28:32Z scymtym_ quit (Ping timeout: 252 seconds)
2015-02-04T10:29:13Z BitPuffin joined #lisp
2015-02-04T10:29:31Z pjb` is now known as pjb
2015-02-04T10:29:49Z egp_ joined #lisp
2015-02-04T10:29:55Z kcj quit (Read error: Connection reset by peer)
2015-02-04T10:33:32Z Ethan- joined #lisp
2015-02-04T10:33:37Z BitPuffin quit (Ping timeout: 245 seconds)
2015-02-04T10:34:37Z tharugrim joined #lisp
2015-02-04T10:37:11Z hiroakip quit (Ping timeout: 245 seconds)
2015-02-04T10:41:22Z arpunk quit (Ping timeout: 255 seconds)
2015-02-04T10:41:53Z egp_ quit (Ping timeout: 250 seconds)
2015-02-04T16:38:53Z ccl-logbot joined #lisp
2015-02-04T16:38:53Z 
2015-02-04T16:38:53Z names: ccl-logbot nikki93 mvilleneuve stardiviner AntiSpamMeta intinig davazp ahungry yrk profess a20150202 nell rick-monster burtons cluck ruste ahungry_ xyh manuel__ rhllor oleo Pyridrym resttime TDog Karl_Dscc LiamH killmaster salva ebrasca quazimodo ggole BitPuffin ynniv nyef Harag urandom__ ``Erik william3 hiyosi lispyone badkins Icon__ Xach vanila hitecnologys impulse frkout butyoudonot egp_ dead__ jgrant H4ns eudoxia trn brucem JuanDaugherty vsync_ froggey_
2015-02-04T16:38:53Z names: honkfestival josteink PinealGl1ndOptic BeLucid_ luis- jim87864` rtra specbot heurist` vaporatorius keen__________53 hellofunk hiroakip scymtym_ harish MutSbeta attila_lendvai zadock yeticry fantazo psy_ pegu fmu grungier mutley89 redeemed sshirokov angavrilov moei kuzy000_ stepnem cdtaylor mrSpec d4ryus___ cadadar fridim_ Posterdati agumonkey Shinmera hvxgr xificurC Cymew mishoo Jesin ehu ben_vulpes metaf5 yenda gko__ dmiles stux|RC-only MrWoohoo
2015-02-04T16:38:53Z names: PaulCapestany theos pillton Qudit314159 ASau Takumo aap pjb ft dpb9cpu MoALTz_ marko-v tsumetai bb010g gluegadget cojy_ splittist endou______ NhanH gz victor_lowther xan_ kjeldahl ecraven ered sivoais mathrick ryankara1on arrdem selat Vutral foom Kruppe jdz joneshf-laptop ryankarason faheem_ Patzy wooden Soft Khisanth adlai salv01 backupthrick bjorkintosh lemoinem edran_ troydm dilated_dinosaur Natch leo bege Vivitron loke tristero jtz bytecrawler Zhivago
2015-02-04T16:38:53Z names: rvirding Jubb cpc26 aksatac farhaven mingvs d4gg4d jasom sjl someone gigetoo soggybread danlentz kbtr ghard hlavaty enfors Amaan newcup jpanest Sgeo joast billstclair milosn easye araujo ThePhoeron __main__ s_e GuilOooo hugod thomas ivan\ ozzloy john-mcaleely fikusz_ j_king hratsimihah nightfly_ eazar001 funnel mgv moomin-aba_ ssake_ sfa theBlackDragon zbrown alakran Mandus_ replcated_ kephra vlion spacebat nydel gabot Tordek_ The_third_man zickzackv mood
2015-02-04T16:38:53Z names: ck_ yawniek_ wasamasa eMBee c74d nisstyre misv_ alex6407 nitrix dunib cods yrdz galdor hyoyoung ivan4th` _death mhi^ setheus emma anunnaki nopf dan64- justinmcp_ yauz_2 dfox gensym pok j0ni mikaelj Colleen CrazyEddy oconnore djinni` shwouchk copec qlkzy Borbus mtd jonh hellome clop kyl_______ jrm wenincode segmond johs enn Fade phadthai katco joshmcmillan_ Krystof z0d jacsib daimrod cmatei jackdaniel aretecode vlnx renopt sismondi ggherdov gregburd_ Neet
2015-02-04T16:38:53Z names: K1rk tstc diginet rotty_ rvchangue_ zacts scymtym lea quasisane jaimef schjetne rtoym kanru otwieracz isoraqathedh snafuchs jsnell larme sharkz gniourf emlow BlastHardcheese dlowe Grue` GGMethos charlie swflint ircbrows- gko_ PuercoPop pchrist abbe tomaw stopbyte whartung yorick dim gabc cyraxjoe clog axion mearnsh bcoburn drmeister zeroish Oladon Tuxedo necronian JohnRimbaud capitaomorte mietek phf TristamWrk AeroNotix Plazma sellout blahzik akkad
2015-02-04T16:38:53Z names: teiresias oGMo Rudolph-Miller_ yeltzooo mmathers sid_cypher kirin` p_l eagleflo joga Cheery reb`` acieroid smull_ Ober bobbysmith0071 SHODAN Ralt decent flip214 kalzz aerique trigen cpt_nemo ramus brandonz_ Bike minion zbigniew redline6561 srcerer nightshade427 schoppenhauer girrig notty viaken vhost- zymurgy |3b| Riviera antoszka cibs ConstantineXVI chameco sbryant nicdev jayne Subfusc Neptu renard_ arrsim fe[nl]ix drdo lpaste sytse Blkt joshe p_l|backup
2015-02-04T16:38:53Z names: clop2 nitro_idiot tkd micro finnrobi housel xorpse samebchase cmbntr ferada cross les peccu2 stokachu tokenrove brent80_plow xristos eak tessier
2015-02-04T16:42:33Z nikki93 quit (Ping timeout: 250 seconds)
2015-02-04T16:45:55Z JuanitoJons joined #lisp
2015-02-04T16:47:12Z DrCode joined #lisp
2015-02-04T16:48:04Z theseb joined #lisp
2015-02-04T16:49:23Z nikki93 joined #lisp
2015-02-04T16:53:09Z eudoxia quit (Quit: Leaving)
2015-02-04T16:53:40Z EvW joined #lisp
2015-02-04T16:54:21Z pillton quit (Ping timeout: 272 seconds)
2015-02-04T16:56:57Z attila_lendvai quit (Quit: Leaving.)
2015-02-04T16:57:57Z enitiz joined #lisp
2015-02-04T16:57:57Z egp_ quit (Read error: Connection reset by peer)
2015-02-04T16:58:41Z paddymahoney joined #lisp
2015-02-04T16:58:43Z paddymahoney quit (Max SendQ exceeded)
2015-02-04T17:00:17Z nikki93 quit (Ping timeout: 245 seconds)
2015-02-04T17:00:31Z ggole quit
2015-02-04T17:00:44Z egp_ joined #lisp
2015-02-04T17:00:51Z william3 quit (Remote host closed the connection)
2015-02-04T17:03:10Z Mon_Ouie joined #lisp
2015-02-04T17:03:32Z nikki93 joined #lisp
2015-02-04T17:05:08Z paddymahoney joined #lisp
2015-02-04T17:05:55Z enitiz quit (Ping timeout: 265 seconds)
2015-02-04T17:06:17Z radioninja joined #lisp
2015-02-04T17:06:52Z redeemed quit (Quit: q)
2015-02-04T17:07:01Z lispyone quit (Remote host closed the connection)
2015-02-04T17:08:22Z zadock quit (Remote host closed the connection)
2015-02-04T17:08:25Z innertracks joined #lisp
2015-02-04T17:09:07Z zadock joined #lisp
2015-02-04T17:10:56Z gingerale joined #lisp
2015-02-04T17:14:26Z pt1 joined #lisp
2015-02-04T17:16:02Z antonv joined #lisp
2015-02-04T17:16:04Z yrk quit (Ping timeout: 245 seconds)
2015-02-04T17:16:25Z enitiz joined #lisp
2015-02-04T17:17:52Z cadadar quit (Quit: Leaving.)
