00:00:19 cl onterface nibrary ? 00:00:32 cloxure? 00:00:35 ontology network? 00:01:17 there is probably important stuff to copy from clojure interfaces 00:01:29 -!- Guthur [~Guthur@cpc11-belf9-2-0-cust855.2-1.cable.virginmedia.com] has quit [Quit: Leaving] 00:01:43 clisure? 00:01:55 injure 00:04:57 injun deface clint clink cling clinic 00:05:40 though clinic should probably be reserved for a CL morphic-like interface 00:07:24 -!- Edward__ [ed@AAubervilliers-154-1-50-103.w90-3.abo.wanadoo.fr] has quit [] 00:13:31 :) 00:15:00 cleips, cl explicit ips 00:15:41 -!- Euthydemus` [~euthydemu@vaxjo3.23.cust.blixtvik.net] has quit [Quit: leaving] 00:16:10 Euthydemus [~euthydemu@vaxjo3.23.cust.blixtvik.net] has joined #lisp 00:16:12 pierrep [~ppasteau@dyn-240-024.wireless.concordia.ca] has joined #lisp 00:16:19 -!- pierrep [~ppasteau@dyn-240-024.wireless.concordia.ca] has quit [Remote host closed the connection] 00:16:53 -!- superflit [~superflit@140.226.49.160] has quit [Quit: superflit] 00:20:43 -!- dan` [~user@c-68-32-54-29.hsd1.nj.comcast.net] has quit [Ping timeout: 240 seconds] 00:21:23 -!- Athas [~athas@82.211.209.226] has quit [Remote host closed the connection] 00:24:38 -!- devinus [~devinus@65.107.181.222.ptr.us.xo.net] has quit [Ping timeout: 264 seconds] 00:28:21 -!- spcshpop` is now known as spcshpopr8tr 00:29:21 ignas [~ignas@ctv-79-132-160-221.vinita.lt] has joined #lisp 00:30:44 lil is my favorite so far 00:30:49 Edward_ [~ed@AAubervilliers-154-1-50-103.w90-3.abo.wanadoo.fr] has joined #lisp 00:33:34 i like lil as well, short package names make me happy, better grab them and get them into quicklisp before they are all gone. 00:33:59 what is lil? 00:34:17 -!- phf [~user@38.104.111.94] has quit [Ping timeout: 255 seconds] 00:34:32 joe4: Lisp Interface Library. scroll up ^ 00:35:16 -!- jeti [~user@p54B4795D.dip.t-dialin.net] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 00:35:22 -!- ikki [~ikki@201.122.132.181] has left #lisp 00:36:07 drewc: so you open a project on cl.net, import the code into a group-shared git repo, and create a page on cliki that includes a list of implemented datastructures? 00:37:20 Fare: sounds about right. 00:37:24 yay 00:37:49 you da man 00:38:11 *Fare* wonders about a hash-consing version of his fmim 00:38:24 and what a "generic" hash-consing library could be like 00:38:33 well, i'm going to start using it heavily, and fare-utils comes with a lot of baggage :P. Might as well get others interested in the idea 00:38:50 of course, has to be parameterized by the equality/hashing interface of the map target 00:39:21 are you going to use alexandria where I was using macros from fare-utils? 00:40:10 most likely yeah, i don't intend to depend on fare-utils at any rate. 00:41:54 ok 00:42:05 wait, I need to commit some crap I have in fare-util 00:42:18 including a basic version of updatef 00:42:28 my pure answer to setf 00:42:42 except the current version doesn't compose well enough. 00:42:43 redline6561 [~redline@c-66-56-55-169.hsd1.ga.comcast.net] has joined #lisp 00:43:25 -!- skalawag [~user@c75-111-102-202.amrlcmta01.tx.dh.suddenlink.net] has quit [Ping timeout: 245 seconds] 00:44:09 -!- katesmith [~katesmith@unaffiliated/costume] has quit [Read error: Connection reset by peer] 00:44:32 a *real* updatef should take three+ arguments: a "zipper" interface, a "state" object, and one (or many) values to purely insert at the place(s) indicated by the zipper into the structure. 00:44:54 katesmith [~katesmith@unaffiliated/costume] has joined #lisp 00:45:08 or maybe four. An interface, an object, a key, and values. 00:45:34 (splitting the key from the interface) 00:46:00 well, i'm going to watch the hockey game and get a bite to eat. 00:47:03 and, i might be persuaded to have a pint or two. 00:47:34 *Fare* sends a beer to drewc using Beer-over-IP 00:47:57 oops, reminds me I have to reactivate my paypal to be able to pay you... 00:48:07 will do after foo 00:48:07 d 00:49:28 jga [~gajon@189.135.75.156] has joined #lisp 00:49:30 there's something onamonapiac about that acronym .. BOIP ... that's what a new yoiker does after a good pint. 00:49:53 onomatopoeic? 00:50:06 sounds like what it means. 00:50:20 -!- schell [~schellsci@70-36-199-232.dsl.static.sonic.net] has quit [Quit: schell] 00:51:21 *drewc* didn't look up the spelling, and now does, and immediately regrets it. 00:51:25 devinus [~devinus@cpe-66-68-82-202.austin.res.rr.com] has joined #lisp 00:51:34 seems there is little consensus on how that word is spelt, 00:51:44 onomato-poetic 00:52:02 I'm sure your rendition is a novel interpretation that should be added to the list! 00:52:11 works for me, guiness time! 00:52:12 -!- devinus [~devinus@cpe-66-68-82-202.austin.res.rr.com] has quit [Remote host closed the connection] 00:52:36 devinus [~devinus@cpe-66-68-82-202.austin.res.rr.com] has joined #lisp 00:56:23 -!- rdd [~user@c83-250-48-164.bredband.comhem.se] has quit [Ping timeout: 245 seconds] 00:57:04 -!- araujo [~araujo@gentoo/developer/araujo] has quit [Ping timeout: 265 seconds] 00:58:08 -!- ignas [~ignas@ctv-79-132-160-221.vinita.lt] has quit [Ping timeout: 255 seconds] 00:58:13 -!- gravicappa [~gravicapp@ppp85-141-167-181.pppoe.mtu-net.ru] has quit [Ping timeout: 265 seconds] 00:59:22 davertron [~david@c-24-218-166-166.hsd1.vt.comcast.net] has joined #lisp 00:59:29 -!- Modius__ [~Modius@cpe-70-123-158-125.austin.res.rr.com] has quit [Quit: I'm big in Japan] 00:59:54 Modius [~Modius@cpe-70-123-158-125.austin.res.rr.com] has joined #lisp 01:01:48 does anyone have experience with quicklisp and lispbuilder-sdl-ttf? 01:02:21 I'm trying to run the simple example on http://lispbuilder.sourceforge.net/lispbuilder-sdl-ttf.html#examplesimple but I get "Symbol "INITIALISE-DEFAULT-FONT" not found in the LISPBUILDER-SDL-TTF package." 01:03:00 not sure if it's an issue with loading lispbuilder-sdl-ttf using quicklisp or what 01:04:46 -!- devinus [~devinus@cpe-66-68-82-202.austin.res.rr.com] has quit [Remote host closed the connection] 01:05:30 -!- housel` is now known as housel 01:06:59 percii [~percii@dxb-as71589.alshamil.net.ae] has joined #lisp 01:07:42 davertron: try an (apropos "initialise-default-font") maybe? 01:07:51 dto [~dto@pool-96-252-62-25.bstnma.fios.verizon.net] has joined #lisp 01:08:00 in my image it's in sdl, not sdl-ttf. 01:08:27 maybe the tutorial is out of date? 01:08:32 -!- tcr [~tcr@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has quit [Quit: Leaving.] 01:09:29 gigamonkey [~user@adsl-99-17-207-80.dsl.pltn13.sbcglobal.net] has joined #lisp 01:10:15 How portable is the C code generated by ECL. I.e. can I generate C code and then distribute that to people to build their own executable? 01:10:46 -!- nyef [~nyef@pool-64-222-164-95.man.east.myfairpoint.net] has quit [Quit: G'night all.] 01:11:08 gigamonkey: theoretically, possibly dependant on your Lisp code (i.e. how much platform-specific stuff is done at lisp-compilation time as opposed to c-compilation) 01:11:59 p_l|home: what do you mean by platform-specific stuff done at lisp-compilation time? 01:12:09 sellout [~greg@c-24-61-13-161.hsd1.ma.comcast.net] has joined #lisp 01:13:02 gigamonkey: #+ #- and macros that depend on OS 01:13:19 I see. 01:13:40 araujo [~araujo@gentoo/developer/araujo] has joined #lisp 01:13:49 So assuming my Lisp code would CL:READ and macro expand the same on every platform, the C code should be portable? 01:14:13 _nix00 [~Adium@61.172.241.100] has joined #lisp 01:14:35 -!- SegFault1X [~SegFaultA@c-98-234-1-162.hsd1.ca.comcast.net] has quit [Quit: leaving] 01:14:45 lnostdal-laptop [~lnostdal@167.80-203-136.nextgentel.com] has joined #lisp 01:14:48 how do I clean quicklisp's software folder? 01:14:51 -!- _nix00 [~Adium@61.172.241.100] has left #lisp 01:15:00 SegFaultAX [~SegFaultA@c-98-234-1-162.hsd1.ca.comcast.net] has joined #lisp 01:15:06 syntard: you can delete it. 01:15:24 xach: and then it'll re-extract in there? 01:15:30 syntard: yes. 01:15:37 gigamonkey: according to the documentation, the C code is supposed to be completely portable 01:15:41 I've never tested this. 01:17:18 -!- dfkjjkfd [~paulh@238-9-ftth.onsnetstudenten.nl] has quit [Quit: Lost terminal] 01:18:21 giga: ECL generates C code for particular platforms -- mainly posix vs. windows. 01:18:36 testing that the generated C code is identical on all platforms would be a cool addition to the test system, provided it's a valid assumption. 01:18:42 giga: In terms of posix systems, I believe that it is fairly portable. 01:20:31 sdsds [~sdsds@dyn-16.sub2.cmts01.cable.TORON10.iasl.com] has joined #lisp 01:20:48 It's scheme, not CL, but gambit generates 100% portable code 01:21:02 (the runtime, of course, being ported, not portable) 01:21:27 i.e. all the portability layer is in macros expanded by cpp, not gambit. 01:21:29 Zhivago: can you cross compile? I.e. generate Windows C code on a non-Windows system. 01:21:52 not sure if it's an issue with loading lispbuilder-sdl-ttf using quicklisp or whaXach: when I run (apropos "initialise-default-font") I get "LISPBUILDER-SDL:INITIALISE-DEFAULT-FONT (fbound)" 01:22:00 gigamonkey: I believe so, but I have not tried. 01:22:18 davertron: that suggests to me that the tutorial might be out of date. 01:22:47 Xach: so do you think it's still showing up under the sdl package or is it under lispbuilder-sdl? 01:23:00 gigamonkey: afaik it should be possible as long as there is nothing OS-specific or when it can check at runtime 01:23:02 -!- kaemo [~kaemo@d38-66.icpnet.pl] has quit [Ping timeout: 255 seconds] 01:23:06 Xach: my lisp package-fu is very weak 01:23:09 -!- Fare [~Fare@64.119.159.126] has quit [Quit: Leaving] 01:23:12 gigamonkey: the code so generated would include necessary C runtime 01:23:23 MetalDust [~metaldust@75-32-203-61.lightspeed.ftwotx.sbcglobal.net] has joined #lisp 01:23:23 *hefner* regrets having not re-established a Windows build environment for ECL apps 01:23:48 davertron: one is a nickname for the other 01:25:33 -!- billitch [~billitch@cpc3-cmbg14-2-0-cust129.5-4.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 01:28:24 it'd be cool to allow the user to specify a mapping from certain features symbols to the C preprocessor and emit #ifdefs when these appear as reader conditionals 01:28:48 b-man_ [~b-man@189.34.55.252] has joined #lisp 01:30:04 -!- HET2 [~diman@cpc1-cdif12-2-0-cust125.5-1.cable.virginmedia.com] has quit [Quit: This computer has gone to sleep] 01:30:46 Looks like it would be hard to cross-compile due to #+msvc, etc. 01:31:05 There shouldn't be any deep reason that you can't, but it doesn't seem set up suitably at the moment. 01:36:44 -!- prip [~foo@host130-134-dynamic.42-79-r.retail.telecomitalia.it] has quit [Ping timeout: 240 seconds] 01:37:58 -!- Yuuhi [benni@p5483D5EA.dip.t-dialin.net] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 01:38:01 drl [~lat@125.167.140.159] has joined #lisp 01:38:52 mbohun [~user@ppp115-156.static.internode.on.net] has joined #lisp 01:41:41 Hi everyone. 01:41:45 Ah, weak. Aside from :msvc a quick scan didn't reveal any other features which would cause problems. There are a few #+win32 and #+mingw32 but they appear to be confined to the compiler and code in the ECL DLL/so 01:46:36 That is to say that I don't believe it affects portability of C files generated from user code, and aside from the need to choose between MSVC and non-MSVC compilers, it's still possible that the C code is portable in the way I thought, non-portable bits being captured in the ECL shared library. I'm checking the manual to see what it claims. 01:48:12 How can I exit slime to be able to use save-lisp-and-die to save a core image? I've asked this before, and remember that it was real easy, but it eludes me now. 01:48:39 drl: the easiest way i know is not to start slime in the first place. 01:48:47 -!- Edward_ [~ed@AAubervilliers-154-1-50-103.w90-3.abo.wanadoo.fr] has quit [] 01:49:06 drl: i used to write a short program to load what i want before save-lisp-and-die. now i use buildapp to write that program for me. 01:49:38 The other issue is supporting delayed compilation. 01:50:29 prip [~foo@host200-196-dynamic.17-79-r.retail.telecomitalia.it] has joined #lisp 01:50:35 Zhivago: can you elaborate? 01:51:33 drl: it might suffice to do M-x slime-disconnect then call save-lisp-and-die from your *inferior-lisp* buffer 01:52:10 echo-area [~user@114.251.86.0] has joined #lisp 01:52:14 Well, it doesn't have any support, iirc, for just producing the C files and then stopping. 01:52:22 (I'm not certain whether slime-disconnect will terminate slime's various threads or not.) 01:52:26 It has support for leaving them behind after compilation, but that's slightly different. 01:56:38 hefner, thanks. That worked. 01:57:30 Xach, I installed Buildapp, but haven't quite learned how to use it. 02:02:49 HET2 [~diman@cpc1-cdif12-2-0-cust125.5-1.cable.virginmedia.com] has joined #lisp 02:04:54 Zhivago: if that's true, it's probably because of the way reading the generated code makes you cringe and wish you were using a different compiler. 02:06:36 -!- pavelludiq [~quassel@91.139.196.22] has quit [Remote host closed the connection] 02:07:30 -!- jolly1 [~amin@75.31.66.53] has left #lisp 02:08:56 -!- chemuduguntar [~user@smtp.touchcut.com] has quit [Ping timeout: 255 seconds] 02:09:54 kenjin2201 [~kenjin@61.99.46.4] has joined #lisp 02:11:22 -!- dborba [~dborba@pool-74-105-202-55.nwrknj.fios.verizon.net] has quit [Disconnected by services] 02:12:08 PCMX [~dborba@pool-74-105-202-55.nwrknj.fios.verizon.net] has joined #lisp 02:13:20 *hefner* is overdue to balance his moaning and snide remarks with a patch or two 02:13:28 BrandLeeJones [~BrandLeeJ@84.114.246.246] has joined #lisp 02:13:53 superflit [~superflit@c-24-9-162-197.hsd1.co.comcast.net] has joined #lisp 02:17:33 -!- jconrad [~jconrad@host109-153-55-121.range109-153.btcentralplus.com] has quit [Quit: jconrad] 02:19:35 -!- b-man_ [~b-man@189.34.55.252] has quit [Remote host closed the connection] 02:22:33 -!- Tristam [~Tristam@cpe-67-242-194-233.nycap.res.rr.com] has quit [Quit: Some days you are the pigeon, other days the statue.] 02:23:05 Tristam [~Tristam@cpe-67-242-194-233.nycap.res.rr.com] has joined #lisp 02:26:43 Bobrobyn [~rsmith05@bas1-guelph22-1177619843.dsl.bell.ca] has joined #lisp 02:27:06 nefo [~nefo@unaffiliated/nefo] has joined #lisp 02:31:06 -!- HET2 [~diman@cpc1-cdif12-2-0-cust125.5-1.cable.virginmedia.com] has quit [Quit: This computer has gone to sleep] 02:32:09 -!- jga [~gajon@189.135.75.156] has quit [Quit: Leaving] 02:32:18 -!- phadthai_ is now known as phadthai 02:34:26 hefnet: Actually, ECL's generated code wasn't particularly obscene, last I looked. 02:36:08 chp [3d87a5ae@gateway/web/freenode/ip.61.135.165.174] has joined #lisp 02:40:33 -!- vroufe [~omx@ppp94-29-9-58.pppoe.spdop.ru] has quit [Ping timeout: 245 seconds] 02:44:50 jolly1 [~amin@75.31.66.53] has joined #lisp 02:44:58 -!- jolly1 [~amin@75.31.66.53] has left #lisp 02:46:15 -!- udzinari [~user@209.158.broadband13.iol.cz] has quit [Ping timeout: 260 seconds] 02:48:53 vroufe [~omx@ppp94-29-9-58.pppoe.spdop.ru] has joined #lisp 02:49:49 -!