2015-02-04T17:18:08Z kobain joined #lisp
2015-02-04T17:18:33Z TDog_ joined #lisp
2015-02-04T17:19:09Z TDog quit (Ping timeout: 246 seconds)
2015-02-04T17:19:12Z kobain quit (Max SendQ exceeded)
2015-02-04T17:19:15Z TDog_ is now known as TDog
2015-02-04T17:19:29Z kobain joined #lisp
2015-02-04T17:22:45Z mvilleneuve quit (Quit: This computer has gone to sleep)
2015-02-04T17:23:53Z nikki93 quit (Ping timeout: 240 seconds)
2015-02-04T17:26:17Z Oddity joined #lisp
2015-02-04T17:29:31Z beach joined #lisp
2015-02-04T17:29:39Z beach: Good evening everyone!
2015-02-04T17:30:00Z xyh quit (Remote host closed the connection)
2015-02-04T17:30:10Z Shinmera: Hello beach
2015-02-04T17:30:17Z Shinmera: Did you get my email yesterday?
2015-02-04T17:30:31Z beach: Shinmera: I did.  Thank you very much!
2015-02-04T17:30:48Z nikki93 joined #lisp
2015-02-04T17:32:39Z JokesOnYou77 joined #lisp
2015-02-04T17:33:30Z Patzy quit (Ping timeout: 246 seconds)
2015-02-04T17:34:27Z Patzy joined #lisp
2015-02-04T17:34:41Z Shinmera: I would've liked to run it on XCL as well, but for some reason I can't even launch it on my machine, it immediately core-dumps.
2015-02-04T17:34:55Z Shinmera: The implementation I mean; I don't even get a REPL.
2015-02-04T17:35:14Z beach: Wow, what happened?
2015-02-04T17:36:15Z beach: I mean, presumably it worked in the past.
2015-02-04T17:37:06Z scharan joined #lisp
2015-02-04T17:37:20Z Shinmera: No, I never was able to get it to work
2015-02-04T17:37:28Z eudoxia joined #lisp
2015-02-04T17:37:35Z beach: Oh! :(
2015-02-04T17:37:37Z zadock quit (Quit: Leaving)
2015-02-04T17:37:39Z Shinmera: I can compile it seemingly without problems, but upon launch it just prints the header and then crashes.
2015-02-04T17:38:10Z JokesOnYou77: I broke my SBCL that way.
2015-02-04T17:38:46Z nyef: Hello beach.
2015-02-04T17:38:52Z lispyone joined #lisp
2015-02-04T17:38:55Z manuel__ quit (Quit: manuel__)
2015-02-04T17:40:15Z badkins quit
2015-02-04T17:42:33Z beach: I have been thinking about lock-free hash tables all day. :)
2015-02-04T17:42:35Z lieven joined #lisp
2015-02-04T17:43:08Z beach: I think it would be possible to test the idea using SXHASH to create some kind of implementation-independent version.
2015-02-04T17:44:00Z yrk joined #lisp
2015-02-04T17:44:31Z yrk quit (Changing host)
2015-02-04T17:44:31Z yrk joined #lisp
2015-02-04T17:44:31Z JokesOnYou77: A cursory google isn't explaining what it means to be lock-free, what is it?
2015-02-04T17:45:02Z oGMo: it means you don't have to lock to safely perform operations
2015-02-04T17:45:18Z JokesOnYou77: concurrent access?
2015-02-04T17:45:23Z beach: JokesOnYou77: It means you don't have to use locks (mutexes) even though you might have several simultaneous threads accessing the table.
2015-02-04T17:45:32Z Xach: JokesOnYou77: http://www.pvk.ca/ has a definition
2015-02-04T17:45:39Z oGMo: JokesOnYou77: more specifically concurrent access _without locking_
2015-02-04T17:45:52Z JokesOnYou77: Cool!
2015-02-04T17:47:17Z xyh joined #lisp
2015-02-04T17:47:20Z xyh: which is our jekyll ?
2015-02-04T17:47:28Z nikki93 quit (Ping timeout: 245 seconds)
2015-02-04T17:48:47Z Alfr joined #lisp
2015-02-04T17:49:08Z nikki93 joined #lisp
2015-02-04T17:51:11Z wasamasa: invent your own in <100 lines of lisp
2015-02-04T17:51:17Z wasamasa: that's how it usually goes
2015-02-04T17:51:39Z burtons quit (Read error: Connection reset by peer)
2015-02-04T17:52:02Z oGMo: xyh: https://github.com/redline6561/coleslaw ?
2015-02-04T17:52:34Z JuanDaugherty: and that works, right?
2015-02-04T17:52:55Z oGMo: what, coleslaw or lock-free structures?
2015-02-04T17:53:20Z JuanDaugherty: lock free access with many contenders
2015-02-04T17:53:23Z oGMo: i'm not sure why either would be in doubt, though i haven't tried the former :P
2015-02-04T17:53:57Z oGMo: JuanDaugherty: right, read up on lock-free algorithms and basically the link above
2015-02-04T17:55:23Z nyef: JuanDaugherty: Part of the design criteria for a lock-free algorithm is ensuring that no thread in contention gets stuck, no matter how many contenders there coulda been.
2015-02-04T17:55:59Z xyh: I mean coleslaw, I just forget its name.
2015-02-04T17:56:01Z taspat joined #lisp
2015-02-04T17:56:03Z manuel__ joined #lisp
2015-02-04T17:56:15Z JuanDaugherty: OK, see there's an equivocation. This what I might call locking that just works versus Walmark Locking
2015-02-04T17:57:34Z JuanDaugherty and noting I had heard the term, thought of it afresh just now though
2015-02-04T17:57:57Z MoALTz_ quit (Quit: Leaving)
2015-02-04T17:58:17Z MoALTz joined #lisp
2015-02-04T18:01:33Z theseb quit (Quit: Leaving)
2015-02-04T18:03:41Z lispyone quit (Remote host closed the connection)
2015-02-04T18:04:25Z lispyone joined #lisp
2015-02-04T18:04:57Z Posterdati: hi
2015-02-04T18:05:09Z JokesOnYou77: Am I being overly cautious in thinking that I'd be better off with a library to remove all URLs from open text input than trying to use a regex?
2015-02-04T18:05:10Z Posterdati: is it possible to create a db with postmoderrn?
2015-02-04T18:06:46Z Xach: Posterdati: yes.
2015-02-04T18:07:40Z mrSpec quit (Remote host closed the connection)
2015-02-04T18:08:17Z Xach: Posterdati: if you have permission, you can connect to an existing database (like "template1") and then issue "create database mydatabase;"
2015-02-04T18:09:44Z nikki93 quit (Ping timeout: 264 seconds)
2015-02-04T18:10:01Z ynniv: JokesOnYou77: Do you mean valid URLs, or just URL-like?