- bgs100 [~ian@unaffiliated/bgs100] has quit [Quit: nighty night] 02:59:04 is Common Lisp the Language, 2nd Edition 02:59:05 by Guy L. Steele Jr. 02:59:12 the standard book on lisp 02:59:24 It refers to a pre-ansi version of CL. 02:59:29 oh, ok. 02:59:31 But it is worth reading. 02:59:42 The hyperspec is probably the most authoritative reference for ansi CL. 02:59:44 which is the latest and comparable version. 02:59:50 You're *much* better off starting with _Practical Common Lisp_ 02:59:53 There isn't one. 02:59:54 oh, the hyperspec. 03:00:02 CLTL2 is in a class of its own. :) 03:00:33 i am reading the "Practical Common Lisp", but felt that the Steele's version had more variety 03:00:50 -!- jweiss_ [~jweiss@cpe-069-134-009-048.nc.res.rr.com] has quit [Ping timeout: 260 seconds] 03:00:58 Speaking as a complete n00b...don't take my word for it 03:00:59 Steele's book was a starting point for the document that became the standard. 03:01:14 Personally I recommend CLTL2 for understanding what CL should have been. 03:01:31 Just refer to the hyperspec for how it actually turned out. 03:01:50 thanks, I will check out the hyperspec. 03:03:04 I think it was Xach (a few years back) who convinced me to buy my copy of PCL...which was more than worth the investment 03:03:10 Thanks, Xach! 03:03:43 It is a very fine book 03:04:43 I'm still pretty much a complete n00b...but it's given me a starting point 03:04:44 -!- lemoinem [~swoog@216.252.89.23] has quit [Remote host closed the connection] 03:05:13 lemoinem [~swoog@216.252.67.62] has joined #lisp 03:12:40 spradnyesh [~pradyus@nat/yahoo/x-dbzxcplftcpzsxqr] has joined #lisp 03:13:05 -!- madnificent [~madnifice@83.101.62.132] has quit [Ping timeout: 260 seconds] 03:14:01 Zhivago: I haven't examined it since before the new compiler and type inference, and always at debug level 2 or 3, otherwise the debugger is useless (no backtraces, even). Debug code contains sizable setup in each function to augment the local environment with function and local variable info. 03:18:20 -!- superflit [~superflit@c-24-9-162-197.hsd1.co.comcast.net] has quit [Ping timeout: 260 seconds] 03:18:39 yates [~yates@nc-76-3-105-199.dhcp.embarqhsd.net] has joined #lisp 03:20:20 -!- attila_lendvai [~attila_le@catv-89-132-188-47.catv.broadband.hu] has quit [Quit: Leaving.] 03:20:41 skalawag [~user@c75-111-102-202.amrlcmta01.tx.dh.suddenlink.net] has joined #lisp 03:20:55 would it be correct to say that an advantage of lisp over compiled languages like C, C++, etc. is that, due to the reader/evaluator partition, it is possible to write code that dynamically generates code? 03:21:15 possible, even encouraged=) 03:22:55 in my opinion, this one capability alone puts lisp in an entirely different (and higher) class than these other languages 03:23:50 someone (stassats?) said a few weeks back to me here that there is no silver bullet in lisp (over other languages). well, here's one! 03:25:10 yates: which lisp implementation do you use? 03:25:13 sbcl 03:25:28 how about you? 03:25:33 then you are executing compiled code, just like C and C++. 03:26:18 could i not write code that creates dynamic forms at runtime? 03:26:24 do taht in C 03:26:28 you can do that in C. 03:26:31 how? 03:26:48 embed a compiler? 03:26:49 write a program that invokes cc on dynamically-written code, and link it up. 03:27:00 what do you think SBCL does? It keeps the compiler around. 03:27:09 I think I mentioned already that I'm a complete lisp n00b 03:27:32 I spend my day-job editing and compiling c++ 03:27:43 sykopomp: ok, maybe in a Turing sense they're equivalent, but practically they're not 03:28:20 Adding a compiler to deal with C code on the fly and re-link just isn't going to happen 03:28:36 bzzbzz [~franco@modemcable240.34-83-70.mc.videotron.ca] has joined #lisp 03:28:36 yates: all it takes is a library or two, if that's what you want to do. 03:29:09 When I allow myself to spend time hacking on some high-level language, it's almost intoxicating 03:29:26 jimrthy: heh - me too 03:29:32 jimrthy: for your own sanity, it may not be best to dabble in lisp, heh 03:29:45 from assembly to lisp... talk about vertigo! 03:29:48 it is not ironic coincidence that intoxication itself provides also new ways of thinking, like vice versa. 03:30:05 -!- kenjin2201 [~kenjin@61.99.46.4] has quit [Ping timeout: 245 seconds] 03:30:06 I go back to the day-job in the morning, I'll edit some code, then get horribly frustrated when the changes don't "just work" 03:30:57 I've spent hours trying to debug some problem, then remember "Oh, yeah. I have to recompile, then sync up my VM again" 03:31:07 I probably shouldn't have admitted that in public 03:31:27 And I freely admit that dabbling in lisp is probably dangerous to my sanity 03:31:44 sykopomp: i'm not sure that's true - depends on the runtime environment you're using. 03:32:18 the linking phase is the potentially troubling one, afaict 03:32:45 @sykopomp: if you'd care to write those libraries, I'd be *thrilled* to pay for them. I might even be able to talk my employer into footing the bill 03:33:27 sykopomp: if i had answered something other than sbcl, would your response been different? 03:33:48 yates: i think the point wrt sbcl is that it is compiled all the time too. 03:33:57 I thought there was a project to write a repl for C at one point... don't know what happened to it, though 03:34:03 yates: you have to do extra work to not compile. 03:34:49 yates: For most CL implementations, my answer would have been similar. 03:34:55 Xach: are you referring to compiling during run-time? 03:34:57 All the ones I can think about do some level of compilation. 03:35:08 yates: all the compilation SBCL does is at runtime. 03:35:11 sykopomp: is scheme any different in this regard? 03:35:33 yates: I don't know. I don't get the impression that Scheme is as prone to native-compiled implementations, though. 03:35:35 yates: pretty much. the repl is more of a rcpl or something. 03:36:04 sykopomp: wouldn't you call the generation of a target elf file a non-runtime compilation? that's the alternative i was thinking of 03:37:25 is there a difference between reduce and apply? 03:37:44 joe4: Reduce repeatedly applies a two-argument function to a list of arguments. 03:37:47 there seem to be quite a lot of functions in common lisp which have a different name but the same function. 03:37:55 apply 'unrolls' the list and applies it as part of the arguments to a single function. 03:38:03 (reduce #'+ (list 2 3 4)) => 9 03:38:24 sykopomp: just tried it at the command prompt and both seemed to give the same result. 03:38:25 joe4: same result doesn't imply same function 03:38:25 beyond that, APPLY is not something you want to use for lists of possibly-arbitrary length. 03:38:34 syntard: yes, it does not. 03:38:42 yates: elf? 03:38:43 symbole_ [~user@ool-182ffe8f.dyn.optonline.net] has joined #lisp 03:38:45 since the argument limit applies to APPLY 03:38:52 Xach: I think he's talking about slad 03:38:55 Good morning everyone! 03:39:04 oh, ok.. so, reduce is the inverse of map. 03:39:14 where as apply is just a simple function. 03:39:39 almost like a foldr/foldl? 03:39:47 joe4: neither of those things seem particularly accurate to me. 03:40:00 reduce is fold, yes. 03:40:04 i have seen reduce likened to foldl 03:40:07 like fold in scheme* 03:40:12 joe4: (reduece #'+ (list 2 3 4)) works like (+ (+ 2 3) 4), (apply #'+ (list 2 3 4)) works like (+ 2 3 4), am I right? 03:40:33 chp: yes. 03:40:42 oh, ok. thanks guys. that clears it up. I should be using reduce more.. 03:41:03 reduce can also handle foldr. 03:45:40 -!- drl [~lat@125.167.140.159] has quit [Quit: Leaving] 03:45:42 xan_ [~xan@rrcs-69-193-205-118.nyc.biz.rr.com] has joined #lisp 03:45:49 -!- araujo [~araujo@gentoo/developer/araujo] has quit [Quit: Leaving] 03:46:08 -!- pizzledizzle [~pizdets@pool-96-250-215-244.nycmny.fios.verizon.net] has quit [] 03:47:13 -!- kenanb [~kenanb@94.54.235.211] has quit [Quit: Leaving.] 03:48:33 hi xach. 03:51:49 -!- SegFaultAX [~SegFaultA@c-98-234-1-162.hsd1.ca.comcast.net] has quit [Quit: Lost terminal] 03:55:04 -!- rme [~rme@pool-70-105-122-27.chi.dsl-w.verizon.net] has quit [Quit: rme] 03:56:08 -!- dys [~andreas@krlh-5f7224e1.pool.mediaWays.net] has quit [Ping timeout: 276 seconds] 03:56:27 dys [~andreas@krlh-5f72fea3.pool.mediaWays.net] has joined #lisp 03:57:02 -!- _pw_ [~user@123.112.77.61] has left #lisp 04:02:55 -!- BrandLeeJones [~BrandLeeJ@84.114.246.246] has quit [Quit: BrandLeeJones] 04:06:20 -!- vroufe [~omx@ppp94-29-9-58.pppoe.spdop.ru] has quit [Ping timeout: 245 seconds] 04:07:17 i am just trying some of the haskell functional operations (compose, map, fold, type, struct) on common lisp and I am finding them to be a piece of cake with common-lisp. I understand that haskell enforces a functional programming style and my programming is a lot better after I learnt it. No doubt about it. I am just curious to know if the concepts are easier in lisp due to dynamic typing? or, is it hype from the haskell community that portra 04:07:29 vroufe [~omx@ppp94-29-9-58.pppoe.spdop.ru] has joined #lisp 04:07:59 i know that these concepts would be hard in C.. but, they seem to be pretty straightforward in common lisp. 04:08:00 _pw_ [~user@123.112.77.61] has joined #lisp 04:08:38 -!- sdsds [~sdsds@dyn-16.sub2.cmts01.cable.TORON10.iasl.com] has quit [Quit: Leaving.] 04:09:35 i am not dissing haskell or common lisp, I am just trying to understand the reasoning behind. 04:09:37 lambda, the ultimate 04:10:15 it's not a coincidence, after all -- they share plenty. 04:10:20 Well, CL is procedural. 04:10:33 It's just that it supports a functional style of programming well. 04:10:36 Zhivago: why do you say that? 04:10:39 Because it is. 04:10:41 it can be used to do both. 04:10:50 But CL has no concept of function ... 04:10:59 What it has are procedures, just like C. 04:11:00 it is just a language and can be used for any paradigm? 04:11:08 It's important to understand this distinction. 04:11:17 so, how is that different from the functions in haskell? 04:11:39 Well, a procedure is essentially a sequence of operations over time. 04:11:52 A function is essentially a mapping from domain to range. 04:12:20 You can implement functions using procedures, but that doesn't mean that you're working with functions. 04:12:35 p_l|home: If you move forward with your lisp system, please consider a build option to disable threading and convert any thread-local variables to globals. In some environments, access to TLS is relatively expensive, requiring a function call rather than just an instruction or several inline. 04:13:35 joe: In terms of type safety, it's mainly a question of where the cost of type discipline lays. 04:13:36 -!- nowl [~nowl@c-71-233-2-216.hsd1.ma.comcast.net] has quit [Remote host closed the connection] 04:14:12 Zhivago: is the distinction just in terms of type? 04:14:26 i could write a function which checks on type? 04:14:37 joe: In CL, the cost of type discipline is placed on the caller. In haskell the cost is type discipline is placed on the callee. More or less. 04:14:42 Sure. 04:14:47 i meant, it checks its' arguments types before executing. 04:15:06 setheus [~setheus@cpe-70-116-140-134.tx.res.rr.com] has joined #lisp 04:15:26 So in CL, if you arrange your data flow right, then type problems don't occur, and you don't have to bother writing them out anywhere specific. 04:15:59 In haskell you need to describe the things that operate on the data in terms of types, and the compiler can complain if your data flow is wrong. 04:16:10 See CHECK-TYPE. 04:17:07 Zhivago: so, if i use check-type or some such functionality in each function, then would procedure = function? 04:17:17 No. 04:17:30 Typechecking has nothing to do with functions vs. procedures. 04:17:32 and if the function is written in a non-side-effects style 04:17:39 Likewise irrelevant. 04:17:52 If you write a procedure that implements a function, then it's still a procedure. 04:18:04 Xach: the linux executable file format 04:18:29 http://en.wikipedia.org/wiki/Executable_and_Linkable_Format 04:18:46 To be able to deal with it as a function you need to be able to express that it is a function in some way. 04:18:48 good night all 04:18:50 -!- yates [~yates@nc-76-3-105-199.dhcp.embarqhsd.net] has quit [Quit: rcirc on GNU Emacs 23.2.1] 04:18:58 CL has no way to express that something is a function. 04:19:25 And without that, you can't apply functional reasoning. 04:19:33 b4|hraban [~b4@a83-163-41-120.adsl.xs4all.nl] has joined #lisp 04:19:36 nel 04:19:40 whoops sorry 04:19:44 At least not at the level of the language -- the programmer can, if it wants to. 04:20:32 Zhivago: so, the next question is, how to make procedure to a function? 04:20:58 in CL, I mean. 04:21:53 You can't. 04:22:00 CL has no concept of functions. 04:22:38 You could extend CL, I guess. 04:22:58 And write a defunctionalizing compiler from funCL to CL. 04:23:42 do you know any place on where I can find more details on it? 04:23:48 On what? 04:24:56 "defunctionalizing compiler" or the distinction between function and procedure. I google'd the "programming procedure vs function" and it is coming up with some irrelevant stuff. 04:25:42 -!- setheus [~setheus@cpe-70-116-140-134.tx.res.rr.com] has quit [Quit: leaving] 04:25:55 -!- az [~az@p4FE4EBD1.dip.t-dialin.net] has quit [Ping timeout: 240 seconds] 04:27:24 hefner: I'm planning on using Plan 9 for "litmus test". It has only coroutines for threading :D 04:27:38 Procedural languages tend to make the distinction between a user defined named code block that returns a value, often called a FUNCTION, and one that does not, denoted by the work PROCEDURE. 04:27:49 that is what I could find from a link 04:27:57 but, I guess that is NOT correct. 04:28:13 joe: No. You're thinking of pascal. 04:29:04 joe: By defunctionalizing compiler, I mean a compiler that converts functions to procedures. 04:29:33 joe: Presumably after having applied whatever useful function composition, decomposition, etc, processes it likes. 04:29:56 p_l|home: notably, TLS access from shared libraries on Linux (to external __thread variables, if it matters) won't get fully optimized the way TLS access in a regular program does, or didn't when I tested it while experimenting on ECL. It sucks if you have something important there, like heap bounds, allocation pointer, or some frequently-accessed environment pointer. 04:30:43 -!- percii [~percii@dxb-as71589.alshamil.net.ae] has quit [Read error: No route to host] 04:31:34 joe: if I take my haskell code and convert it to lisp using the same compose, map or reduce operators, I do not understand how that is not functional code? I am sure I am not understanding something. 04:32:02 -!- spradnyesh [~pradyus@nat/yahoo/x-dbzxcplftcpzsxqr] has left #lisp 04:32:20 joe: What guarantees do functions make that procedures don't make? 04:32:21 there's probably more to it, but procedures allow side-effects and sequential evaluation within their bodies. 04:32:25 s/joe/Zhivago/ 04:32:50 joe: If you can't tell what procedures make the same guarantees, then you can't make use of those guarantees. 04:32:56 (adds insult to injury on x86 when GCC's global register variable extension can't be relied on to store a per-thread pointer and you're painfully short on registers anyway) 04:33:17 Zhivago: "What guarantees do functions make that procedures don't make?" - I do not understand. 04:33:27 joe: Well, that's because you don't know what functions are. 04:33:28 superflit [~superflit@c-24-9-162-197.hsd1.co.comcast.net] has joined #lisp 04:33:33 -!- sellout [~greg@c-24-61-13-161.hsd1.ma.comcast.net] has quit [Quit: sellout] 04:33:43 joe: Any function can be replaced by a dictionary. 04:33:55 az [~az@p5796C3AB.dip.t-dialin.net] has joined #lisp 04:34:02 joe: This is probably the simplest way to understand it. 04:34:20 How can you tell which procedures can be replaced by a dictionary? 04:35:00 spradnyesh [~pradyus@nat/yahoo/x-ryrdmkbnivhwfyzh] has joined #lisp 04:35:02 what do you mean by "dictionary" - something like a symbol dictionary or forth word dictionary? 04:35:20 hefner: So far, I don't see any real use for system-provided TLS 04:35:23 I mean a mapping. 04:35:26 other than FFI 04:35:28 Zhivago: sandbox them, run all possible inputs forever, see if outputs are same and they don't attempt anything funny 04:35:50 syntard: you're entering halting problem territory, I think 04:35:55 joe: e.g., dictionary[key] -> value 04:36:17 p_l|home: in the VM, or in general? :) 04:36:22 p_l|home: I'm ready to run it forever, so it's ok :) 04:36:33 joe: plus_1[2] == 3, and so on. 04:37:01 Zhivago: oh, ok. then, is it not just a matter of how the code looks or functions?? 04:37:22 How the code looks is irrelevant. How it functions is relevant. 04:37:31 for example (defun plus_1 (x) (+ 1 x)) -- is this not a function? 04:37:36 If you could replace the code with a dictionary, then it's a function. 04:37:52 joe: That's a procedure implementing a function, because CL has no _concept_ of function. 04:38:10 joe: CL can't differentiate that from (defun plus_1 (x) (incf y) (+ x 1)) 04:38:22 Zhivago: Are you talking about how prolog implements its subroutines? 04:38:29 joe: Which means that it cannot reason about plus_1 as a function. 04:38:30 pattern-match and show a result 04:38:36 joe: Do you have a severe reading problem? 04:39:00 -!- vroufe [~omx@ppp94-29-9-58.pppoe.spdop.ru] has quit [Ping timeout: 240 seconds] 04:39:43 (incf y): just because nobody is checking that this piece of code does not exist in the function, it is called a procedure? 04:40:32 joe: They're both procedures -- you just can't tell which one is equivalent to a function. 04:41:20 joe: Also, CL allows you to redefine procedures, which means that even if you looked at the code, it wouldn't tell you about the future unless you constrained it not to redefine that function. 04:44:15 p_l|home: on x86 there just aren't enough registers to go around with all the extra pointers you need to keep handy, plus I've a fascination with adding a small number of 'fast registers' to the VM, accessed through specially 2-address instructions where each src/dest pair is hardcoded with a unique opcode, of which I'd like as many as possible. 04:45:38 assume that I have a wrapper that can check or 'not' allow redefine's or incf or decf or set or such operators - but, still there is no guarantee that it is a function? Do you mean to say that plus_1 is not a function because there is nothing built in CL that can confirm that it is a function? 04:46:06 p_l|home: in my experiments with that approach I saw a huge performance gain versus indexing a stack-allocated register file, and though the icache pressure of all those thousands of extra instructions might sting, I can't see it swamping the advantage 04:46:47 joe: It's a procedure because that's how it operates -- it executes operations in a sequence. 04:47:23 joe: It procedes along a sequence of operations -- this is why it's called a procedure. 04:47:45 joe: A procedure that implements a function usually known as an algorithm. 04:48:07 joe: A function is a relationship from one set to another set. 04:48:19 joe: Can you understand the difference? 04:49:05 joe: You can say "the position of object X is a function of time T". 04:49:35 joe: That's different from having a procedure to work out the position of X by following a set of instructions, based on the value of T. 04:50:32 hefner: what about a "register file" designed so that it's always in L1d ? 04:51:05 Zhivago: yes, I understand the difference now. Thanks a lot.. I heard a Simon Peyton Jones' talk where he said that in the backend of haskell he does a lot of imperative programming though it is functional when we are writing code in it. I was trying to think about the CL stuff in the same way. 04:51:12 bah. compile to native code, it's not hard. 04:51:42 pkhuong_: you mean my byte-code stuff? 04:51:56 joe: There's nothing wrong with writing procedures in a functional style -- it gives many advantages. 04:52:22 p_l|home: right. 04:52:28 p_l|home: I imagine it'd almost always be in L1 once the VM warmed up 04:52:43 joe: Likewise, using monads to make the process of time implicit allows writing functions in a procedural style -- it can be convenient. 04:54:00 pkhuong_: well, it is an option, but I figured I'd first try pute-bytecode variant 04:54:28 Zhivago: do you think that having a funCL layer is very hard to do? 04:54:29 I'm leaving a backdoor to access native code though, something I thought up for my Dalvik native extensions 04:54:31 (I'll admit to having not yet researched prior art in making VMs go fast, aside from discussions of dispatch, threading, and the new Javascript engines, but hopefully I'm not inventing square wheels) 04:54:54 joe: The real question is -- what benefit would it bring? 04:55:22 joe: But, no -- writing a functional language in lisp isn't very hard. 04:55:47 p_l|home: I'd read mike pall's rant on bytecode interpreters. 04:57:09 p_l|home: basically, he argues that the niche for bytecode interpreters between real interpreters and (jit) compilers is vanishingly small. 04:57:24 Zhivago: just curious, do you know of anyone who has undertaken such a task or thought about it.. Everything seems to be already there but for a few missing pieces? 04:57:41 joe: You can find functional lisps around the place. 04:58:04 joe: You can think of haskell as more or less being a highly evolved one. 04:59:11 Zhivago: yes, found some. am reading up on them. Thanks a lot. Sorry for being so thick. 04:59:24 -!- spradnyesh [~pradyus@nat/yahoo/x-ryrdmkbnivhwfyzh] has left #lisp 04:59:50 spradnyesh [~pradyus@nat/yahoo/x-ryrdmkbnivhwfyzh] has joined #lisp 05:00:14 joe: That's ok. 05:00:34 _nix00 [~Adium@61.172.241.100] has joined #lisp 05:01:15 phuong: With tracing jits, I'm inclined to agree -- bytecode seems only useful for code compression for cache coherence, where the active core would otherwise be too large to fit. 05:01:41 thom_logn [~thom@pool-74-100-140-188.lsanca.fios.verizon.net] has joined #lisp 05:02:22 And with the size of cache these days, that's fairly monstrously large. 05:02:42 Although a shift toward massively multicore with small caches might change the balance. 05:03:45 pkhuong_: well, JIT *might* be added later. The thing is, bytecode interpreter gives the shortest path for the planned target without the added complexity of JIT (and having to use hw write barrier or inlined function instead of checking one bit in every pass). 05:04:03 -!- nuba [~nuba@pauleira.com] has quit [Read error: Connection reset by peer] 05:04:16 nuba [~nuba@pauleira.com] has joined #lisp 05:04:52 an AST interpreter seems much simpler, actually. 05:07:05 -!- theBlackDragon [~dragon@83.101.84.195] has quit [Read error: Operation timed out] 05:07:14 -!- _nix00 [~Adium@61.172.241.100] has left #lisp 05:07:27 -!- acieroid [~acieroid@wtf.awesom.eu] has quit [Read error: Connection reset by peer] 05:07:40 acieroid [~acieroid@wtf.awesom.eu] has joined #lisp 05:08:49 theBlackDragon [~dragon@83.101.84.195] has joined #lisp 05:09:57 pkhuong_: how are you defining bytecode interpreter, versus "real interpreter"? 05:10:26 Surely an AST interpreter is a "real interpreter," but what about a MIPS emulator? 05:11:58 hefner: the bytecode interpreter adds a non-trivial compilation step between parsing and the execution. At this point, you're already paying a lot of the costs associated with compilers, without getting anywhere near as much benefits as you could. 05:12:13 -!- lolsuper_ [~super_@unaffiliated/lolsuper-/x-9881387] has quit [Ping timeout: 245 seconds] 05:12:16 pkhuong: Yeah, but AST interpreters don't get the same kind of cache coherence. 05:12:40 pkhuong: (Which is the only point of using a byte-code interpreter that I can see) 05:13:34 Oops, I misread the argument above. 05:14:40 kushal [~kdas@fedora/kushal] has joined #lisp 05:14:58 Efthymios [~efthymios@c-98-207-146-68.hsd1.ca.comcast.net] has joined #lisp 05:15:08 good morning all 05:15:26 hello kushal 05:15:36 beach, hi 05:15:43 I think the only arguments for byte-code compilers are the same ones for microcode. 05:15:54 For years I scorned systems that interpreted bytecode instead of using native code, but finally I've softened up and envisioned a sweet spot in simplicity, portability, compactness, and stability. Sort of 'worse is better', but with the expectation that in a real system I'd combine the interpreter with a JIT and/or code translated to C/C++ for performance. 05:17:41 hefner: the argument isn't bytecode VS native code, but rather asking where bytecode lands between ast interpreters and native code compilers. 05:22:14 in this case, it's an issue of "bytecode is easier to combine with 3rd-party native code that might have issues with a native runtime that isn't libc" 05:22:24 <_3b> davertron: don't use the lispbuilder docs on sourceforge, current ones are at http://code.google.com/p/lispbuilder/ 05:23:48 pkhuong_: I still value performance sufficiently that AST interpreters can't compete, so it's close enough to that. 05:24:54 pkhuong_: can you give me the link to said rant? What I've found suggests the route that I am taking (fast bytecode interpreter + JIT) 05:28:05 -!- silenius [~silenus@cpe-76-169-35-122.socal.res.rr.com] has quit [Read error: Connection reset by peer] 05:28:24 p_l|home: There's , but ISTR a longer post on a mailing list. 05:29:41 scgilardi [~scgilardi@c-71-192-85-172.hsd1.ma.comcast.net] has joined #lisp 05:30:38 V-ille [~ville@70.42.89.16] has joined #lisp 05:31:00 but, actually, that points seems to be more against simple compilers than bytecode. 05:31:42 iaindalton [~user@host-72-174-169-102.cdc-ut.client.bresnan.net] has joined #lisp 05:32:02 comparing luajit 1 and luajit 2, the difference seems to be that luajit 1 depends on the interpreter instead of a simple (but fast) native code compiler. 05:32:17 _nix00 [~Adium@61.172.241.100] has joined #lisp 05:36:39 I think that comment supports my position 05:37:12 Vicfred [~Vicfred@201.102.80.138] has joined #lisp 05:37:32 hefner: the interpreter is hand-coded in asm. 05:37:59 well, important platforms would get asm versions of the interpreter 05:38:07 ..as time and energy permit. 05:39:30 -!- logia_th [~nmo@83.34.227.127] has quit [Quit: return 0;] 05:39:39 I don't know... gcc seems to be getting better support for custom runtime code frobbing (e.g. asm labels). I'm not sure which would be the lowest hanging fruit, given enough ingenuity. 05:41:03 -!- _nix00 [~Adium@61.172.241.100] has left #lisp 05:41:20 but even if a simple non-optimizing native compiler came out somewhat faster than the finely tuned and admittedly more complex bytecode system, which the linked discussion depicts as unlikely, you've still traded some portability for, retargetting the codegen for each CPU, and more complex OS-specific support code 05:43:50 sorry, I've been considering systems that compile to c, and then perform some inlining at runtime. Seems like that might be somewhere on the pareto front for simplicity, portability and performance. 05:44:36 .. and in the end you're back to all the issues that trouble the native code systems - relocation vs. PIC-code, stopping for GC, reading the thread context for the GC, signal handling woes, flaky operating systems breaking your software.. 05:44:44 -!- joe4 [~joe@c-24-126-150-146.hsd1.ga.comcast.net] has quit [Ping timeout: 240 seconds] 05:45:48 PIC (I thought everything was PIC nowadays ;)? thread context? signals? Just do it all in software, and presreve a hope of embedding your product 05:45:52 pkhuong_: that sounds potentially exciting and new 05:47:56 -!- xan_ [~xan@rrcs-69-193-205-118.nyc.biz.rr.com] has quit [Ping timeout: 276 seconds] 05:47:58 CL is weird in that respect, but I think most other languages tend to go with software write barriers, etc, especially when compiling to C. 05:48:04 yeah, I guess that kind of system, making no attempt to lead the performance pack, might as well stick to low tech implementation techniques 05:48:51 I'm not convinced that going through the MMU makes as much sense nowadays; embedded targets might not have any, and my xeon really doesn't like TLB flushes. 05:49:21 I'm afraid we might be underestimating the hidden cost of playing with mprotect. 05:50:37 setheus [~setheus@cpe-70-116-140-134.tx.res.rr.com] has joined #lisp 05:51:28 -!- setheus [~setheus@cpe-70-116-140-134.tx.res.rr.com] has quit [Client Quit] 05:52:13 udzinari [~user@209.158.broadband13.iol.cz] has joined #lisp 05:52:15 there's a subtle psychological factor - the cost of mprotect is hard to measure, invisible and a little magical, whereas the software write barrier is scattered all over memory waiting to become a scapegoat for any performance problem that comes along 05:52:57 Harag [~Harag@wbs-41-208-211-15.wbs.co.za] has joined #lisp 05:53:38 hefner: right, that's how I feel also. Costs that are harder to measure are easier to ignore; sort of a lamppost effect. 05:54:03 On the other hand, Factor uses a software write barrier (I think..) and gets along just fine. 05:54:10 right. 05:54:21 Factor goes with (byte) card marking. 06:02:44 I favor compiling to C++. Maybe there's hidden dealbreakers, but I'm fascinated with making a descriptor class for pointers/immediates that hides most tagging and type checking, arithmetic with overloaded operators, with specialized descriptor classes (ImmFixnum, ImmFloat, ConsPtr, DInteger, ...) with known types that elide checks and optimize methods (like immediate float arithmetic) 06:03:16 -!- MetalDust [~metaldust@75-32-203-61.lightspeed.ftwotx.sbcglobal.net] has quit [] 06:03:34 I just installed quicklisp, but the load it adds to my init file fails with: ASDF could not load sb-posix because Not an absolute pathname #P"~/.clc/systems/". 06:03:34 What do I need to do to fix this? 06:03:36 aside from being a compiler target, you get a nice C++ interface for accessing lisp objects for free 06:03:37 yeah.. on the other hand, C++ is hell to debug, from an asm level. 