2015-02-04T18:10:53Z hitecnologys quit (Quit: hitecnologys)
2015-02-04T18:11:13Z Posterdati: Xach: ok, thanks
2015-02-04T18:11:38Z nikki93 joined #lisp
2015-02-04T18:12:00Z JokesOnYou77: I mean I'm trying to do text classification on social media text and I wanted to try and replace all URLs with some standin to see how it impacts the classifier
2015-02-04T18:12:55Z Bicyclidine joined #lisp
2015-02-04T18:13:08Z fragamus joined #lisp
2015-02-04T18:14:17Z badkins joined #lisp
2015-02-04T18:14:33Z otwieracz quit (Ping timeout: 265 seconds)
2015-02-04T18:15:02Z ynniv: For cases when validity matters you want to use a library. It sounds like if you have a few false positives or negatives it won’t greatly affect the output, so I would use something from http://stackoverflow.com/questions/161738/what-is-the-best-regular-expression-to-check-if-a-string-is-a-valid-url
2015-02-04T18:17:03Z JokesOnYou77: ynniv, good point, I guess I can afford to miss few...Would be best to get everything, but that's a loosing battle, thank you
2015-02-04T18:17:53Z enitiz quit (Ping timeout: 245 seconds)
2015-02-04T18:18:37Z EvW quit (Ping timeout: 250 seconds)
2015-02-04T18:19:04Z otwieracz joined #lisp
2015-02-04T18:19:36Z xyh quit (Remote host closed the connection)
2015-02-04T18:19:42Z Cymew quit (Ping timeout: 252 seconds)
2015-02-04T18:24:05Z holomorph joined #lisp
2015-02-04T18:24:50Z ghard quit (Ping timeout: 252 seconds)
2015-02-04T18:25:23Z nikki93 quit (Ping timeout: 245 seconds)
2015-02-04T18:26:39Z nikki93 joined #lisp
2015-02-04T18:27:22Z EvW joined #lisp
2015-02-04T18:28:45Z Posterdati: Xach: so, if I can connect to postgres db I could create other db... But it refuses to connect-toplevel!
2015-02-04T18:29:37Z enitiz joined #lisp
2015-02-04T18:29:48Z burtons joined #lisp
2015-02-04T18:30:51Z Xach: Posterdati: in what way does it refuse?
2015-02-04T18:31:20Z Posterdati: Database error 28P01: password authentication failed for user "postgres"
2015-02-04T18:31:38Z intinig quit (Remote host closed the connection)
2015-02-04T18:31:42Z Xach: Posterdati: Then you must use the correct password.
2015-02-04T18:31:51Z hiyosi quit (Quit: My Mac has gone to sleep. ZZZzzz…)
2015-02-04T18:31:55Z Posterdati: I used it
2015-02-04T18:32:07Z intinig joined #lisp
2015-02-04T18:32:20Z Posterdati: is the postgres user one
2015-02-04T18:32:21Z Xach: Does that password work in psql from the command line?
2015-02-04T18:32:29Z Posterdati: yes
2015-02-04T18:32:48Z Aerathis joined #lisp
2015-02-04T18:32:51Z oGMo: does postmodern use the same connection method?
2015-02-04T18:32:58Z Posterdati: host    all             all             127.0.0.1/32            md5
2015-02-04T18:33:38Z Posterdati: in pg_hba.conf
2015-02-04T18:33:38Z JokesOnYou77: Maybe turn off password checking?  Change MD% to trust (i think)
2015-02-04T18:33:51Z H4ns: does "psql -h 127.0.0.1" work as well?
2015-02-04T18:33:52Z Posterdati: ok yes, I found it, thanks
2015-02-04T18:34:40Z Posterdati: H4ns: no
2015-02-04T18:34:44Z H4ns: see
2015-02-04T18:34:59Z Tristam joined #lisp
2015-02-04T18:35:01Z Posterdati: but for another db worked
2015-02-04T18:35:34Z H4ns: i'd use "trust" anyway
2015-02-04T18:35:35Z oGMo: thus pgsql probably doesn't allow postgres logins except over the domain socket or whatnot by default
2015-02-04T18:36:03Z Harag quit (Ping timeout: 252 seconds)
2015-02-04T18:36:03Z nikki93 quit (Ping timeout: 264 seconds)
2015-02-04T18:36:16Z Posterdati: H4ns: worked with trust
2015-02-04T18:36:21Z Posterdati: thanks people
2015-02-04T18:36:36Z BlueRavenGT joined #lisp
2015-02-04T18:36:55Z beach left #lisp
2015-02-04T18:37:00Z oGMo: isn't that also unencrypted? don't you need hostssl or something (does postmodern support that?)
2015-02-04T18:37:26Z devll joined #lisp
2015-02-04T18:37:28Z enitiz quit (Ping timeout: 245 seconds)
2015-02-04T18:37:30Z nikki93 joined #lisp
2015-02-04T18:37:37Z JokesOnYou77: Awesome! Glad we could help.
2015-02-04T18:37:43Z Posterdati: it works with postmodern too!
2015-02-04T18:38:20Z JokesOnYou77: And I'm not sure if it changes encryption, it may.  But it's primary function is only to set authentication.
2015-02-04T18:39:39Z Posterdati: now I need to undestand if a specific database already exists and if not create it :)
2015-02-04T18:40:20Z Tristam quit (Max SendQ exceeded)
2015-02-04T18:40:23Z intinig quit (Ping timeout: 240 seconds)
2015-02-04T18:40:42Z JokesOnYou77: I would caution though that you don't really want the authentication set to trust on a production system.  If this is for testing, that's fine but otherwise you may want someone else to look at it
2015-02-04T18:41:29Z Posterdati: well only a program could access the db
2015-02-04T18:41:44Z Posterdati: no other kind of accesses
2015-02-04T18:41:49Z H4ns: don't worry.  trust is just fine for production systems as well.
2015-02-04T18:44:16Z JokesOnYou77: rlly? I mean, i know it's whitelisted to localhost, but I feel like you'd be better off not trusting any other process even on the host machine.
2015-02-04T18:45:18Z H4ns: JokesOnYou77: if you want containment, use virtualization.  multi-user hosts are always insecure, and trying to defend against local attacks is a waste of time.
2015-02-04T18:45:48Z JokesOnYou77: Fair enough.
2015-02-04T18:48:26Z salv01 left #lisp
2015-02-04T18:48:51Z nikki93 quit (Ping timeout: 245 seconds)
2015-02-04T18:50:05Z Bicyclidine quit (Ping timeout: 256 seconds)
2015-02-04T18:51:31Z sdemarre joined #lisp
2015-02-04T18:51:44Z Bicyclidine joined #lisp
2015-02-04T18:53:05Z intinig joined #lisp
2015-02-04T18:54:58Z davazp quit (Ping timeout: 245 seconds)
2015-02-04T18:55:22Z toors joined #lisp
2015-02-04T18:55:23Z kirin` quit (Ping timeout: 245 seconds)
2015-02-04T18:56:44Z kirin` joined #lisp
2015-02-04T18:59:29Z antonv: hi
2015-02-04T18:59:37Z antonv: could anyone suggest good imap client?
2015-02-04T18:59:52Z fragamus quit (Quit: Computer has gone to sleep.)
2015-02-04T19:00:08Z dlowe: imap client library, you mean?