06:05:04 -!- lnostdal-laptop [~lnostdal@167.80-203-136.nextgentel.com] has quit [Ping timeout: 240 seconds] 06:05:08 -!- tritchey [~tritchey@c-68-51-119-127.hsd1.in.comcast.net] has quit [Ping timeout: 245 seconds] 06:05:10 iaindalton: check the answer here http://groups.google.com/group/quicklisp/browse_thread/thread/bb045ef9ca89d741 06:06:04 joe4 [~joe@c-24-126-150-146.hsd1.ga.comcast.net] has joined #lisp 06:07:28 konr [~user@187.106.38.106] has joined #lisp 06:08:12 pkhuong_: I don't think that would happen here, it isn't like there'd be multiple inheritance, hairy complex constructors/destructors, or whatever else might get messy at the assembly level. 06:09:23 stassats [~stassats@wikipedia/stassats] has joined #lisp 06:15:00 syntard: thx 06:17:06 setheus [~setheus@cpe-70-116-140-134.tx.res.rr.com] has joined #lisp 06:17:46 pkhuong_: btw, the interpreter in my case will be also written in assembly - first a generic C version for easy porting, then assembly for x86-64, x86 and ARM (rest depends on other people, I think, cause I have only those on hand) 06:17:53 churib [~tg@dslb-088-071-150-055.pools.arcor-ip.net] has joined #lisp 06:18:19 Lycurgus [~juan@cpe-72-228-177-92.buffalo.res.rr.com] has joined #lisp 06:19:37 as for TLB flushes, OSes manually edit TLB these days, iirc (that is, avoiding full TLB flush etc) 06:19:53 Jubb [~ghost@129.21.85.144] has joined #lisp 06:19:59 most other languages depending on software write barriers are interpreted it seems 06:20:06 or use reference counting 06:20:14 I doubted the gains of rewriting the interpreter in assembly on an x86, but I guess I can look to LuaJIT for a rough idea 06:20:23 -!- pchrist [~spirit@gentoo/developer/pchrist] has quit [Read error: Connection reset by peer] 06:20:29 (or do other weird stuff) 06:20:38 LLVM 06:20:43 p_l: What problem are you trying to solve with this implementation? 06:20:54 the GC I'm looking into, VCGC, doesn't really need a write barrier other than synchronisation 06:21:27 Zhivago: interop with C/C++ code, especially the kind of code that might not like mprotect() et al, nor non-libc runtimes 06:21:47 (that is, different calling conventions etc) 06:21:49 p_l: Doesn't ECL already solve those problems? 06:22:03 Zhivago: not fully 06:22:09 *stassats* doesn't buy into fantasy-talk, waits for the real code 06:22:33 Zhivago: the idea started with hefner complaining about issues he sometimes hits on ECL, I think :) 06:23:00 pchrist [~spirit@gentoo/developer/pchrist] has joined #lisp 06:23:25 have you tried resolving this issues in ECL? 06:23:48 I have no objection to it -- I was just wondering what the point was supposed to be. 06:24:18 stassats: some things can be resolved, but I can't change the license. 06:24:36 Hmm. Mark and sweep ... 06:24:51 ramkrsna [~ramkrsna@nat/redhat/x-auveodjpvzgzcnhc] has joined #lisp 06:24:51 -!- ramkrsna [~ramkrsna@nat/redhat/x-auveodjpvzgzcnhc] has quit [Changing host] 06:24:51 ramkrsna [~ramkrsna@unaffiliated/ramkrsna] has joined #lisp 06:24:55 So you want a highly multi-threaded interpreter? 06:25:00 stassats: it would be a great starting point, if it weren't for that 06:25:30 What's the problem with the license? 06:26:59 have you spoken with the authors about changing it/dual licensing? 06:28:30 simplechat [~simplecha@unaffiliated/simplechat] has joined #lisp 06:29:09 it strongly discourages static linking which I think is inappropriate for a language runtime (and costs size/performance), it's incompatible with running on closed platforms (game consoles being the obvious example), it feels like shaky legal, still ground open to interpretation without much or any legal precedent behind it 06:29:50 Doesn't that only apply to the linkage of the ecl library? 06:29:52 don't forget the tired debate about how LGPL applies to lisp systems, where things are so tightly bound 06:30:03 Not a problem for ECL. 06:30:14 Doesn't apply to the generated code. 06:30:33 sure, I know that. 06:31:13 So basically the problem is that you can't statically link with libecl without GPLing your code? 06:31:18 hefner: does ECL has "v2 or later" clause? 06:31:37 anyway, I think it'd be better served by either the GPL plus static linking exception (classpath-style license, if I understand it right), or a non-copyleft BSD/MIT/whatever license 06:31:51 setheus_ [~setheus@cpe-70-116-140-134.tx.res.rr.com] has joined #lisp 06:32:41 -!- setheus_ [~setheus@cpe-70-116-140-134.tx.res.rr.com] has quit [Client Quit] 06:32:43 Zhivago: that's what started annoying me, and lately I'm fascinated and annoyed even more by this notion that you just can't legally run LGPL on certain machines, because there's no way to comply with it 06:32:50 hefner: GPLv3 gives linking exception by default to "language/system runtimes", not sure if it applies to static linking 06:32:50 (the same for LGPLv3) 06:33:42 ..not that I'm likely to port my silly lisp strategy game to the Playstation 3 any time soon, but it's the principle that bothers me. 06:33:45 Ok. So the issue would be to apply for a gcc exception for runtime support linkage. 06:34:01 (nevermind the principle that MS, Sony, Nintendo, and arguably Apple are bastards..) 06:34:01 That sounds like it might be doable. 06:34:13 hefner: what about GC issues you mentioned as well? 06:34:44 -!- fisxoj [~fisxoj@cpe-24-59-205-231.twcny.res.rr.com] has quit [Ping timeout: 276 seconds] 06:36:20 -!- udzinari [~user@209.158.broadband13.iol.cz] has quit [Ping timeout: 245 seconds] 06:36:45 ah, right - more practically, I don't like it being tied to the Boehm GC, using mprotect and signal handlers, yet for whatever reason not being able to use it in generational mode - a surprise I didn't notice when initially evaluating ECL. 06:38:48 GC pauses were imperceptible, animation was gorgeous and smooth, and I rejoiced. Then over time the program grew, GC pauses become noticeable (though not a problem, yet), and now I'm afraid they're tied proportionally to the size of my heap and will grow 3x-4x longer in the finished product 06:38:51 well, my idea drops hw write barriers completely, replacing them with single uncached read. I suspect uncached read is faster than doing a complete context switch etc. 06:39:27 granted, that's firmly in the camp of something I should try fixing myself rather than complaining about on IRC. 06:40:29 that is, you enable the generational GC, it becomes very, very crashy. Not sure if it finishes building itself. 06:40:29 -!- syntard [~chatzilla@204.51.92.251] has quit [Read error: Connection reset by peer] 06:40:44 -!- prip [~foo@host200-196-dynamic.17-79-r.retail.telecomitalia.it] has quit [Ping timeout: 240 seconds] 06:40:50 -!- ltriant [~ltriant@lithium.mailguard.com.au] has quit [Quit: Get MacIrssi - http://www.sysctl.co.uk/projects/macirssi/] 06:41:13 anyway, ECL is a unique and cool CL worth using, in case I've scared anyone off. 06:41:24 -!- setheus [~setheus@cpe-70-116-140-134.tx.res.rr.com] has quit [Quit: leaving] 06:41:41 -!- gigamonkey [~user@adsl-99-17-207-80.dsl.pltn13.sbcglobal.net] has quit [Ping timeout: 265 seconds] 06:41:49 syntard [~chatzilla@204.51.92.251] has joined #lisp 06:42:07 Davidbrcz [~david@212-198-92-25.rev.numericable.fr] has joined #lisp 06:42:53 -!- Vicfred [~Vicfred@201.102.80.138] has quit [Quit: Leaving] 06:43:50 -!- dto [~dto@pool-96-252-62-25.bstnma.fios.verizon.net] has quit [Remote host closed the connection] 06:47:16 -!- Modius [~Modius@cpe-70-123-158-125.austin.res.rr.com] has quit [Quit: I'm big in Japan] 06:47:18 drl [~lat@125.167.140.159] has joined #lisp 06:48:49 dnm_ [~dnm@c-68-34-57-169.hsd1.va.comcast.net] has joined #lisp 06:49:40 ..but, I do believe that a well-executed new CL with a fast bytecode interpreter, embeddable everywhere from big apps to little phones, with a liberal public-domainish license, will slowly seep to fill all the cracks and gaps where the current lisps don't fit 06:50:27 araujo [~araujo@gentoo/developer/araujo] has joined #lisp 06:51:04 -!- Davidbrcz [~david@212-198-92-25.rev.numericable.fr] has quit [Ping timeout: 240 seconds] 06:52:07 p_l: Yes, but the point is that your implementation is targetted toward very heavily multithreaded programs? 06:53:14 omg. android-like phone with a CL api instead of java? sign me up 06:53:16 prip [~foo@host103-122-dynamic.35-79-r.retail.telecomitalia.it] has joined #lisp 06:53:27 ..accumulating users, hackers, and new optional features as it goes - LLVM JIT, ECL-style compilation to C/C++, ObjC bridge, etc., and eventually take over the world 06:53:32 Then I suggest contributing to the Shen project. 06:53:53 Shen is getting a lisp kernel language implementation for the dvm. 06:54:03 Having a CL target that shouldn't be hard. 06:54:48 hefner: The only real problem with that is the CL bit. 06:54:56 kriyative [~kriyative@ip70-181-85-174.oc.oc.cox.net] has joined #lisp 06:56:37 Zhivago: not necessarily. I hope to make it conditional, with "full" build making few versions of the runtime (so you can test various variants) and when you make an executable you simply choose which version you want. So there would be unithreaded and multithreaded, portable interpreter and assembly one etc. 06:56:52 tritchey [~tritchey@c-68-51-119-127.hsd1.in.comcast.net] has joined #lisp 06:57:44 -!- abend [~user@alpha.muted.org] has quit [Ping timeout: 240 seconds] 06:58:03 with unithreaded I can drop the write barrier and use simpler means etc. 06:58:56 the craziest idea I had was to farm all I/O into async calls and do GC while waiting :P 06:59:03 Zhivago: Yeah, heh. I wouldn't actually implement CL if I were the sole user and implementer of such a system, but something slightly simplified, cleaned up in the obvious places, but not gratuitously different, and capable of hosting and interoperating with an ANSI CL once someone got around to implementing it. 07:00:52 -!- ramkrsna [~ramkrsna@unaffiliated/ramkrsna] has quit [Ping timeout: 272 seconds] 07:01:06 Well, I hope to make it as modular as possible 07:02:20 hefner: Perhaps defining a sensible subset of CL would be a good start. 07:02:36 Although it's possible that Qi/Shen actually makes better sense for games. 07:03:16 -!- konr [~user@187.106.38.106] has quit [Remote host closed the connection] 07:04:41 p_l: I think that implementations that try to be all things to all people tend to fail. 07:04:44 -!- kriyative [~kriyative@ip70-181-85-174.oc.oc.cox.net] has quit [Quit: kriyative] 07:05:04 billr [43a45d64@gateway/web/freenode/ip.67.164.93.100] has joined #lisp 07:05:19 p_l: Personally, I also think that you'd probably be much better off using a multi-process model ala erlang, with respect to gc. 07:05:43 p_l: When you have small pools without contention a lot of problems disappear. 07:06:17 I'm not trying to be all things for all people, though. Part of the GC stuff is just so I can experiment with it before trying to force it upon SBCL, etc. 07:06:21 Also you can then use message passing and scale up to multi-mchine systems. 07:06:31 -!- b4|hraban [~b4@a83-163-41-120.adsl.xs4all.nl] has quit [Quit: Leaving] 07:06:47 If the aim is to get something into SBCL, wouldn't an experimental branch make more sense? 07:07:14 I have to include threading because I can't expect all systems to be VMS or linux patched with acall(), I have to make with unithreaded for Plan9 :-) 07:07:25 SBCL currently supports two GCs simultaneously (iirc), one for large objects and one for normal objects. 07:07:38 Perhaps adding a third arena would not be so hard. 07:07:41 Zhivago: goal is to learn enough about implementing certain things, and hopefully adapting some of that to SBCL later on 07:07:41 ehu [~ehuels@ip118-64-212-87.adsl2.static.versatel.nl] has joined #lisp 07:07:45 arix [~aric@16.sub-75-203-6.myvzw.com] has joined #lisp 07:07:59 Well, if it's for educational purposes, it doesn't matter how stupid an idea it is :) 07:08:35 *hefner* supposes the ultimate world-conquering CL/borg hybrid would absorb a JVM (mapped to the local bytecode) and a hacked Clojure that would happily live alongside CL and BorgLisp, seeing each others namespaces and calling each others functions 07:08:42 daniel__ [~daniel@p5082BE1A.dip.t-dialin.net] has joined #lisp 07:09:08 I think that's part of the "single location" disease afflicting lisp design. 07:09:37 fnordus [~dnall@S01060023693bfad4.va.shawcable.net] has joined #lisp 07:09:52 If you're making something new, I suggest strongly that message passing and multiple processes be as fundamental as possible. 07:10:08 (Not necessarily os level processes) 07:10:22 single location disease? 07:10:41 Zhivago: I'd probably have to go with copying messages, then 07:10:58 (instead of Erlang-style "pass by reference") 07:11:14 I like message passing, but I also like sharing complex data structures. I'm not sure how to do that efficiently from a different address space. 07:11:20 Well, only if you don't have immutable objects. 07:11:41 -!- daniel [~daniel@p5B3278BC.dip.t-dialin.net] has quit [Ping timeout: 250 seconds] 07:11:51 hefner: You don't need separate address spaces -- just separate mutable object transitive closures. 07:12:25 You can share the address space at an os-level as long as you don't allow external references to mutable objects. 07:12:38 Zhivago: well, I don't see CL working without immutability, though I suspect it might be doable... 07:13:00 Well, CL works without immutablity for most things at the moment. 07:13:22 fancy term for a graph of objects 07:13:41 kraison [~kraison@70.90.182.149] has joined #lisp 07:13:44 You just need to add support for immutable pairs, imutable vectors, etc. 07:13:50 ramkrsna [~ramkrsna@unaffiliated/ramkrsna] has joined #lisp 07:14:03 I want to have full ANSI CL + MOP, if it could be done immutably while having all of the portable CL code still running, I'd be happy 07:14:41 p_l: That MOP can handle immutable slots. 07:15:02 You'd probably need a build-and-freeze approach, but that's ok. 07:15:10 anyway, I think the appeal of shared address spaces, aside from being genuinely useful for HPC with small numbers of cores, is just that you don't have to think about it. One less thing to design. 07:15:36 -!- arix [~aric@16.sub-75-203-6.myvzw.com] has left #lisp 07:15:41 On the other hand, you need to think about what you'd normally send in messages -- probably vectors of atoms. 07:15:58 hefner: That's the seductive trap. 07:16:19 hefner: You don't have to think about it, and by the time you do, it's too late. 07:16:54 I do see your point, yes. 07:17:00 -!- kraison [~kraison@70.90.182.149] has left #lisp 07:18:13 Zhivago: well, I can add message-passing later as long as I keep necessary infrastructure for it to make it run fast 07:18:16 *hefner* considers wandering off to learn Erlang, having obviously fallen off the train to the future 07:18:53 p_l: Once you have message passing and multiple heaps, most of your gc issues disappear, and a standard copy-compact gc should kick the crap out of anything else. 07:19:12 Well, erlang isn't, but it's a harbinger of things to come. 07:19:51 Javascript probably is a better case study. 