2015-02-04T19:01:02Z nikki93 joined #lisp
2015-02-04T19:01:34Z fragamus joined #lisp
2015-02-04T19:02:00Z antonv: dlowe: yes
2015-02-04T19:02:22Z fragamus quit (Read error: Connection reset by peer)
2015-02-04T19:02:25Z intinig quit (Remote host closed the connection)
2015-02-04T19:02:59Z intinig joined #lisp
2015-02-04T19:03:18Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T19:03:43Z kirin` joined #lisp
2015-02-04T19:04:48Z lispyone quit (Remote host closed the connection)
2015-02-04T19:05:35Z 18WAAAU5J joined #lisp
2015-02-04T19:06:07Z TDog_ joined #lisp
2015-02-04T19:06:30Z Aerathis left #lisp
2015-02-04T19:06:54Z 18WAAAU5J: Hi! I am very happy i get ECL working in Android. Now every body will see that cl works also there. https://play.google.com/store/apps/details?id=ar.com.playnu.clrepl
2015-02-04T19:07:07Z ghard joined #lisp
2015-02-04T19:07:39Z 18WAAAU5J is now known as playnu_com_ar
2015-02-04T19:07:39Z TDog quit (Ping timeout: 256 seconds)
2015-02-04T19:08:02Z TDog_ is now known as TDog
2015-02-04T19:08:30Z gko__ quit (Quit: Connection closed for inactivity)
2015-02-04T19:08:31Z pt1 quit (Remote host closed the connection)
2015-02-04T19:09:49Z kirin` quit (Ping timeout: 245 seconds)
2015-02-04T19:10:00Z antonv: playnu_com_ar: CONGRATS!
2015-02-04T19:10:01Z dlowe: playnu_com_ar: pretty cool
2015-02-04T19:10:14Z Patzy quit (Ping timeout: 245 seconds)
2015-02-04T19:10:49Z kirin` joined #lisp
2015-02-04T19:11:05Z Patzy joined #lisp
2015-02-04T19:11:14Z playnu_com_ar: thanks (:
2015-02-04T19:11:35Z intinig quit (Ping timeout: 265 seconds)
2015-02-04T19:11:58Z antonv: playnu_com_ar: what are your plans? are you going to create some usefull app?
2015-02-04T19:12:10Z antonv: Have you connected from desktop slime to android swank yet?
2015-02-04T19:12:33Z antonv: does quicklisp work on android?
2015-02-04T19:12:54Z pt1 joined #lisp
2015-02-04T19:14:53Z playnu_com_ar: in the repo are the parches for get working swank
2015-02-04T19:15:08Z playnu_com_ar: but i am not the original creator and also i didn't test that
2015-02-04T19:15:29Z playnu_com_ar: I think the best way to do the thinks is to crosscompile from a desktop
2015-02-04T19:16:01Z playnu_com_ar: just beacause you can't have really workin dev enviroment inside a android emulator
2015-02-04T19:17:40Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T19:17:41Z antonv: playnu_com_ar: I see
2015-02-04T19:18:00Z antonv: playnu_com_ar: on unrelated topic: do you know a good imap client library for CL?
2015-02-04T19:18:03Z cluck quit (Ping timeout: 264 seconds)
2015-02-04T19:18:03Z kirin` joined #lisp
2015-02-04T19:18:30Z playnu_com_ar: nop, sorry
2015-02-04T19:18:59Z antonv: :)
2015-02-04T19:19:39Z billitch joined #lisp
2015-02-04T19:21:00Z antonv: oh, I've just successfully logged-in to my mailbox via IMAP with mel-base!
2015-02-04T19:21:04Z nikki93 quit (Ping timeout: 245 seconds)
2015-02-04T19:21:14Z antonv: it wasn't easy - no sensible error messages
2015-02-04T19:21:20Z Pyridrym quit (Read error: Connection reset by peer)
2015-02-04T19:22:59Z kirin` quit (Ping timeout: 252 seconds)
2015-02-04T19:23:02Z devll quit (Remote host closed the connection)
2015-02-04T19:23:31Z antonv: let's try to find and read some message...
2015-02-04T19:24:22Z kirin` joined #lisp
2015-02-04T19:25:17Z cluck joined #lisp
2015-02-04T19:26:17Z pegu quit (Ping timeout: 252 seconds)
2015-02-04T19:30:14Z kirin` quit (Ping timeout: 245 seconds)
2015-02-04T19:31:30Z kirin` joined #lisp
2015-02-04T19:32:59Z pt1 quit (Remote host closed the connection)
2015-02-04T19:35:15Z eazar001 quit (Quit: WeeChat 1.1.1)
2015-02-04T19:35:22Z Vutral quit (Ping timeout: 244 seconds)
2015-02-04T19:37:06Z eazar001 joined #lisp
2015-02-04T19:38:19Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T19:38:42Z kirin` joined #lisp
2015-02-04T19:40:28Z intinig joined #lisp
2015-02-04T19:40:31Z jtza8 joined #lisp
2015-02-04T19:41:57Z lispyone joined #lisp
2015-02-04T19:42:51Z intinig quit (Remote host closed the connection)
2015-02-04T19:43:20Z intinig joined #lisp
2015-02-04T19:45:02Z Pyridrym joined #lisp
2015-02-04T19:45:11Z kirin` quit (Ping timeout: 244 seconds)
2015-02-04T19:45:41Z kirin` joined #lisp
2015-02-04T19:49:17Z Vutral joined #lisp
2015-02-04T19:49:56Z intinig quit (Ping timeout: 264 seconds)
2015-02-04T19:50:17Z lispyone quit (Remote host closed the connection)
2015-02-04T19:51:23Z kirin` quit (Ping timeout: 240 seconds)
2015-02-04T19:52:44Z kirin` joined #lisp
2015-02-04T19:57:01Z yenda: after a month of lisp, I'm now officialy unable to pratice java programming, I'm now trying to C-w C-y in Netbeans and put parens everywhere it's a mess
2015-02-04T19:58:23Z kirin` quit (Ping timeout: 240 seconds)
2015-02-04T19:58:23Z jewel joined #lisp
2015-02-04T19:58:41Z a20150202 quit (Quit: Page closed)
2015-02-04T19:58:43Z antonv: yenda: rehabilitation program is needed
2015-02-04T19:59:28Z antonv: yenda: side note: why Netbean, use IntelliJ
2015-02-04T19:59:49Z kirin` joined #lisp
2015-02-04T20:00:32Z nyef: The greatest Java IDE ever is /bin/rm
2015-02-04T20:01:10Z nyef: Guaranteed to improve the quality of any Java code it is run against. d-:
2015-02-04T20:01:21Z dim: hey, that's an interesting use of my XML processing tool suite!
2015-02-04T20:02:46Z yenda: because I'm already 5 days late and I don't care I hate Java I just want to put an end to this misery
2015-02-04T20:03:50Z jtza8 quit (Remote host closed the connection)
2015-02-04T20:04:19Z burtons quit (Read error: Connection reset by peer)
2015-02-04T20:05:14Z JuanDaugherty quit (Remote host closed the connection)
2015-02-04T20:05:52Z burtons joined #lisp
2015-02-04T20:06:13Z kirin` quit (Ping timeout: 245 seconds)
2015-02-04T20:06:34Z paddymahoney quit (Remote host closed the connection)
2015-02-04T20:06:40Z playnu_com_ar quit (Remote host closed the connection)
2015-02-04T20:07:33Z JokesOnYou77: yenda, I'm not crazy about java, but I did change my keys to C-c C-x C-v :P
2015-02-04T20:07:48Z JokesOnYou77 Currently staring at Java :/
2015-02-04T20:07:49Z kirin` joined #lisp
2015-02-04T20:08:49Z yenda: well if you can't you should avoid
2015-02-04T20:08:52Z yenda: it's so braindead
2015-02-04T20:11:33Z intinig joined #lisp
2015-02-04T20:12:07Z Quadrescence joined #lisp
2015-02-04T20:12:07Z JokesOnYou77: I really don't like coding in it, but I do like the structure it imposes on projects.  When you can architect something to be object oriented that structure can be very elegant.  Ofc, when you can't it's like a nightmare :P
2015-02-04T20:12:31Z rhllor: it's good for structure
2015-02-04T20:14:28Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T20:14:51Z kirin` joined #lisp
2015-02-04T20:14:54Z yenda: it just make the braindead parts of programs longer to write, and the ones that do the magic are still often unintelligble
2015-02-04T20:15:21Z yenda: the advantages of it are clearly the jvm and the libraries
2015-02-04T20:15:47Z mishoo_ joined #lisp
2015-02-04T20:16:10Z yenda: it's like the C virus, but it's a new generation :D
2015-02-04T20:16:13Z JokesOnYou77: Well, it does have a lot of libs.
2015-02-04T20:16:16Z JokesOnYou77: lol
2015-02-04T20:16:18Z dim: then use ABCL?
2015-02-04T20:16:21Z mishoo quit (Ping timeout: 265 seconds)
2015-02-04T20:16:49Z akkad: that depends on closed source software from Oracle
2015-02-04T20:17:22Z rhllor: why not clojurescript then? :D
2015-02-04T20:17:48Z akkad: still requires java
2015-02-04T20:18:19Z rhllor: what I mean is, if you're using Java, why not CS to help it along?
2015-02-04T20:18:43Z nikki93 joined #lisp
2015-02-04T20:20:10Z rhllor: it's as close to lisp as you're gonna get
2015-02-04T20:21:44Z AntiSpamMeta quit (Read error: Connection reset by peer)
2015-02-04T20:21:49Z JokesOnYou77: I've actually found Scala to look very interesting.
2015-02-04T20:21:58Z AntiSpamMeta joined #lisp
2015-02-04T20:22:06Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T20:22:19Z pegu joined #lisp
2015-02-04T20:22:30Z kirin` joined #lisp
2015-02-04T20:22:49Z wasamasa: I've found Scala to look very complicated
2015-02-04T20:23:12Z rhllor: that's gonna be my next project, after swift
2015-02-04T20:23:28Z wasamasa: I'd be a lot more enthusiastic about clojurescript if it wouldn't depend on clojure
2015-02-04T20:23:38Z sdemarre quit (Ping timeout: 252 seconds)
2015-02-04T20:23:55Z Petit_Dejeuner joined #lisp
2015-02-04T20:24:05Z pt1 joined #lisp
2015-02-04T20:26:00Z selat quit (Quit: Lost terminal)
2015-02-04T20:26:56Z nikki93 quit (Remote host closed the connection)
2015-02-04T20:29:06Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T20:29:30Z kirin` joined #lisp
2015-02-04T20:31:52Z Quadrescence quit (Quit: This computer has gone to sleep)
2015-02-04T20:35:23Z TDog quit (Ping timeout: 240 seconds)
2015-02-04T20:36:49Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T20:37:12Z kirin` joined #lisp
2015-02-04T20:38:04Z meow joined #lisp
2015-02-04T20:38:08Z manuel__ quit (Quit: manuel__)
2015-02-04T20:38:19Z kuzy000_ quit (Ping timeout: 256 seconds)
2015-02-04T20:39:40Z axion: hi all. can someone help me remove the duplicate hyphens in the second example here? http://paste.lisp.org/display/145631
2015-02-04T20:39:55Z fantazo quit (Quit: Verlassend)
2015-02-04T20:41:58Z meow quit
2015-02-04T20:43:16Z edgar-rft joined #lisp
2015-02-04T20:43:48Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T20:44:11Z rhllor: try checking for a - between spaces. of course, that make take some complicated code, so I'd just stick with that unless you really needed it
2015-02-04T20:44:12Z Grue`: heh, I had to implement the exact same thing recently
2015-02-04T20:44:12Z kirin` joined #lisp
2015-02-04T20:44:24Z Grue`: but I obviously used ppcre
2015-02-04T20:44:48Z wasamasa: ,(downcase (replace-regexp-in-string "\\W+" "-" "This is a test! - Part 2"))
2015-02-04T20:44:57Z axion: this snippet was offered by someone in here a couple years ago when I was having a hard time...I never encountered a hyphen in the input until now
2015-02-04T20:45:17Z Grue`: https://github.com/tshatrov/codos/blob/master/src/models.lisp#L168
2015-02-04T20:45:21Z wasamasa: oh damnit, I thought I was in #emacs
2015-02-04T20:45:29Z wasamasa: don't mind me
2015-02-04T20:45:49Z intinig quit (Remote host closed the connection)
2015-02-04T20:46:11Z manuel__ joined #lisp
2015-02-04T20:46:24Z axion: thanks.
2015-02-04T20:46:42Z axion wonders if i can get what i want without ppcre for this one time use
2015-02-04T20:46:49Z scymtym_ quit (Ping timeout: 250 seconds)
2015-02-04T20:47:09Z wasamasa: axion: I'd search for slugs, or rather, slugification
2015-02-04T20:47:14Z wasamasa: axion: usually done with RE magic
2015-02-04T20:47:28Z Grue`: sure, just do a loop across and print chars into another string
2015-02-04T20:47:47Z JokesOnYou77: axion,  #\SPACE isn't alphanumeric, why not just get rid of the substitution?
2015-02-04T20:47:49Z Grue`: if there're several dashes in a row, dont print them
2015-02-04T20:48:24Z pt1 quit (Remote host closed the connection)
2015-02-04T20:48:49Z axion: ok thanks
2015-02-04T20:48:57Z pt1 joined #lisp
2015-02-04T20:49:35Z JokesOnYou77: Oh, I see, nvm
2015-02-04T20:49:49Z kirin` quit (Ping timeout: 245 seconds)
2015-02-04T20:51:08Z rhllor quit (Ping timeout: 264 seconds)
2015-02-04T20:51:11Z kirin` joined #lisp
2015-02-04T20:54:21Z nikki93 joined #lisp
2015-02-04T20:55:54Z rhllor joined #lisp
2015-02-04T20:57:45Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T20:58:09Z kirin` joined #lisp
2015-02-04T20:59:41Z JuanitoJons quit (Ping timeout: 245 seconds)
2015-02-04T21:00:28Z eudoxia quit (Quit: Leaving)
2015-02-04T21:03:10Z cdtaylor quit (Remote host closed the connection)
2015-02-04T21:04:15Z AntiSpamMeta quit (Read error: Connection reset by peer)
2015-02-04T21:04:28Z AntiSpamMeta joined #lisp
2015-02-04T21:04:36Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T21:05:00Z kirin` joined #lisp
2015-02-04T21:08:29Z pt1 quit (Remote host closed the connection)
2015-02-04T21:08:47Z resttime quit (Read error: Connection reset by peer)
2015-02-04T21:09:17Z resttime joined #lisp
2015-02-04T21:11:28Z kirin` quit (Ping timeout: 244 seconds)
2015-02-04T21:13:06Z kirin` joined #lisp
2015-02-04T21:14:39Z thatsnotjack joined #lisp
2015-02-04T21:14:52Z dead__ quit (Ping timeout: 245 seconds)
2015-02-04T21:17:27Z zeroish quit (Ping timeout: 264 seconds)
2015-02-04T21:20:02Z angavrilov quit (Remote host closed the connection)
2015-02-04T21:20:17Z kirin` quit (Ping timeout: 245 seconds)
2015-02-04T21:20:45Z TDog joined #lisp
2015-02-04T21:21:20Z kirin` joined #lisp
2015-02-04T21:25:15Z MutSbeta quit (Quit: Leaving.)