07:20:58 Javascript went for classic shared-space multithreading 07:21:16 at least that's what I've seen 07:21:33 Then you're blind and deranged. 07:21:38 -!- prip [~foo@host103-122-dynamic.35-79-r.retail.telecomitalia.it] has quit [Ping timeout: 264 seconds] 07:21:52 Javascript uses the complete opposite. 07:22:10 edlinde [~edlinde@90-227-7-243-no15.tbcn.telia.com] has joined #lisp 07:22:16 Single-threaded with i/o not occurring within user code. 07:22:46 And the only parallelism you get is with web workers who you talk to via immutable message passing. 07:22:50 ah, now I went through the spec, it uses message-passing 07:23:05 adf [~root@188.149.62.99] has joined #lisp 07:23:57 -!- adf [~root@188.149.62.99] has quit [Client Quit] 07:23:58 unfortunately, the JS world is rather alien to me, given that a big chunk of it seems dominated by web designers or other website specialists (and I need to constantly remind myself that I shall not take any "website" jobs) 07:24:15 -!- Jubb [~ghost@129.21.85.144] has quit [Quit: Jubb] 07:24:32 otoh, they apparently have the same costs as normal threads or even processes 07:25:35 flip214 [~marek@2001:858:107:1:baac:6fff:fe6b:9183] has joined #lisp 07:25:36 -!- flip214 [~marek@2001:858:107:1:baac:6fff:fe6b:9183] has quit [Changing host] 07:25:36 flip214 [~marek@unaffiliated/flip214] has joined #lisp 07:26:13 Depends on how you implement them, but not really since contention is largely eliminated. 07:26:42 Beetny [~Beetny@ppp118-208-8-91.lns20.bne1.internode.on.net] has joined #lisp 07:26:46 Zhivago: yeah, but basically there is no guarantee that you can spam them easily, like in Erlang 07:28:02 freddie111 [~user@ppp-94-64-152-69.home.otenet.gr] has joined #lisp 07:28:52 -!- symbole_ [~user@ool-182ffe8f.dyn.optonline.net] has quit [Remote host closed the connection] 07:28:54 it's really ballooned out in the last 2-3 years 07:29:24 major layers of stuff are in js now; in mozilla for example they pack it in jars 07:29:57 Lycurgus: in mozilla it has been like that for big chunk of last decade 07:30:31 ah, I just started building it for my own purposes in the last month or so 07:30:45 trigen [~MSX@87.209.144.213] has joined #lisp 07:30:46 -!- iaindalton [~user@host-72-174-169-102.cdc-ut.client.bresnan.net] has quit [Quit: rcirc on GNU Emacs 24.0.50.1] 07:32:21 it's 4-5 million SLOC, kina awsome to see the thing work, the build and the mgt of the developers 07:32:28 *awesome 07:32:40 mcsontos [~mcsontos@nat/redhat/x-nxmjgthzicjghyjv] has joined #lisp 07:32:42 not all js of course 07:32:51 -!- Bobrobyn [~rsmith05@bas1-guelph22-1177619843.dsl.bell.ca] has quit [Quit: Leaving] 07:33:32 *p_l|home* was forced to compile Firefox for himself since 2 years ago to make it usable 07:33:32 seems like there was a lisp browser a few years back 07:34:13 i did it for money, supporting a branded version 07:34:44 prip [~foo@host28-120-dynamic.35-79-r.retail.telecomitalia.it] has joined #lisp 07:34:48 there still is 07:35:04 URL? 07:35:14 <_pw_> Does that lisp browser run on win32? 07:35:39 -!- jdz [~jdz@host84-111-dynamic.2-87-r.retail.telecomitalia.it] has quit [Ping timeout: 240 seconds] 07:35:57 -!- cmm [~cmm@bzq-79-182-202-208.red.bezeqint.net] has quit [Ping timeout: 250 seconds] 07:36:06 <_pw_> does it expose REPL to normal users? 07:36:14 _pw_: I guess you can get it to run with Xming or similar, or (maybe with some modification) LW or ACL 07:36:34 it needs CLIM to work, dunno about REPL 07:36:57 cmm [~cmm@bzq-79-182-202-208.red.bezeqint.net] has joined #lisp 07:36:59 ah; meaning mcclim? 07:37:44 well, it works with McCLIM, might need some patching for Franz or LispWorks CLIM 07:37:58 and McCLIM's win32 backend isn't really functional, afaik 07:38:38 so it's closure? 07:39:49 yep 07:39:50 i must have some downlevel links, looks like unchanged in 5 years 07:40:07 So, what bearing might the old micro vs. monolithic kernel debate have on today's enthusiasm for message passing? What was the lesson of that early skirmish? That elegant message passing designs are in fact inferior to big balls of mud (for an OS kernel, at least)? Or, just a another example of worse being better - growing faster, more aggressive, starving out the nascent competition? 07:40:13 well, it still worked ~2y ago when I tried it 07:40:38 hefner: the true lesson was "don't believe in Mach" 07:41:04 I would have also accepted "Hurd sucks." 07:41:15 kind of different hdw situation since then 07:41:27 as for the rest, there's no true monolithic kernel in use on any desktop platform 07:41:30 hefner: The winner of that fight was the "loadable kernel module design". 07:41:30 different underlying economics 07:41:53 But then, kernels aren't meant to scale. 07:42:02 Zhivago: hybrid kernels, not LKM. Loadable modules were there before microkernels 07:42:23 The_Jon_Smith [~The_Jon_S@ip24-250-13-137.ri.ri.cox.net] has joined #lisp 07:42:26 rrice [~rrice@adsl-69-221-165-176.dsl.akrnoh.ameritech.net] has joined #lisp 07:42:35 Yes, and LKM remain after microkernels. 07:42:48 I don't see much in the same of hybrid kernels, although I'm not sure what you mean by that. 07:43:08 has anyone used CommonQT with qtwebkit? 07:43:23 Essentially the scale of a kernel is whatever is required to support a process. 07:44:13 i've got regular qt working, but I can't seem to get WebView to do anything but crash miserably 07:44:35 -!- nefo [~nefo@unaffiliated/nefo] has quit [Read error: Operation timed out] 07:44:40 hefner: inside a CPU is very different than over a network 07:45:02 Zhivago: Linux 2.6 is a hybrid kernel (and 2.4 had an older variant, but didn't use it so much), NT kernel is a hybrid one since beginning, VMS arguably is the weirdo out there that doesn't exactly fit any definition, XNU is basically Mach after hammering it to run fast (which so far always resulted in a hybrid due to Mach's abysmal IPC perf.) 07:45:05 also, microkernels didn't fail at one thing - being inherently safer and cleaner designs 07:45:23 and usually more reliable 07:45:37 -!- spradnyesh [~pradyus@nat/yahoo/x-ryrdmkbnivhwfyzh] has left #lisp 07:45:41 -!- REPLeffect [~REPLeffec@69.54.115.254] has quit [Read error: Connection reset by peer] 07:45:42 now they fall flat on their face speed-wise if they're a true microkernel 07:45:43 p_l: A hybrid of what? 07:45:58 but speed isn't everything always and forever 07:46:35 QNX is a decent example 07:46:50 Zhivago: hybrid kernel - a mix of monolithic and microkernel design 07:47:04 So, which part of linux is microkernel design? 07:47:07 Adamant: also, L4 didn't fall on its face despite being a true microkernel, the same for QNX 07:47:11 I'm not sure they succeeded in making XNU fast. This brand new Mac with an i5 and 4 gigs of ram is the most sluggish machine I've used since my 486 with Win95 and 12 megs of ram. 07:47:22 spradnyesh [~pradyus@nat/yahoo/x-nnfadaghklkfacuy] has joined #lisp 07:47:24 p_l|home: well speed-wise they do suck 07:47:27 AFAIK 07:47:29 -!- daniel__ is now known as daniel 07:47:45 (I don't think I'm exaggerating, but it's tricky to remember) 07:48:30 but their other virtues are still there. there's a proved version of L4, called L4.verified. accomplishing that for a Linux-style kernel, or THE Linux kernel, would require press-ganging the world's programmers 07:48:42 Adamant: I should remind you that both QNX and L4 are used in latency-constrained environments. And Spring showed that properly done microkernel can do IPC much faster than context switch 07:48:57 ..and I see I'm a scant 500 megs away from filling up my swap and crashing the machine. :) 07:49:10 Latency control doesn't imply fast. 07:49:16 p_l|home: not familiar with Spring 07:49:26 maybe they've come up with a way 07:49:48 unless that way is "do everything in a type-safe language and avoid switching" 07:49:49 Adamant: they cut down big chunk of IPC to few cycles, afaik 07:50:03 in which case I've seen that one before 07:50:06 and no, it was capable of running POSIX apps 07:50:08 ah 07:51:55 as for microkernel-based parts in Linux - kernel threads in 2.6 are basically processes running in kernel space 07:52:58 there are also several interfaces to run various subsystems considered part of the base OS in userspace (FUSE, Character Device in Userspace thingy etc.) 07:53:00 p_l|home: thanks! I need to look up Spring 07:53:15 Adamant: it was short-lived research OS at Sun 07:53:17 oh yeah, but most OS's do that at this point 07:53:44 Adamant: and I did mention that most OSes on desktop machines are hybrid :) 07:53:50 :P 07:53:54 fair enough 07:54:22 Anyway, speaking of Closure, lichtblau had done some maintenance work on it relatively recently - perhaps a "thank you" after ripping out so much of its code to package as libraries. Insanely cool piece of software. Interesting opportunity to attempt adding Javascript support using cl-js. 07:54:58 p_l: Do linux kernel threads run in separate address spaces? 07:55:05 Adamant: and as I said, VMS denies clear-cut categorisation as any process can enter kernel mode at any time 07:55:29 -!- edlinde [~edlinde@90-227-7-243-no15.tbcn.telia.com] has quit [Quit: edlinde] 07:55:54 Zhivago: no (though it is possible, it's simply not done). It's basically the same thing that XNU does, with different interfaces. The same as DEC OSF/1 etc. 07:56:09 -!- Adamant [~Adamant@unaffiliated/adamant] has quit [Quit: Adamant] 07:56:59 p_l: So, which part of that is microkernel design? 07:57:10 it's still classified as "hybrid" 07:57:46 So, anything with threads in it is a hybrid design? 07:58:04 Doesn't sound very meaningful. 07:58:07 Zhivago: no 07:58:22 microkernel doesn't actually specify separate address space 07:58:41 So, what does microkernel mean? 07:58:43 (Mach team was drunk on MMU kool-aid) 07:59:47 Zhivago: kernel divided into multiple chunks, "processes", capable of running simultaneusly and not only in the sense of "two applications called into syscall at the same time" 08:00:43 So, where are these divisions with kernel threads? 08:01:45 REPLeffect [~REPLeffec@69.54.115.254] has joined #lisp 08:01:46 the exact level of division depends on particular system - it's not a single "category" but more of a spectrum between two extremal choices 08:02:43 So, essentially what you're saying is that adding threading to a LKM design magically made it into a hybrid microkernel design, even through these things predate microkernels? 08:02:57 setheus [~setheus@cpe-70-116-140-134.tx.res.rr.com] has joined #lisp 08:03:28 XNU for example is "hybrid cause Mach ctx switch sucks balls" i.e. they placed Mach servers into kernel mode and one address space so IPC became a function call (still with some extra costs, afaik, due to Mach's model), while some OSes might be different 08:03:48 Zhivago: mind you, I never said "hybrid microkernel". I said "hybrid kernel" 08:04:08 and if we played with tracing stuff back, then Exokernels were first etc. 08:04:20 definitions got fleshed out as the problem space was explored 08:04:29 Hun [~Hun@80.81.19.29] has joined #lisp 08:04:34 Zhivago: hybrid kernel - a mix of monolithic and microkernel design 08:05:04 I think that your terminology here is completely stupid. 08:05:15 And you've diluted microkernel to the point of meaninglessness. 08:05:25 You might want to rethink it. 08:05:40 Zhivago: I'm not the one who made the terms 08:07:39 homie` [~user@xdsl-195-14-197-168.netcologne.de] has joined #lisp 08:07:51 As wikipedia notes, the term is controversial and largely held to be marketing nonsense. 08:07:57 I recommend not using it any more. 08:08:31 Zhivago: I'd say the same about most terms in CS 08:09:12 (and there's probably a lot of "lost in translation" going on as well, as with everything that isn't clearly defined) 08:09:15 wbooze` [~user@xdsl-195-14-197-168.netcologne.de] has joined #lisp 08:09:44 -!- wbooze [~user@xdsl-87-79-233-152.netcologne.de] has quit [Ping timeout: 240 seconds] 08:09:45 p_l: Is a graph marketing nonsense? 08:09:51 HET2 [~diman@cpc1-cdif12-2-0-cust125.5-1.cable.virginmedia.com] has joined #lisp 08:10:09 Zhivago: It's maths 08:10:35 -!- homie [~user@xdsl-87-79-233-152.netcologne.de] has quit [Ping timeout: 260 seconds] 08:10:41 not a purely CS term :) 08:11:07 _danb_ [~user@124-149-32-186.dyn.iinet.net.au] has joined #lisp 08:11:14 You are one confused bastard. 08:12:22 I take offence to the term bastard, no need to accuse any 3rd parties :P 08:13:21 Given a file containing english text, how can I feed it to the substitute function to convert spaces to newlines? 08:14:17 (as for "hybrid" - My reaction to someone describing their kernel as hybrid would be "so describe to me the exact way it is a hybrid one") 08:14:18 drl: Start by reading it into a sting using with-output-to-string. 08:14:43 *string 08:15:13 drl: open the file... (make-array (file-length ...) :element-type 'character) ... call read-sequence to read the file, filling the array... now pass it to substitute 08:15:13 p_l: e.g., you understand it to be meaningless drivel and you still use the term. 08:15:50 s/e.g./i.e./ 08:16:19 well, guilty as charged, but I blame the OS books I've used in the past which used the same term 08:16:59 aerique [euqirea@xs2.xs4all.nl] has joined #lisp 08:17:05 and it might be the case of different perception of the term due to differing language/cultural background (I've hit few of those before, especially on definitions, so...) 08:17:39 drl: Did you get that? 08:17:43 beach and hefner, thanks. 08:17:56 jdz [~jdz@193.206.22.97] has joined #lisp 08:18:18 beach, yes. I'm a very slow typist. 08:18:24 let's get back to CL 08:18:56 Adamant [~Adamant@unaffiliated/adamant] has joined #lisp 08:19:45 *beach* had forgotten about file-length. 08:20:17 aurelien [~Libre@fsf/member/aurelien] has joined #lisp 08:20:36 mvilleneuve [~mvilleneu@LLagny-156-36-4-214.w80-14.abo.wanadoo.fr] has joined #lisp 08:21:33 subtleties await with multibyte-encoded characters, but one lesson at a time 08:24:58 kiuma [~kiuma@85-18-55-37.ip.fastwebnet.it] has joined #lisp 08:25:34 hefner: Oh, right, file-length doesn't take those into account does it? 08:26:43 you need to scan the whole file to get length in characters 08:27:18 beach: no, but read-sequence should return the correct number 08:28:31 *hefner* dashes out and about 08:29:28 -!- mcsontos [~mcsontos@nat/redhat/x-nxmjgthzicjghyjv] has quit [Ping timeout: 265 seconds] 08:31:05 Or you could read it piecemeal, and concatenate at the end. 08:31:50 -!- HET2 [~diman@cpc1-cdif12-2-0-cust125.5-1.cable.virginmedia.com] has quit [Quit: This computer has gone to sleep] 08:33:52 good morning 08:37:05 lolsuper_ [~super_@unaffiliated/lolsuper-/x-9881387] has joined #lisp 08:37:33 -!