2015-02-04T21:25:23Z yeticry quit (Ping timeout: 240 seconds)
2015-02-04T21:26:52Z taspat` joined #lisp
2015-02-04T21:27:18Z mordocai joined #lisp
2015-02-04T21:28:00Z rhllor quit (Ping timeout: 246 seconds)
2015-02-04T21:28:01Z taspat quit (Ping timeout: 255 seconds)
2015-02-04T21:28:25Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T21:28:32Z Alfr quit (Quit: Leaving)
2015-02-04T21:28:34Z rhllor joined #lisp
2015-02-04T21:28:48Z kirin` joined #lisp
2015-02-04T21:35:06Z kirin` quit (Ping timeout: 245 seconds)
2015-02-04T21:35:53Z kirin` joined #lisp
2015-02-04T21:36:23Z lispyone joined #lisp
2015-02-04T21:38:17Z TDog quit (Ping timeout: 252 seconds)
2015-02-04T21:42:46Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T21:43:09Z kirin` joined #lisp
2015-02-04T21:44:27Z taspat` quit (Ping timeout: 264 seconds)
2015-02-04T21:44:46Z bgs100 joined #lisp
2015-02-04T21:50:04Z zacts: hi
2015-02-04T21:50:09Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T21:50:12Z zacts: what books cover CLOS well?
2015-02-04T21:50:24Z H4ns: minion: keene
2015-02-04T21:50:25Z minion: keene: No definition was found in the first 5 lines of http://www.cliki.net/keene
2015-02-04T21:50:33Z kirin` joined #lisp
2015-02-04T21:50:38Z H4ns: narf.  sonya keene's clos book is good.
2015-02-04T21:50:46Z zacts: ok cool
2015-02-04T21:51:09Z Zhivago: I'd also look at the scheme tinyclos implementation.
2015-02-04T21:51:11Z zacts: is the 1989 edition Ok?
2015-02-04T21:51:18Z zacts: Zhivago: ah cool
2015-02-04T21:51:18Z H4ns: zacts: yes.
2015-02-04T21:51:22Z zacts: neat
2015-02-04T21:51:23Z Zhivago: It makes a lot of the details that are obscure in clos explicit.
2015-02-04T21:51:38Z zacts: how about the art of the metaobject protocol?
2015-02-04T21:51:56Z H4ns: zacts: that covers the mop.  it is very good as well.
2015-02-04T21:51:57Z Zhivago: I'd read it in conjunction with the scheme tinyclos.
2015-02-04T21:52:19Z H4ns: zacts: keene's book covers clos only, not the mop.
2015-02-04T21:53:23Z gingerale quit (Ping timeout: 240 seconds)
2015-02-04T21:57:06Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T21:57:30Z kirin` joined #lisp
2015-02-04T21:58:06Z nell quit (Quit: WeeChat 1.2-dev)
2015-02-04T22:01:37Z Xach: CLOS is used a lot. The MOP is not directly used nearly as much.
2015-02-04T22:03:00Z kirin` quit (Ping timeout: 252 seconds)
2015-02-04T22:04:25Z spanter joined #lisp
2015-02-04T22:04:43Z lispyone quit (Remote host closed the connection)
2015-02-04T22:04:56Z kirin` joined #lisp
2015-02-04T22:05:07Z hiyosi joined #lisp
2015-02-04T22:06:55Z yrk quit (Quit: ERC (IRC client for Emacs 25.0.50.1))
2015-02-04T22:06:58Z fragamus joined #lisp
2015-02-04T22:09:04Z AntiSpamMeta quit (Quit: Bug hunting)
2015-02-04T22:09:18Z AntiSpamMeta joined #lisp
2015-02-04T22:11:42Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T22:11:57Z fmu quit (Ping timeout: 245 seconds)
2015-02-04T22:12:07Z kirin` joined #lisp
2015-02-04T22:15:48Z Lokathor joined #lisp
2015-02-04T22:17:44Z JuanitoJons joined #lisp
2015-02-04T22:18:43Z wasamasa: http://arrdem.com/2014/11/28/the_future_of_the_lispm/
2015-02-04T22:19:01Z kirin` quit (Ping timeout: 264 seconds)
2015-02-04T22:19:38Z wasamasa: you know, this approach doesn't sound that unreasonable
2015-02-04T22:19:53Z wasamasa: tl;dr: mezzano is pretty alright
2015-02-04T22:20:20Z kirin` joined #lisp
2015-02-04T22:21:16Z nikki93 quit (Remote host closed the connection)
2015-02-04T22:21:20Z Patzy quit (Remote host closed the connection)
2015-02-04T22:21:28Z Patzy joined #lisp
2015-02-04T22:21:52Z ecraven: concerning emacs: how do I correctly setup slime with use-package so that it is auto-loaded, yet (slime-setup 'fancy) is still called?
2015-02-04T22:22:11Z wasamasa: ecraven: did you already ask on #emacs?
2015-02-04T22:23:06Z wasamasa: ecraven: also, doesn't use-package try autoloading everything?
2015-02-04T22:23:28Z ecraven: wasamasa: not by default. I'm asking on #emacs now, just thought someone here might know too
2015-02-04T22:23:53Z ecraven: from my first trials, it seems not entirely straightforward to call slime-setup at the right time, before it is too late
2015-02-04T22:24:36Z wasamasa: also, isn't it sufficient to just set the slime-contribs variable to the right value?
2015-02-04T22:24:36Z ynniv quit (Quit: ynniv)
2015-02-04T22:24:47Z wasamasa: slime should take care of the rest in the next emacs session
2015-02-04T22:25:37Z kirin` quit (Ping timeout: 264 seconds)
2015-02-04T22:26:21Z profess quit (Ping timeout: 245 seconds)
2015-02-04T22:27:05Z ecraven: wasamasa: I don't know enough slime, but I'll try that! thanks
2015-02-04T22:27:06Z kirin` joined #lisp
2015-02-04T22:27:46Z wasamasa: ecraven: https://github.com/wasamasa/dotemacs/blob/master/init.org#slime
2015-02-04T22:28:17Z ecraven: wasamasa: that looks useful :) thank you!
2015-02-04T22:28:21Z wasamasa: ecraven: I only needed to use the slime-setup function to reload those when experimenting with slime configuration after launching a slime connection
2015-02-04T22:28:48Z wasamasa: but for setting up defaults, it wasn't needed at all
2015-02-04T22:29:03Z wasamasa: also, why do I have it twice in there
2015-02-04T22:29:06Z kcj joined #lisp
2015-02-04T22:29:41Z ecraven: better safe than sorry? :)
2015-02-04T22:30:02Z wasamasa: ok, fixed it
2015-02-04T22:30:11Z ecraven: wasamasa: works perfectly, thank you!
2015-02-04T22:30:38Z ecraven: just trying to reduce my emacs startup time by deferring everything possible ;)
2015-02-04T22:30:44Z ecraven: not that I start emacs that often :-/
2015-02-04T22:31:05Z wasamasa: well, the side effect of having a simpler config is nice to have
2015-02-04T22:32:21Z ehu_ joined #lisp
2015-02-04T22:33:08Z rhllor quit (Quit: rhllor)
2015-02-04T22:33:51Z Pyridrym quit (Quit: leaving)
2015-02-04T22:34:06Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T22:34:17Z Pyridrym joined #lisp
2015-02-04T22:34:23Z ehu quit (Ping timeout: 240 seconds)
2015-02-04T22:34:29Z kirin` joined #lisp
2015-02-04T22:36:36Z ehu_ quit (Ping timeout: 246 seconds)
2015-02-04T22:38:51Z EvW quit (Ping timeout: 245 seconds)
2015-02-04T22:41:02Z fragamus quit (Quit: Computer has gone to sleep.)
2015-02-04T22:41:44Z zRecursive joined #lisp
2015-02-04T22:41:53Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T22:42:16Z kirin` joined #lisp
2015-02-04T22:42:25Z nyef: "The type tag is 6 bits wide, providing 64 tag values of which 53 have been allocated and 13 remain unused, indicating that I can't count, because that comes out to 66 tag values."
2015-02-04T22:47:43Z Shinmera quit (Quit: しつれいしなければならないんです。)
2015-02-04T22:47:44Z pillton joined #lisp
2015-02-04T22:49:18Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T22:49:24Z nyef: Ahh... Found where the extra two tags came from. Two of the values were defined for use but also still on the freelist.
2015-02-04T22:49:27Z |3b|: nyef: or that some are allocated multiple times
2015-02-04T22:49:41Z kirin` joined #lisp
2015-02-04T22:50:50Z badkins quit (Read error: Connection reset by peer)
2015-02-04T22:52:05Z scymtym_ joined #lisp
2015-02-04T22:52:53Z Bicyclidine quit (Ping timeout: 240 seconds)
2015-02-04T22:56:23Z Karl_Dscc quit (Remote host closed the connection)
2015-02-04T22:57:23Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T22:57:48Z kirin` joined #lisp
2015-02-04T23:00:05Z Qudit314159 quit (Remote host closed the connection)
2015-02-04T23:00:23Z Bicyclidine joined #lisp
2015-02-04T23:00:39Z kobain quit (Ping timeout: 264 seconds)
2015-02-04T23:02:44Z kirin` quit (Ping timeout: 245 seconds)
2015-02-04T23:04:30Z kirin` joined #lisp
2015-02-04T23:04:57Z Bicyclidine quit (Ping timeout: 246 seconds)
2015-02-04T23:05:14Z manuel__ quit (Quit: manuel__)
2015-02-04T23:07:04Z Bicyclidine joined #lisp
2015-02-04T23:08:27Z mishoo_ quit (Ping timeout: 264 seconds)
2015-02-04T23:11:12Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T23:11:22Z badkins joined #lisp
2015-02-04T23:11:27Z xificurC quit (Ping timeout: 264 seconds)
2015-02-04T23:11:35Z kirin` joined #lisp
2015-02-04T23:11:37Z slyrus joined #lisp
2015-02-04T23:12:38Z kobain joined #lisp
2015-02-04T23:15:01Z mishoo_ joined #lisp
2015-02-04T23:15:16Z nikki93 joined #lisp
2015-02-04T23:17:17Z dagnachew joined #lisp
2015-02-04T23:17:28Z dagnachew quit (Client Quit)
2015-02-04T23:18:12Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T23:18:35Z cluck: ecraven: just make sure you're using autoloads instead of requires and that the big stuff is byte compiled (most everything else is just a lot of work for little gains)
2015-02-04T23:18:36Z kirin` joined #lisp
2015-02-04T23:19:53Z burtons quit (Ping timeout: 240 seconds)
2015-02-04T23:21:20Z cluck: nyef: if history has taught us anything is that one of those should be reserved to mean "hey, the next few ones are also tags", so we can extend it at will, you know «kinda like unicode»
2015-02-04T23:22:40Z nyef: Umm... No. This is supposed to be an exhaustive partition of a six-bit address space.
2015-02-04T23:23:27Z nugnuts joined #lisp
2015-02-04T23:23:49Z cluck: nyef: i'm just being a smartass, because everybody wants 127 bit tags on 64 bit platforms, right? ;)
2015-02-04T23:24:33Z kirin` quit (Ping timeout: 246 seconds)
2015-02-04T23:25:11Z thatsnotjack quit (Remote host closed the connection)
2015-02-04T23:25:50Z LiamH quit (Quit: Leaving.)
2015-02-04T23:26:15Z kirin` joined #lisp
2015-02-04T23:26:20Z nyef: For certain things they want as few tag bits as they can get away with.
2015-02-04T23:26:29Z nyef: For others, there's a bit of leeway.
2015-02-04T23:26:59Z cluck: nyef: on a more serious note, are you working on getting sbcl onto mezzano? (ie wth is sbcl-os? did i misunderstand something)
2015-02-04T23:27:17Z nyef: The latter.
2015-02-04T23:27:22Z nyef: You misunderstood something.
2015-02-04T23:27:41Z |3b|: running sbcl on mezzano would be amusing though :)
2015-02-04T23:27:57Z nyef: SBCL-OS was my attempt at writing an OS in CL, by bastardizing the SBCL build process to produce a cold-core that would start up on a carefully prepared machine.
2015-02-04T23:28:02Z eudoxia joined #lisp
2015-02-04T23:28:09Z manuel__ joined #lisp
2015-02-04T23:28:15Z nyef: This was something like nine years ago now.
2015-02-04T23:28:26Z akkad: nyef: use the cmucl build process for entropy?
2015-02-04T23:28:36Z cadadar joined #lisp
2015-02-04T23:28:40Z nyef: akkad: Umm... No.
2015-02-04T23:29:05Z jasom: is it possible to get a more precise location for a condition in sldb (without recompiling with debug set higher)
2015-02-04T23:29:25Z jasom: even an offset into the function disassembly would be useful
2015-02-04T23:30:03Z nyef: jasom: You might have to bounce to the native debugger and start poking around to get that offset.
2015-02-04T23:30:18Z vaporatorius quit (Remote host closed the connection)
2015-02-04T23:30:24Z nyef: jasom: Is this on SBCL, and if so is it an internal-error of some sort?
2015-02-04T23:30:59Z jasom: yes sbcl, it's object-not-cons
2015-02-04T23:31:22Z jasom: I actually found the error trap; is it possible to find all places that could trap to there?
2015-02-04T23:31:32Z nyef: cluck: Anyway, right now I'm working on figuring out some of the low-level parts of Mezzano so that I can try my hand at improving it.
2015-02-04T23:32:19Z nyef: jasom: Within the disassembly? If you used sb-disassem:disassemble-code-component then all of the code should be there (including the arg-parsing code), at which point the branch labels should ALSO be there, and you can search for those...
2015-02-04T23:32:33Z john_cephalopoda joined #lisp
2015-02-04T23:32:52Z john_cephalopoda: Hey I got a problem with clisp programming. I am trying to use hash tables.
2015-02-04T23:33:09Z Xach: john_cephalopoda: i want to help! but i will feel so much better if you call it "CL"
2015-02-04T23:33:28Z john_cephalopoda: I am using this to set a value in the hash table: (setf(gethash "15" ruleset)
2015-02-04T23:33:31Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T23:33:54Z kirin` joined #lisp
2015-02-04T23:33:55Z Xach: john_cephalopoda: did you make the ruleset hash table with a test of 'equal or 'equalp? if you didn't, you won't be able to use strings as keys effectively.
2015-02-04T23:34:09Z jasom: yeah, I found it now... I'm confused by the code-gen; it's calling sb-kernel::%adjoin-test which I assume to be generated from pushnew
2015-02-04T23:34:11Z john_cephalopoda: Ah, okay. I didn't specify any.
2015-02-04T23:34:27Z john_cephalopoda: The actual function was (setf(gethash "12" ruleset) "a")
2015-02-04T23:34:50Z john_cephalopoda: (gethash "12" ruleset) didn't output anything so I got confused.
2015-02-04T23:34:51Z lifenoodles joined #lisp
2015-02-04T23:35:09Z Xach: john_cephalopoda: the default test is EQL, and (eql "12" "12") is not necessarily true.
2015-02-04T23:35:24Z john_cephalopoda: Does it only check pointers?
2015-02-04T23:35:27Z zRecursive left #lisp
2015-02-04T23:35:48Z pjb: john_cephalopoda: is this a clisp question or a Common Lisp question?
2015-02-04T23:36:07Z Xach: john_cephalopoda: no. it does more than that. but what it does does not extend to strings.