- aurelien [~Libre@fsf/member/aurelien] has quit [Quit: ()] 08:38:03 dfox [~dfox@ip-89-176-11-19.net.upcbroadband.cz] has joined #lisp 08:46:06 marijnjh [~user@p5799EE2F.dip.t-dialin.net] has joined #lisp 08:47:31 rdd [~rdd@c83-250-48-164.bredband.comhem.se] has joined #lisp 08:51:49 mcsontos [~mcsontos@nat/redhat/x-tmrudrlqwcymhwtr] has joined #lisp 08:53:12 lnostdal-laptop [~lnostdal@46.66.23.118.tmi.telenormobil.no] has joined #lisp 08:55:05 konr [~user@187.106.38.106] has joined #lisp 08:56:57 morphling [~stefan@gssn-5f757ffa.pool.mediaWays.net] has joined #lisp 08:59:11 mstevens [~mstevens@89.145.84.152] has joined #lisp 08:59:13 -!- mstevens [~mstevens@89.145.84.152] has quit [Changing host] 08:59:13 mstevens [~mstevens@fsf/member/pdpc.active.mstevens] has joined #lisp 08:59:49 -!- Ralith [~ralith@S010600221561996a.vc.shawcable.net] has quit [Read error: Connection reset by peer] 08:59:51 I guess I didn't get it after all. I think I need an example. Let's simplify a bit: Given a file containing english text, convert all spaces to newlines, and write resulting text to a new file. 09:00:05 -!- mcsontos [~mcsontos@nat/redhat/x-tmrudrlqwcymhwtr] has quit [Ping timeout: 240 seconds] 09:01:04 lhz [~shrekz@c-dba672d5.021-158-73746f34.cust.bredbandsbolaget.se] has joined #lisp 09:02:07 Ralith [~ralith@S010600221561996a.vc.shawcable.net] has joined #lisp 09:02:25 drl: Start by reading the file into a string: (with-open-file (stream) (let ((string (make-array (file-length stream) :element-type 'character))) (read-sequence string stream))) 09:02:43 varjag [~eugene@122.62-97-226.bkkb.no] has joined #lisp 09:02:59 drl: er, add string at the end so that it returns the string. 09:03:03 -!- SpitfireWP [~Spitfire@wikipedia/spitfire] has quit [Ping timeout: 250 seconds] 09:03:27 drl: Then use (substitute #\Newline #\Space ...) on the result. 09:03:54 http://paste.lisp.org/submit 09:04:09 uhh 09:04:14 http://paste.lisp.org/display/116535 09:04:15 pavelludiq [~quassel@91.139.196.22] has joined #lisp 09:04:18 gravicappa [~gravicapp@ppp85-141-167-181.pppoe.mtu-net.ru] has joined #lisp 09:04:32 drl: Finally use with-open-file again but this time for the output. 09:04:43 i guess i'm missing some context 09:05:10 jdz: He wanted to use substitute for some reason I don't know. 09:05:33 jdz: If that requirement goes away, your version is fine. 09:06:27 drl: reading the whole file at once is not a good idea, in general 09:07:09 drl: but using a buffer and read-sequence/write-sequence should be fine 09:07:34 SpitfireWP [~Spitfire@wikipedia/spitfire] has joined #lisp 09:09:04 hmm. clozure supports mmap via map-file-to-ivector, but that one only works read-only. would be pretty cool to have that as a portable library... 09:09:15 spiaggia, many thanks! jdz, thanks also. Actually, I don't need to use substitute. This gives me two ways to do this. 09:10:33 -!- lolsuper_ [~super_@unaffiliated/lolsuper-/x-9881387] has quit [Ping timeout: 245 seconds] 09:13:43 jconrad [~jconrad@host109-153-55-121.range109-153.btcentralplus.com] has joined #lisp 09:16:03 -!- sigjuice [~sigjuice@c-24-130-52-60.hsd1.ca.comcast.net] has quit [Ping timeout: 250 seconds] 09:19:58 SpitfireWP_ [~Spitfire@wikipedia/spitfire] has joined #lisp 09:20:41 -!- rread [~rread@c-24-130-52-79.hsd1.ca.comcast.net] has quit [Read error: Connection reset by peer] 09:22:44 -!- SpitfireWP [~Spitfire@wikipedia/spitfire] has quit [Ping timeout: 255 seconds] 09:25:18 -!- mstevens [~mstevens@fsf/member/pdpc.active.mstevens] has quit [Quit: leaving] 09:25:26 read-sequence will be faster by far 09:26:48 Krystof [~csr21@cpc1-bour2-0-0-cust414.15-1.cable.virginmedia.com] has joined #lisp 09:27:11 edlinde [~edlinde@90-227-7-243-no15.tbcn.telia.com] has joined #lisp 09:28:11 SpitfireWP__ [~Spitfire@wikipedia/spitfire] has joined #lisp 09:28:46 kaemo [~kaemo@d38-66.icpnet.pl] has joined #lisp 09:30:21 -!- edlinde [~edlinde@90-227-7-243-no15.tbcn.telia.com] has quit [Client Quit] 09:30:21 -!- Euthydemus [~euthydemu@vaxjo3.23.cust.blixtvik.net] has quit [Read error: No route to host] 09:30:40 Euthydemus [~euthydemu@vaxjo3.23.cust.blixtvik.net] has joined #lisp 09:30:58 hlavaty [~user@77-22-100-178-dynip.superkabel.de] has joined #lisp 09:31:16 edlinde [~edlinde@90-227-7-243-no15.tbcn.telia.com] has joined #lisp 09:31:20 -!- SpitfireWP_ [~Spitfire@wikipedia/spitfire] has quit [Ping timeout: 265 seconds] 09:31:31 plediii [~plediii@nat-128-42-156-192.rice.edu] has joined #lisp 09:32:10 mixo [~mshiburi@2001:4200:7000:3:213:72ff:fe87:5055] has joined #lisp 09:38:33 tcr [~tcr@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has joined #lisp 09:39:21 SpitfireWP_ [~Spitfire@wikipedia/spitfire] has joined #lisp 09:41:24 -!- lanthan_afh [~ze@p50992b91.dip0.t-ipconnect.de] has quit [Read error: Connection reset by peer] 09:42:01 lanthan_afh [~ze@p50992b91.dip0.t-ipconnect.de] has joined #lisp 09:42:26 -!- SpitfireWP__ [~Spitfire@wikipedia/spitfire] has quit [Ping timeout: 276 seconds] 09:45:03 ILU needed with sbcl ? 09:45:44 ILU? 09:45:54 International Lisper Union? 09:46:38 zomgbie [~jesus@chello212186106204.11.vie.surfer.at] has joined #lisp 09:46:59 i can not find the [m]acronym expansion but something about modules and lightweight processes 09:47:13 http://dbpubs.stanford.edu:8091/~testbed/ilu/ilu20doc/manual_toc.html and corba. 09:47:33 this is all i see when searching "lisp lightweight processes" 09:49:38 humasect: what was your question, again? 09:49:42 seems like some Lisp-related component model 09:49:48 Athas [~athas@82.211.209.226] has joined #lisp 09:50:41 jdz: hmm. i dont want to use erlang for a particular project anymore, but the other teammates are sold on lightweight processes (the rest they are sold on can be done with CL) 09:50:59 freiksenet [~freiksene@cs181130165.pp.htv.fi] has joined #lisp 09:51:19 humasect: do you want to lose Erlang itself or just change the language? 09:51:38 -!- smithzv [~smithzv@c-98-245-201-232.hsd1.co.comcast.net] has quit [Ping timeout: 264 seconds] 09:51:39 p_l|home: i am not sure if i could trust LFE quite yet.. but maybe i can =) 09:52:05 nikodemus [~nikodemus@cs181058025.pp.htv.fi] has joined #lisp 09:52:42 well, there's LFE and Reia 09:53:09 though Reia is similar to Ruby/Smalltalk rather than Lisp 09:53:12 smithzv [~smithzv@c-98-245-201-232.hsd1.co.comcast.net] has joined #lisp 09:53:29 yeah, hmm 09:53:44 carlocci [~nes@93.37.197.198] has joined #lisp 09:58:05 -!- _danb_ [~user@124-149-32-186.dyn.iinet.net.au] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 09:59:24 sigjuice [~sigjuice@c-71-198-22-233.hsd1.ca.comcast.net] has joined #lisp 10:00:36 nha [~prefect@imamac13.epfl.ch] has joined #lisp 10:01:30 Silicon_Valley_N [~Silicon_V@99-89-3-67.lightspeed.sntcca.sbcglobal.net] has joined #lisp 10:01:53 whats up all 10:02:00 helloworld? 10:02:33 Does anyone know where is a good place to start learning LISP? 10:02:48 Practical Common Lisp 10:02:58 pocket_ [~pocket_@p4252-ipbf3210hodogaya.kanagawa.ocn.ne.jp] has joined #lisp 10:06:29 rread [~rread@c-24-5-127-116.hsd1.ca.comcast.net] has joined #lisp 10:13:26 whats that? 10:13:45 lol, nooB here 10:13:55 to Lisp 10:16:01 Silicon_Valley_N: <--- PCL 10:16:58 billitch [~billitch@cpc3-cmbg14-2-0-cust129.5-4.cable.virginmedia.com] has joined #lisp 10:20:54 lol is not that good for a beginner ... you have to be able to read macros to understand that one 10:21:04 -!- morphling [~stefan@gssn-5f757ffa.pool.mediaWays.net] has quit [Remote host closed the connection] 10:21:23 flip214: Land of Lisp? 10:21:24 Do I save memory by doing "make-array" once and assigning it the content of a line from the input file, processing this line and then resetting the vector before reading in the new line? or do I just do a make-array for every line read in? 10:21:36 mcsontos [~mcsontos@nat/redhat/x-mvowowlissybvcpv] has joined #lisp 10:22:00 stassats, didn't read it yet. I answered Silicon_Valley_N: "lol, noob here" 10:22:17 edlinde: are your lines are of constant length? 10:22:17 oh, sorry ... I meant "Let over Lambda", not "Land of Lisp" 10:22:19 that's bad 10:22:26 we've got a collision of acronyms 10:22:32 flip214: Silicon_Valley_N is already confused enough, don't confuse him more 10:22:42 stassats: no, the lines will look like "a b c d" or "a b c" etc 10:22:55 very confused... 10:22:57 so just characters or numbers space seperated on each line 10:23:18 edlinde: just use read-line, care about space when you need to care 10:23:37 ? 10:24:15 stassats: from what I read I saw that I could use "read" to read in s-expressions... so I was going to format the input file for "a b c" to be (a b c) or something 10:24:43 enupten [~neptune@117.192.74.5] has joined #lisp 10:24:47 why do you want to do that? 10:24:52 The thing is that this file could have lots of lines in it and I need only one at a time, after which I can free the memory for it 10:25:10 GC would take care of your memory 10:25:13 edlinde: you don't manage memory manually in lisp 10:25:28 -!- p_l|home [~pl@pp84.internetdsl.tpnet.pl] has quit [Ping timeout: 265 seconds] 10:25:36 stassats: I would like to read in these lines one at a time, and then I have to process each char or number in this line 10:25:58 I am confused because I am pretty sure someone in this channel mentioned that there is no GC at all in C Lisp 10:26:30 iNtERrUpT [~interrupt@116.201.171.122] has joined #lisp 10:26:34 so you guys are saying I could just do a "make-array" as I loop through the lines of the file and it would work for me? 10:26:37 C Lisp makes no sense. CL is the common acronym for Common Lisp. clisp is a specific implementation of Common Lisp -- there are several 10:26:52 ok CL 10:27:10 p_l|home [~pl@pp84.internetdsl.tpnet.pl] has joined #lisp 10:27:10 edlinde: you can just call READ-LINE and it will work. no need to call MAKE-ARRAY 10:27:20 edlinde: you are confused 10:27:30 maybe he just meant to say gc is a lower level construct in lisp, rather than being in lisp itself ? 10:28:03 (if your file is massive and you run into performance issues, that's a different thing -- but get it working first, optimize only when you know you need to) 10:28:10 that was me who said that CL has no GC, and that is true, the standard doesn't describe any way to manage memory 10:28:28 but all implementation of CL have a garbage collector 10:28:38 Athas` [~athas@82.211.209.226] has joined #lisp 10:29:02 stassats: ok so in my case Clozure would do it for me? 10:29:14 yes 10:29:21 ok thats fine then 10:29:23 cool 10:29:32 -!- iNtERrUpT [~interrupt@116.201.171.122] has quit [Remote host closed the connection] 10:29:35 just with read-line... isn't it going to return a string to me? 10:29:38 things get pretty hairy when the GC doesn't work. just assume it does for now, as it usually does 10:29:42 clhs read-line 10:29:43 http://www.lispworks.com/reference/HyperSpec/Body/f_rd_lin.htm 10:29:50 edlinde: Lisp doesn't need GC, but keeping track of when to call eralis was tedious so they invented GC 10:30:14 -!- Athas [~athas@82.211.209.226] has quit [Ping timeout: 255 seconds] 10:30:55 BrianRice [~water@c-98-246-165-205.hsd1.or.comcast.net] has joined #lisp 10:31:12 ok 10:31:57 -!- TraumaPony [~TraumaPon@124-171-212-5.dyn.iinet.net.au] has quit [Ping timeout: 265 seconds] 10:32:36 TraumaPony [~TraumaPon@124-148-40-6.dyn.iinet.net.au] has joined #lisp 10:33:27 (eralis being supposedly the equivalent of free() in one of the first, pre-GC LISP implementations) 10:33:46 stassats: sorry stupid qn :) , isn't "read-line" going to return to me a string, do I then have to convert this string to a vector? 10:33:57 a string is a vector 10:33:58 clhs string 10:33:59 http://www.lispworks.com/reference/HyperSpec/Body/a_string.htm 10:34:13 Hun: ah yeah true 10:34:35 lolsuper_ [~super_@unaffiliated/lolsuper-/x-9881387] has joined #lisp 10:35:15 -!- Athas` is now known as Athas 10:35:26 is it better to store such strings in a vector or a list? esp if I know that I will be processing one character at a time? 10:35:37 just wonder what you guys think is more efficient? 10:35:48 vector 10:36:02 and every alternate value would be a space character yeah? 10:36:13 which I would have to skip while processing the vector... is that right? 10:36:48 -!- zomgbie [~jesus@chello212186106204.11.vie.surfer.at] has quit [Quit: Lost terminal] 10:36:50 try to abstract that out. parsing is easier if you do a seperate tokenizer and interpreter 10:36:53 if that's what you want to do 10:36:54 edlinde: Haskell implements strings as Lists. This became the reason why everyone who needs any serious processing of strings switched to BinaryStrings, which are arrays :) 10:37:02 -!- prip [~foo@host28-120-dynamic.35-79-r.retail.telecomitalia.it] has quit [Ping timeout: 276 seconds] 10:37:08 p_l|home: how is this relevant? 10:37:32 stassats: 10:18 < edlinde> is it better to store such strings in a vector or a list? <--- my reaction to this 10:37:49 ok its something to get started from 10:37:52 it doesn't answer edlinde's question 10:38:22 hmmm.... I guess I'm starting to get cryptic 10:38:23 stassats: I think p_l|home was referring to Haskell because I mentioned I used it earlier 10:38:25 edlinde: you seem to care too much about optimization, prematurely 10:38:28 stats thats for the link. just found it 10:38:55 stassats: I have a C background and I am new to lisp, infact I just started learning it a week ago 10:39:08 rwallace [~rwallace@89.100.128.108] has joined #lisp 10:39:16 edlinde: use straightforward code first 10:39:16 cheers to that 10:39:19 stassats: so I am used to thinking about how to alloc and dealloc my memory 10:39:37 yeah, we kinda figured 10:39:38 stassats: yeah with a GC things should be ok. 10:39:48 so, get used to not thinking about it 10:40:02 hehe ok :) 10:40:19 oh btw whats the protocol of posting questions here ? 10:40:26 get used to thinking about not leaving references to objects you don't need 10:40:38 edlinde: 10:40:52 I mean I understand that you guys are way too advanced for these newbie questions 10:40:55 stassats: that leaves a space. 10:41:10 stassats: well last night I was told off for posting "too frequently" 10:41:21 edlinde: we aren't. if you have a question, put it. try to do your homework though 10:41:24 -!- pchrist [~spirit@gentoo/developer/pchrist] has quit [Quit: leaving] 10:41:24 Hun: it's a clause separator, the space would've been 10:41:29 stassats: :) 10:41:54 edlinde: asking every trivial detail isn't good, yes 10:41:55 yeah its just that if say it was up to me now, I would have been thinking in terms of memory and using things like read instead of read-line 10:41:59 pchrist [~spirit@gentoo/developer/pchrist] has joined #lisp 10:42:47 edlinde: think in memory if you need to. usually, you don't need to 10:43:09 that's probably because you haven't read "Practical Common Lisp" yet 10:43:28 nah thats not a problem now, if there is a GC in place then I am fine... as long as I don't run out of memory for just reading in a huge file 10:43:46 so the protocol is, in fact: | 10:43:48 stassats: I have.. but only upto chapter 12 I think 10:43:59 Joreji [~thomas@67-052.eduroam.RWTH-Aachen.DE] has joined #lisp 10:44:05 -!- TraumaPony [~TraumaPon@124-148-40-6.dyn.iinet.net.au] has quit [Ping timeout: 240 seconds] 10:44:19 -!- ineiros [~itniemin@james.ics.hut.fi] has quit [Ping timeout: 240 seconds] 10:44:25 if only I had time to read the entire thing.. I am doing two other languages at the same time as part of a functional programming course 10:44:32 edlinde: if you need to parse stuff fast, Clozure also supports mmap. though you likely don't want to use that because of portability 10:44:41 for a project I opted to use Common lisp instead of Erlang or Haskell 10:44:51 PCL was written so that there will be less questions for #lisp and more time for discussions about gun control and such 10:45:03 stassats: :) 10:45:20 but it turned not as expected 10:45:37 don't forget nanowrimo :P 10:45:41 ok thanks... I will get back to trying some code 10:46:09 ineiros [~itniemin@james.ics.hut.fi] has joined #lisp 10:48:19 I was wondering, do you guys work in industry? or in academia? 10:48:41 I mean are there companies using common lisp for major products? 10:48:49 yes, there are 10:49:05 prip [~foo@host167-83-dynamic.11-87-r.retail.telecomitalia.it] has joined #lisp 10:49:38 stassats: so you work in industry? 10:50:03 no, i work for myself, though the pay is lousy 10:50:16 ah ok like a freelance contractor 10:50:25 computer networking here 10:50:28 trying to branch off 10:50:58 edlinde: yes, only i contract myself to write software for my own needs 10:51:03 _nix00 [~Adium@61.172.241.100] has joined #lisp 10:51:31 stassats: and your boss is an idiot! 10:51:50 ROFL 10:52:03 so stassats' boss is no exceptional value 10:52:06 I hate my boss 10:52:10 do you hate your boss? 10:52:31 edlinde: piano, sitegrinder3, and inspiredata are three boxed software products you can buy that are powered by CL. 10:52:36 i do, i also hate my employee for not delivering on time 10:52:43 -!- spradnyesh [~pradyus@nat/yahoo/x-nnfadaghklkfacuy] has quit [Ping timeout: 240 seconds] 10:52:55 it's a VM, but for employee relations 10:52:55 edlinde: well, maybe you personally can't buy them. 10:52:57 do yourself a favor then. slap him... 10:53:05 ok 10:53:37 the panama canal for a time was managed by a CL product 10:53:39 maybe it still is 10:53:46 :) 10:54:09 Xach: you have to give the Gulf War example 10:54:12 -!- _nix00 [~Adium@61.172.241.100] has left #lisp 10:54:24 nah its just that I am a student and so was wondering if any of you guys were still at uni or doing research? 10:55:06 Adamant: Don't know about that. 10:55:06 cl sounds like a good language to know to whip up some quick prototypes and see things working 10:55:26 -!- Joreji [~thomas@67-052.eduroam.RWTH-Aachen.DE] has quit [Ping timeout: 265 seconds] 10:55:45 edlinde: true. it's also a good language to take the quick prototypes and make them fast 10:55:57 Xach: the one where an expert system based on Lisp did enough optimization of logistics compared to a previous system that it justified the DoD's entire investment in Lisp and AI 10:56:09 with some left over 10:56:54 but it sponsored a war! how good is that? 10:57:04 absolutely nothin' 10:57:32 stassats: if it was Gulf War II I wouldn't use the example 10:58:41 but it was Gulf War I 10:59:18 "any sufficiently advanced technology is indistinguishable from war technology" 10:59:37 more like any sufficently advanced technology will be put to military use 11:00:02 and any sufficiently expensive technology will usually be put to military use first. 11:01:51 =p 11:01:53 -!- cataska [~cataska@210.64.6.233] has quit [Quit: leaving] 11:03:17 -!- billstclair [~billstcla@unaffiliated/billstclair] has quit [Quit: Ex-Chat] 11:03:18 rTypo [~rTypo@pc54.clicknet.iasi.rdsnet.ro] has joined #lisp 11:07:38 -!- lolsuper_ [~super_@unaffiliated/lolsuper-/x-9881387] has quit [Ping timeout: 245 seconds] 11:07:45 Adamant: the exact quote was that the system "saved enough money to repay all the AI research funded by DARPA till Gulf War" 11:07:55 ah 11:07:58 (well, maybe not exact, but that was the spirit of it :D) 11:08:10 yeah, that's better than going from my memory 11:08:35 of course Lisp research was AI research at that time and vice versa 11:09:13 not really, it just happened that Lisp was back then still one of the most used languages for AI 11:10:04 p_l|home: I thought most AI research until the early 80's used Lisp or Lisp-derived languages like PLANNER 11:10:18 that was what I meant 11:10:22 was planner ever implemented? 11:10:47 "good for AI" is another way to say "changes design a lot", which games and many other kinds of interactive materials can be said to be, which is good to have a language that can scale to the complexity of said application 11:11:29 stassats: I thought it was, just not fully 11:11:53 it's not a language which scales better, but programming techniques, in some languages some techniques are easier to implement 11:11:58 Adamant: there was micro-planner 11:12:23 sounds about right. insomnia strikes. 11:13:52 stassats: i dont know the vocabulary for it, but how DSLs are made naturally to follow the various stages of an application being developed 11:14:26 DSL is just another buzzword to me 11:14:54 yep, forgive me 11:16:07 lisp build a building material, each stage or layer of foundation for an application or library being a new (non-descrete, usually) layer of language which depends on previous ones. even C devs know some of this albeit subconsciously 11:17:00 p_l|home: the inventor of lisp was a AI professor, i think. so maybe that is a clue 11:17:30 like how erlang came from server devs ? 11:18:16 -!- pavelludiq [~quassel@91.139.196.22] has quit [Remote host closed the connection] 11:20:30 -!- mbohun [~user@ppp115-156.static.internode.on.net] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 11:21:42 petercoulton [~petercoul@152.105.49.241] has joined #lisp 11:23:14 skaboy [~grant@c0144.aw.cl.cam.ac.uk] has joined #lisp 11:26:02 Joreji [~thomas@85-113.eduroam.RWTH-Aachen.DE] has joined #lisp 11:26:58 hmm was trying something like ... (read-delimited-list #\Return in nil) 11:27:27 was thinking this function was more suited for my needs as it skips whitespaces .. instead of read-line 11:27:37 but the above doesn't work 11:27:56 spradnyesh [~pradyus@nat/yahoo/x-slgekbmhdxfrytgu] has joined #lisp 11:27:58 what does read-delimited-list expect ? 11:28:48 clhs read-delimited-list 11:28:48 http://www.lispworks.com/reference/HyperSpec/Body/f_rd_del.htm 11:29:09 "The consequences are undefined if char has a syntax type of whitespace[2] in the current readtable." 11:29:16 so, you can't use #\Readline 11:29:41 if you wan to read a line of space-delimited words, use read-line and then split the string 11:29:50 using split-sequence, for example 11:30:05 #\Return 11:30:08 -!- gravicappa [~gravicapp@ppp85-141-167-181.pppoe.mtu-net.ru] has quit [Ping timeout: 245 seconds] 11:30:12 Yuuhi [benni@p5483A6BB.dip.t-dialin.net] has joined #lisp 11:31:02 ok 11:31:25 thanks 11:32:56 -!- spradnyesh [~pradyus@nat/yahoo/x-slgekbmhdxfrytgu] has quit [Ping timeout: 276 seconds] 11:37:01 lc0180643 [~lc@41.35.197.69] has joined #lisp 11:38:24 Hello, if I do (defun five () (+ 2 3)) the REPL parrots back five and that seems to be ok, how ever when I try to check it using five the REPL tells me that it's not defined 11:38:28 what did I miss? 11:39:04 "I try to check it using five" what does that mean? 11:39:33 ignas [~ignas@78-60-36-123.static.zebra.lt] has joined #lisp 11:39:40 -!- chp [3d87a5ae@gateway/web/freenode/ip.61.135.165.174] has quit [Quit: Page closed] 11:39:50 I mean, I type back in the REPL five to see the result of (+ 2 3) 11:39:59 -!- hefner [~hefner@ppp-58-9-182-122.revip2.asianet.co.th] has quit [Ping timeout: 255 seconds] 11:39:59 you need (five) 11:40:05 does that mean that you're entering "five" instead of "(five)"? 11:40:08 :/ my bad! 11:40:25 stassats: yes, that's pretty much what I did 11:40:35 lc0180643: try to include the prompt in your code examples. that makes it clearer 11:40:49 as in > (defun five () (+ 2 3)) => FIVE 11:40:53 you're defining a functions, you may define a variable with (defvar five (+ 3 2)) 11:40:57 (five) => 5 11:41:00 Hun: gotcha 11:41:00 -!- humasect [~humasect@S01060018f870b75e.rd.shawcable.net] has quit [Remote host closed the connection] 11:41:00 then "five" would work 11:41:42 question, parntheses are always there wither I input data or commands right? 11:42:06 parenthesis are where they need to be 11:42:27 no. parens have meaning. not like in C-derived languages, you can not add or remove them as you like 11:42:49 Lately I am wondering more and more about if 'x is really better syntactic choice than 'x'. 11:43:18 Zhivago: the meaning of '(foo) is clear. what is the meaning of '(foo')? 11:43:43 of '(foo)'? 11:43:52 that one also 11:44:10 i'd thing '(foo') is meaningless because of the wrong nesting 11:44:13 think 11:44:19 billstclair [~billstcla@gw3.tacwap.org] has joined #lisp 11:44:19 -!- billstclair [~billstcla@gw3.tacwap.org] has quit [Changing host] 11:44:19 billstclair [~billstcla@unaffiliated/billstclair] has joined #lisp 11:44:49 Hun: Um. Examine your prejudices and try again. 11:44:55 thanks folks 11:44:57 -!- lc0180643 [~lc@41.35.197.69] has quit [Read error: Connection reset by peer] 11:45:07 anyone tried clsql on a x86_64 box ? 11:45:55 Zhivago: my objection is just that you have two different paren-style things that interweave then 11:46:10 then: what would you write for '(foo 'bar)? 11:46:13 kushal: why are you asking? 11:46:23 '(foo 'bar')'? 11:46:37 hefner [~hefner@ppp-58-9-113-71.revip2.asianet.co.th] has joined #lisp 11:46:59 haha, imagine people's complaints about lisp syntax if it had that quote style 11:47:20 not only is it full of nail clippings, it's full of bird droppings aswell 11:47:25 oh the noes 11:47:30 stassats, http://fpaste.org/NK0R/ (downloaded through quicklisp ) 11:47:37 OliverUv: they don't complain about single '? 11:47:42 i bet they do 11:47:50 Hun: It depends on what you want to quote ... 11:48:05 you can always use (quote ..) and (function ...), if you're quirky enough 11:48:13 it is trying to build 32bit whereas the system is a x86_64 11:48:20 stassats: I looked at http://www.cliki.net/SPLIT-SEQUENCE and tried the last example and got an error 11:48:39 stassats: is this a standard function or do I have to download something to use this? 11:48:45 -!- hlavaty [~user@77-22-100-178-dynip.superkabel.de] has quit [Ping timeout: 260 seconds] 11:49:14 edlinde: the latter 11:49:34 People seem perfectly happy with "hello, 'I like cheese'." so that embedding argument doesn't hold water. 11:50:12 what would 'foo bar' mean? 11:50:19 '|foo bar| or '(foo bar)? 11:50:24 hlavaty [~user@77-22-100-178-dynip.superkabel.de] has joined #lisp 11:50:28 -!- enupten [~neptune@117.192.74.5] has quit [Remote host closed the connection] 11:50:45 Zhivago: people also have fault-ignoring parsers ;) 11:50:45 what is '|foo bar| ? 11:51:01 OliverUv: the symbol with the name "foo bar" 11:51:14 ah ok, thanks 11:51:23 kushal: i don't know how to persuade it not do this, but installing 32-bit libraries may help 11:51:24 note that there are no caps 11:51:37 stassats, yes 11:52:03 Hun: it's hot in here, so no caps 11:52:29 '|foo bar|, naturally. 11:52:36 ohhhhhh, so when I "comment out" code with #|(some code bla bla bla)|# am I actually just putting in the uninterned symbol with the name some code bla bla bla? 11:52:43 Or, rather 'foo\ bar 11:52:44 no 11:53:05 i imagined it wasn't, but it didn't feel too unlikely either 11:53:15 note the difference between #| and | 11:53:16 OliverUv: no. you'd do that with | 11:53:32 Zhivago: with or without caps? there's further ambiguity 11:53:48 stassats: yeah I thought it might have the same difference as when doing #:asdf instead of :asdf 11:53:53 stassats: crap I don't know where to place this asdf file 11:53:54 '#:|comment bar| would do what you're describing 11:54:02 mm ok 11:54:10 edlinde: use quicklisp 11:54:15 good to konw, thanks 11:54:20 heading back to bed now 11:55:49 -!- petercoulton [~petercoul@152.105.49.241] has quit [Quit: Leaving] 11:56:23 Hun: Not with 'foo\ bar 11:56:40 Running the latest clisp on Windows, it works fine when passed a program as commandline argument, but when I just type clisp by itself to get interactive mode, I get a 'Windows -- no disk' error dialog 11:56:43 stassats, fixed with glibc.i686 11:56:52 Anyone else here come across that, and how to fix it? Google search doesn't show up anything 11:56:55 still don't know why it requires that 11:57:04 Hun: Besides which, CL's case stupidity needs killing with an axe. 11:57:14 -!- pjb [~t@81.202.16.46.dyn.user.ono.com] has quit [Remote host closed the connection] 11:57:19 kushal: though i'd still consider it a problem and contacting the authorities would be good 11:57:28 Zhivago: that's true. upper-case-only ftw 11:57:33 yes 11:57:38 -!- Krystof [~csr21@cpc1-bour2-0-0-cust414.15-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 11:57:42 I will write to the mailing list 11:58:15 -!- simplechat [~simplecha@unaffiliated/simplechat] has quit [Read error: Connection reset by peer] 11:58:28 also I should starting blogs with these various issues you all helped me to solve 11:58:41 -!- SpitfireWP_ [~Spitfire@wikipedia/spitfire] has quit [Quit: It's getting dark, too dark to see.] 12:01:48 dfkjjkfd [~paulh@236-9-ftth.onsnetstudenten.nl] has joined #lisp 12:02:34 pjb [~t@81.202.16.46.dyn.user.ono.com] has joined #lisp 12:02:47 pavelludiq [~quassel@91.139.196.22] has joined #lisp 12:02:47 -!- Harag [~Harag@wbs-41-208-211-15.wbs.co.za] has quit [Remote host closed the connection] 12:03:03 Harag [~Harag@wbs-41-208-211-15.wbs.co.za] has joined #lisp 12:04:32 keep your keyboards fresh! 12:05:12 -!- homie` [~user@xdsl-195-14-197-168.netcologne.de] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 12:05:20 -!- wbooze` [~user@xdsl-195-14-197-168.netcologne.de] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 12:07:06 -!- euphidime [u178@gateway/web/irccloud.com/x-efydvegdifwbkixl] has quit [Max SendQ exceeded] 12:09:20 homie [~user@xdsl-195-14-197-168.netcologne.de] has joined #lisp 12:12:14 wbooze [~user@xdsl-195-14-197-168.netcologne.de] has joined #lisp 12:14:51 b-man_ [~b-man@189.34.49.137] has joined #lisp 12:15:33 someone just said (2 3 +) is how lisp is written can you believe this i am terribly appalled 12:17:13 it's hebrew lisp 12:17:55 Krystof [~csr21@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has joined #lisp 12:22:48 lol 12:22:55 -!- REPLeffect [~REPLeffec@69.54.115.254] has quit [Ping timeout: 265 seconds] 12:23:25 and actual lisp syntax is reverse hebrew notation? 12:23:47 -!- freddie111 [~user@ppp-94-64-152-69.home.otenet.gr] has quit [Remote host closed the connection] 12:23:50 -!- fmu_ [u89@gateway/web/irccloud.com/x-cncielffrjwdgfzs] has quit [Max SendQ exceeded] 12:23:56 sellout [~greg@c-24-61-13-161.hsd1.ma.comcast.net] has joined #lisp 12:24:17 <_3b> (2 3 +) evaluates fine here :p 12:24:49 <_3b> (after setting *read-base* to 2, defining a function named 2, and a variable named 3 at least :) 12:25:20 fmu_ [u89@gateway/web/irccloud.com/x-aruzmdtqxeedpgcp] has joined #lisp 12:25:34 unicode [~user@95.214.7.92] has joined #lisp 12:25:35 -!