2015-02-04T23:36:40Z jasom: (pushnew child-key (cdr (gethash key *parts-md5-hash*)) :test #'string=)  <-- I don't see *anything* in that line that needs to be a cons
2015-02-04T23:36:52Z pjb: john_cephalopoda: Try: (let ((key "12")) (setf (gethash key rules) "twelve") (gethash key rules))
2015-02-04T23:37:03Z john_cephalopoda: pjb: I am not sure. I am using clisp but since it implements common lisp and this seems to be a common lisp-specific function, it should be CL-related
2015-02-04T23:38:15Z nikki93 quit (Remote host closed the connection)
2015-02-04T23:39:27Z kirin` quit (Ping timeout: 245 seconds)
2015-02-04T23:39:33Z john_cephalopoda: pjb: Could I just do :test #'equal instead?
2015-02-04T23:39:46Z pjb: Of course, this is what you should do.
2015-02-04T23:39:50Z rtra quit (Ping timeout: 265 seconds)
2015-02-04T23:40:09Z john_cephalopoda: Nice. Thanks.
2015-02-04T23:40:26Z pjb: john_cephalopoda: I'm not aware of any divergence of clisp there.  However, clisp has an extension of CL, in that you can specify three other test functions to make-hash-table: EXT:FASTHASH-EQ EXT:FASTHASH-EQL EXT:FASTHASH-EQUAL
2015-02-04T23:40:31Z nell joined #lisp
2015-02-04T23:40:41Z |3b|: jasom: can't set the CDR of NIL, so you will get an error if KEY isn't found
2015-02-04T23:40:50Z pjb: NIL is not a cons.
2015-02-04T23:41:18Z rtra joined #lisp
2015-02-04T23:41:28Z john_cephalopoda: Lisp is very different from the language that are usually used for stuff, like Java and C(++).
2015-02-04T23:41:32Z kirin` joined #lisp
2015-02-04T23:41:41Z jasom: |3b|: thanks, I'm blind
2015-02-04T23:41:50Z jasom: pjb: I know nil is not a cons, I missed that it was setting the cdr
2015-02-04T23:41:50Z john_cephalopoda: But CL seems pretty logical.
2015-02-04T23:42:39Z pjb: john_cephalopoda: definitely.  Once you've understood a few basic rules, there remains almost no problem in CL.
2015-02-04T23:42:53Z jasom: I will squash this bug now.  It only happens about once a week, so I didn't want to leave the stacktrace until I had found it
2015-02-04T23:43:02Z john_cephalopoda: The only flaw of CL is, that compiling is kinda difficult.
2015-02-04T23:43:29Z H4ns: john_cephalopoda: how is compiling difficult?
2015-02-04T23:43:45Z nikki93 joined #lisp
2015-02-04T23:43:50Z |3b| finds it more difficult to not compile things
2015-02-04T23:43:50Z lispyone joined #lisp
2015-02-04T23:43:53Z john_cephalopoda: There are only a few compilers that actually compile to a standalone executable and I haven't found a way to get .o files yet.
2015-02-04T23:44:05Z oGMo: john_cephalopoda: you're confused
2015-02-04T23:44:11Z H4ns: john_cephalopoda: that is not "compiling"
2015-02-04T23:44:38Z jasom: john_cephalopoda: why do you want a .o file?
2015-02-04T23:45:25Z john_cephalopoda: So I can link it together with some other code. Not sure if call conventions match though
2015-02-04T23:45:44Z H4ns: john_cephalopoda: you can easily call c libraries from lisp code.
2015-02-04T23:45:46Z Bicyclidine: they don't, in general. ecl can, though, probably.
2015-02-04T23:45:54Z oGMo: finding a compiler that outputs a .so or .a is hard (ecl probably, ccl i think), but that's an _entirely_ different subject than compilation
2015-02-04T23:45:59Z Bicyclidine: "uses the C ABI" is a lot more specific a requirement than "compiles to native code"
2015-02-04T23:46:28Z oGMo: the commercial compilers possibly
2015-02-04T23:46:35Z jasom: john_cephalopoda: loading C .so files into a lisp image is fairly easy; doing it the otherway around, slightly harder
2015-02-04T23:46:40Z jasom: oGMo: ecl does it no problem
2015-02-04T23:47:03Z oGMo: jasom: i was under the impression
2015-02-04T23:47:17Z jasom: its actually fairly easy now too with asdf integration
2015-02-04T23:47:55Z jasom: it it permissible to modify a hash table while looping over its keys with LOOP?
2015-02-04T23:48:03Z kirin` quit (Read error: Connection reset by peer)
2015-02-04T23:48:11Z oGMo: it's also regrettable various compilers _don't_ produce ABI entries and ELF shared objects or whatnot .. shouldn't be terribly hard in theory, but
2015-02-04T23:48:26Z kirin` joined #lisp
2015-02-04T23:48:37Z Xach: jasom: you can modify the entry for the current key
2015-02-04T23:49:34Z Xach: jasom: http://l1sp.org/cl/3.6 has the scoop
2015-02-04T23:50:40Z jasom: can I modify other entries?  That's somewhat unclear
2015-02-04T23:51:02Z defaultxr joined #lisp
2015-02-04T23:51:08Z jasom: If it doesn't add/remove an entry that is
2015-02-04T23:51:57Z jasom: since it's not adding or deleting a new element
2015-02-04T23:52:06Z jasom: (what is deleting a new element anyway?)
2015-02-04T23:53:00Z mordocai quit (Ping timeout: 252 seconds)
2015-02-04T23:53:00Z ebrasca quit (Read error: Connection reset by peer)
2015-02-04T23:53:15Z cadadar quit (Quit: Leaving.)
2015-02-04T23:54:17Z ebrasca joined #lisp
2015-02-04T23:54:22Z JokesOnYou77: john_cephalopoda, what lisp implementation are you using?
2015-02-04T23:54:49Z john_cephalopoda: The one in GNU CLISP 2.49
2015-02-04T23:54:50Z JokesOnYou77 just integrated his Lisp project as an executable in a python project :P
2015-02-04T23:55:01Z kirin` quit (Ping timeout: 264 seconds)
2015-02-04T23:55:11Z john_cephalopoda: :D
2015-02-04T23:55:24Z jasom: JokesOnYou77: I once ported part of a python library to lisp because I couldn't for the life of me figure out how to deliver a python standalone
2015-02-04T23:55:42Z john_cephalopoda: :D
2015-02-04T23:55:47Z jasom: It had the fringe benefit of being about 8x faster too
2015-02-04T23:55:48Z kirin` joined #lisp
2015-02-04T23:56:16Z Bicyclidine quit (Ping timeout: 265 seconds)
2015-02-04T23:57:10Z Pyridrym quit (Quit: leaving)
2015-02-04T23:57:17Z jasom: Anyway, here's the offending code: https://github.com/skypher/weblocks/blob/8198458041975d0ce16b90d327dba96722d66cae/src/utils/html-parts.lisp#L106
2015-02-04T23:57:35Z JokesOnYou77: john_cephalopoda, Hm.... that one I'm not sure about.  I have no idea if that implementation supports building images.  Compiling in lisp is different from delivering an executable.  In lisp executables are typically referred to as images and they're essentially a copy of the state of the lisp in ram.  SBCL has a lisp-executable library
2015-02-04T23:57:57Z JokesOnYou77: and the proprietary implementations have built-ins for building images.
2015-02-04T23:58:29Z john_cephalopoda: I once built an executable with some lisp implementation. Not sure which one though.
2015-02-04T23:59:04Z JokesOnYou77: jasom, I bet it was faster.  And python isn't typically delivered as a standalone, though it can be set up to be installed locally.