- edlinde [~edlinde@90-227-7-243-no15.tbcn.telia.com] has quit [Quit: edlinde] 12:26:06 humasect [~humasect@S01060018f870b75e.rd.shawcable.net] has joined #lisp 12:26:16 heh 12:27:50 -!- skaboy [~grant@c0144.aw.cl.cam.ac.uk] has quit [Ping timeout: 260 seconds] 12:27:56 *_3b* wonders if that counts for that hacker test question about changing the value of 4 in a language other than fortran 12:28:38 REPLeffect [~REPLeffec@69.54.115.254] has joined #lisp 12:29:32 *rwallace* thinks it does ^.^ 12:32:14 -!- mega1 [~quassel@catv4E5CABA2.pool.t-online.hu] has quit [Ping timeout: 276 seconds] 12:36:45 -!- kiuma [~kiuma@85-18-55-37.ip.fastwebnet.it] has quit [Quit: Bye bye ppl] 12:38:47 -!- fmu_ [u89@gateway/web/irccloud.com/x-aruzmdtqxeedpgcp] has quit [Max SendQ exceeded] 12:39:07 fmu_ [u89@gateway/web/irccloud.com/x-kgtohlzpwekgnwio] has joined #lisp 12:39:21 multicore lisp? 12:39:36 huh, when searching "anagram" in google it says "Did you mean: nag a ram" 12:39:41 sorry for the off-topic 12:39:41 humasect: sciener CL. 12:39:46 ouu 12:40:13 2008 last release hmmm 12:41:17 lichtblau: are you about? 12:41:44 vroufe [~omx@ppp94-29-9-58.pppoe.spdop.ru] has joined #lisp 12:41:50 -!- lnostdal-laptop [~lnostdal@46.66.23.118.tmi.telenormobil.no] has quit [Ping timeout: 260 seconds] 12:43:11 lispworks 12:43:30 -!- cYmen [~cymen@squint.a-oben.org] has quit [Ping timeout: 272 seconds] 12:43:45 SpitfireWP [~Spitfire@wikipedia/spitfire] has joined #lisp 12:43:59 cYmen [~cymen@squint.a-oben.org] has joined #lisp 12:45:54 jweiss_ [~jweiss@cpe-069-134-009-048.nc.res.rr.com] has joined #lisp 12:46:13 hi 12:46:48 -!- Jasko [~tjasko@c-174-59-223-208.hsd1.pa.comcast.net] has quit [Quit: Leaving] 12:47:26 francogrex [~user@109.130.117.141] has joined #lisp 12:47:42 hi 12:48:13 -!- davertron [~david@c-24-218-166-166.hsd1.vt.comcast.net] has quit [Remote host closed the connection] 12:48:41 this is amazing: http://www.youtube.com/watch?v=zg_Pbcfhx98 and http://www.youtube.com/watch?v=kS4cqa9CDVY 12:48:45 Jasko [~tjasko@c-174-59-223-208.hsd1.pa.comcast.net] has joined #lisp 12:48:57 do you think something like that can be done using common lisp ? 12:49:02 yes 12:49:04 lichtblau: someone here at $work reported that he'd taken a look at sb-heapdump, and that it didn't currently work 12:49:15 that wasn't much of a problem because he was really after it for reading rather than using, but still 12:50:30 i got sb-heapdump semi-working on then-current sbcl, it's not working because of the changes in internals 12:50:35 -!- pavelludiq [~quassel@91.139.196.22] has quit [Read error: Connection reset by peer] 12:51:15 <_3b> main issue i could see with doing that in lisp is GCs that want half the heap free for a full GC make hugfe data sets a bit more of a hassle 12:51:27 <_3b> *huge data sets 12:52:06 -!- unicode [~user@95.214.7.92] has quit [Quit: leaving] 12:52:15 it's from http://www.flightgear.org/ the src code shows mainly c/c++ 12:52:42 great work really 12:52:52 <_3b> yeah, wouldn't expect it to be lisp, but that is just due to lisp not being popular with games/gfx types 12:53:15 _3b: but in principle it can be done with lisp right? 12:53:29 kral [~kral@93.185.125.210] has joined #lisp 12:53:32 namaste 12:53:37 namaste 12:53:49 francogrex: lisp is a general-purpose language. you know what the words mean, right? 12:53:52 <_3b> well, in principle, you could just write a c compiler in lisp and compile the c code :p 12:55:07 francogrex: type hinting would help optimization a lot, which is important for gaming 12:55:19 <_3b> lisp does 3d graphics just fine, and at least a few can do math well enough to run the simulation, so just a SMOP from that point aside from the huge data set issue 12:55:32 attila_lendvai [~attila_le@catv-89-132-188-47.catv.broadband.hu] has joined #lisp 12:56:00 some lisps have static spaces 12:56:26 *_3b* is assuming that is actually one of the ones with huge data sets, i don't really remember which flight sim is which 12:56:43 any world is generally a pretty big data set 12:56:55 <_3b> depends on how detailed it is 12:57:01 -!- fmu_ [u89@gateway/web/irccloud.com/x-kgtohlzpwekgnwio] has quit [Max SendQ exceeded] 12:57:04 <_3b> and how much heap you have :) 12:57:05 fmu_ [u89@gateway/web/irccloud.com/x-mxateypdumcvbmyp] has joined #lisp 12:57:14 -!- sixpoint8 [~jaykub@ip68-12-204-131.ok.ok.cox.net] has quit [Ping timeout: 255 seconds] 12:57:21 _3b: so from your pov, it's the GC and the huge data sets that may be an obstacle then... 12:57:57 <_3b> not an insurmountable one, just a hassle 12:58:42 ok 12:58:49 <_3b> probably mostly involving hiding the stuff from the GC and handling it manually 12:59:12 sixpoint8 [~jaykub@ip68-12-204-131.ok.ok.cox.net] has joined #lisp 12:59:14 <_3b> but you would probably be doing most of that work for LoD anyway 12:59:53 LoD? 12:59:54 gravicappa [~gravicapp@ppp85-141-167-181.pppoe.mtu-net.ru] has joined #lisp 13:00:48 Life or Death? 13:01:20 Lisp on Drugs? 13:01:29 -!- Athas [~athas@82.211.209.226] has quit [Remote host closed the connection] 13:01:56 jweiss__ [~jweiss@cpe-069-134-009-048.nc.res.rr.com] has joined #lisp 13:02:34 that was a fun game 13:03:07 I totally felt prepared for becoming a surgeon after that 13:03:21 that and watching Weird Al's parody videos of Madonna songs 13:05:05 <_3b> sorry, 'Level of Detail' 13:05:59 Adamant: type hinting, that shouldn't be a problem 13:06:40 well, not compared to what happens when you give a patient allergic to shellfish a CT scan, no 13:07:01 <_3b> keep in mind that there is also lots of stuff that would be a hassle in a project like that in any language, not just lisp :) 13:07:18 *stassats* stops understanding a thing in this channel 13:07:56 stassats: the difference between me and a Surrealist is that I am not a Surrealist. 13:08:55 _3b: ok to wrap up: then would be be safe to say, if resources are there (people willing to work on such large project), with what we have in Cl right now, it's not impossoble to do what the guys at flightgear.org have done... 13:09:22 nothing is impossible 13:09:31 <_3b> yeah, not really a meaningful question 13:09:40 <_3b> with enough resources, you could makit out of rocks :p 13:09:47 <_3b> *make it 13:10:00 no, with enough resources and xkcd's wanderer 13:10:17 and a really deep level of attention 13:10:19 *_3b* was thinking more of melting them to make computer chips, but that works too 13:10:45 ah that's quicker 13:10:46 rme [~rme@pool-70-105-122-27.chi.dsl-w.verizon.net] has joined #lisp 13:13:16 ok enough bull... I go back to play the awesome games... I bet several here have downloaded and are playing this right now 13:13:34 francogrex: how much did you bet? 13:13:42 *_3b* suspects the people who would want to, already have 13:13:50 pretty much 13:14:28 *stassats* played flightgear three years ago 13:14:30 jdz: you owe me 50 $ 13:14:35 francogrex: i'd bet people from this play http://ai-contest.com/index.php instead 13:14:44 from this channel 13:14:56 nefo [~nefo@2402:f000:5:8f01::143:81] has joined #lisp 13:14:56 -!- nefo [~nefo@2402:f000:5:8f01::143:81] has quit [Changing host] 13:14:56 nefo [~nefo@unaffiliated/nefo] has joined #lisp 13:15:16 the only games I'm playing involve dead zombies 13:15:23 ohh, lisp entry is still at #1 13:15:25 oh, and there's also IRC 13:15:43 it's a MMOC 13:15:50 massive multiplayer online chat 13:15:55 -!- churib [~tg@dslb-088-071-150-055.pools.arcor-ip.net] has quit [Ping timeout: 245 seconds] 13:16:04 *stassats* is about to go to play some squash 13:16:09 computer games are boring 13:16:26 so is squash, but have fun anyway 13:16:40 good luck! :) 13:17:26 -!- SpitfireWP [~Spitfire@wikipedia/spitfire] has quit [Ping timeout: 264 seconds] 13:17:34 -!- ramkrsna [~ramkrsna@unaffiliated/ramkrsna] has quit [Remote host closed the connection] 13:18:28 -!- francogrex [~user@109.130.117.141] has quit [Remote host closed the connection] 13:18:29 churib [~tg@dslb-088-071-150-055.pools.arcor-ip.net] has joined #lisp 13:19:07 -!- joe4 [~joe@c-24-126-150-146.hsd1.ga.comcast.net] has quit [Quit: leaving] 13:20:11 SpitfireWP [~Spitfire@wikipedia/spitfire] has joined #lisp 13:26:45 HG` [~HG@xdsl-92-252-105-72.dip.osnanet.de] has joined #lisp 13:29:26 -!- beach [~user@ABordeaux-552-1-85-252.w90-38.abo.wanadoo.fr] has quit [Ping timeout: 276 seconds] 13:30:12 gigamonkey [~user@adsl-99-17-207-80.dsl.pltn13.sbcglobal.net] has joined #lisp 13:30:19 -!- mixo [~mshiburi@2001:4200:7000:3:213:72ff:fe87:5055] has quit [Quit: Ex-Chat] 13:30:38 -!- _pw_ [~user@123.112.77.61] has quit [Ping timeout: 264 seconds] 13:34:13 -!- kaemo [~kaemo@d38-66.icpnet.pl] has quit [Remote host closed the connection] 13:37:26 xan_ [~xan@rrcs-69-193-205-118.nyc.biz.rr.com] has joined #lisp 13:38:01 ctp_ [~user@talula.plus.com] has joined #lisp 13:38:50 Athas [~athas@95.142.154.222] has joined #lisp 13:39:48 -!- flip214 [~marek@unaffiliated/flip214] has quit [Remote host closed the connection] 13:41:46 -!- gigamonkey [~user@adsl-99-17-207-80.dsl.pltn13.sbcglobal.net] has quit [Ping timeout: 272 seconds] 13:44:00 Hrmm, closer-mop doesn't seem to implement (setf class-direct-superclasses) 13:45:20 edlinde [~edlinde@90-227-7-243-no15.tbcn.telia.com] has joined #lisp 13:46:01 -!- vroufe [~omx@ppp94-29-9-58.pppoe.spdop.ru] has quit [Ping timeout: 250 seconds] 13:46:11 is it an accessor? 13:46:39 through my quick skimming through mop, it has only a reader 13:46:47 -!- lemoinem [~swoog@216.252.67.62] has quit [Remote host closed the connection] 13:47:12 lemoinem [~swoog@58-73-252-216.dsl.colba.net] has joined #lisp 13:47:22 ziarkaen [~ziarkaen@stu301.queens.ox.ac.uk] has joined #lisp 13:47:30 stassats: It's an accessor. You mean skimming through AMOP? 13:47:53 stassats: pg 285. 13:48:06 -!- Athas [~athas@95.142.154.222] has quit [Remote host closed the connection] 13:48:13 My program finally substitutes newlines for spaces, but my greek words now look like back slashes and numbers; the encoding is messed up. What could be causing that? 13:49:03 What's the coding system used? 13:49:30 pjb, utf-8 13:49:44 sellout: i meant http://www.alu.org/mop/ 13:49:52 _nix00 [~Adium@114.92.125.65] has joined #lisp 13:50:31 -!- V-ille [~ville@70.42.89.16] has quit [Quit: Leaving.] 13:50:44 and it's only a reader in sb-mop too 13:51:42 and i don't have time to look into amop, i'm already late 13:51:45 stassats: Yeah, it looks like closer-mop doesn't implement class-direct-superclasses for anything  every impl implements it. But I hadn't realized that "MOP" was different from what's in AMOP. 13:51:48 -!- _nix00 [~Adium@114.92.125.65] has left #lisp 13:51:49 pjb, (with-open-file (stream-in "/home/lat/l/i/scm.txt" :external-format :utf-8) 13:51:50 Disappointed :( 13:52:05 -!- ctp_ [~user@talula.plus.com] has quit [Ping timeout: 250 seconds] 13:52:30 sellout: does sbcl implement it? 13:52:49 stassats: I dunno yet  using CCL here (which doesn't) 13:53:33 The function (COMMON-LISP:SETF SB-MOP:CLASS-DIRECT-SUPERCLASSES) is undefined. 13:54:46 -!- HG` [~HG@xdsl-92-252-105-72.dip.osnanet.de] has quit [Quit: Leaving.] 13:55:03 So, now I'm wondering if there's any way for me to take a class hierarchy and add some "mix-ins" to it without making a complete parallel hierarchy. 13:55:10 urandom__ [~user@p548A2E28.dip.t-dialin.net] has joined #lisp 13:55:44 can't you use ensure-class-using-class? 13:56:17 page 285 in AMOP is closette 13:56:29 which is related to but different from CLOS and its MOP 13:56:40 ensure-class is the way you want to go 13:57:01 Right  thanks guys. 13:57:01 beach has a paper somewhere about "stealth mixins"; Pascal's contextl will have some machinery to do that too (because that's how layers are implemented) 13:57:49 dlowe [~dlowe@ita4fw1.itasoftware.com] has joined #lisp 14:01:50 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 264 seconds] 14:02:49 jajcloz [~jaj@pool-108-7-68-199.bstnma.fios.verizon.net] has joined #lisp 14:03:51 CTP [~user@talula.plus.com] has joined #lisp 14:05:29 petercoulton [~petercoul@cpc2-midd16-2-0-cust169.11-1.cable.virginmedia.com] has joined #lisp 14:07:10 Bronsa [~bronsa@host165-180-dynamic.4-87-r.retail.telecomitalia.it] has joined #lisp 14:09:50 KalifG [~user@helios.isc.tamu.edu] has joined #lisp 14:09:50 BrandLeeJones [~BrandLeeJ@chello062178064156.22.11.vie.surfer.at] has joined #lisp 14:11:05 -!- mcsontos [~mcsontos@nat/redhat/x-mvowowlissybvcpv] has quit [Ping timeout: 260 seconds] 14:11:19 drdo [~user@2.208.54.77.rev.vodafone.pt] has joined #lisp 14:14:46 -!- sellout [~greg@c-24-61-13-161.hsd1.ma.comcast.net] has quit [Quit: sellout] 14:16:42 -!- Xach [~xach@pdpc/supporter/professional/xach] has quit [Ping timeout: 265 seconds] 14:17:07 Xach [~xach@pdpc/supporter/professional/xach] has joined #lisp 14:18:23 hargettp [~hargettp@dhcp-174.mirrorimage.net] has joined #lisp 14:19:35 -!- xan_ [~xan@rrcs-69-193-205-118.nyc.biz.rr.com] has quit [Ping timeout: 265 seconds] 14:20:24 bobbysmith007 [~russ@216.155.97.1] has joined #lisp 14:21:06 mbohun [~mbohun@ppp115-156.static.internode.on.net] has joined #lisp 14:21:17 -!- p_l|home [~pl@pp84.internetdsl.tpnet.pl] has quit [Ping timeout: 276 seconds] 14:23:28 madnificent [~madnifice@83.101.62.132] has joined #lisp 14:23:50 -!- Xach [~xach@pdpc/supporter/professional/xach] has quit [Ping timeout: 245 seconds] 14:24:01 Xach [~xach@pdpc/supporter/professional/xach] has joined #lisp 14:25:53 -!- churib [~tg@dslb-088-071-150-055.pools.arcor-ip.net] has quit [Ping timeout: 255 seconds] 14:27:20 -!- Adamant [~Adamant@unaffiliated/adamant] has quit [Quit: Adamant] 14:27:23 -!- Harag [~Harag@wbs-41-208-211-15.wbs.co.za] has quit [Ping timeout: 255 seconds] 14:28:35 -!- dfox [~dfox@ip-89-176-11-19.net.upcbroadband.cz] has quit [Ping timeout: 265 seconds] 14:30:08 -!- Krystof [~csr21@cpc5-bour5-2-0-cust340.15-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 14:31:54 -!- Xach [~xach@pdpc/supporter/professional/xach] has quit [Quit: leaving] 14:32:52 -!- jweiss__ [~jweiss@cpe-069-134-009-048.nc.res.rr.com] has quit [Quit: Ex-Chat] 14:33:55 -!- redline6561 [~redline@c-66-56-55-169.hsd1.ga.comcast.net] has quit [Ping timeout: 240 seconds] 14:35:20 V-ille [~ville@ville-hp.dhcp.fnal.gov] has joined #lisp 14:36:26 SegFaultAX [~SegFaultA@c-98-234-1-162.hsd1.ca.comcast.net] has joined #lisp 14:39:56 chp [~chp@114.113.65.235] has joined #lisp 14:40:02 drl: Try: (with-open-file (stream-in "/home/lat/l/i/scm.txt" :external-format :utf-8) (loop for ch = (read-char stream-in nil nil) while ch do (princ (char-name ch)) (princ " ") finally (terpri) (finish-output))) 14:40:25 If you get greek characters, then perhaps the problem is rather with the output. 14:40:25 HG` [~HG@xdsl-92-252-105-72.dip.osnanet.de] has joined #lisp 14:42:53 sellout [~greg@64.134.66.79] has joined #lisp 14:43:16 -!- b-man_ [~b-man@189.34.49.137] has quit [Read error: Operation timed out] 14:43:31 -!- edlinde [~edlinde@90-227-7-243-no15.tbcn.telia.com] has quit [Quit: edlinde] 14:45:57 -!- setheus [~setheus@cpe-70-116-140-134.tx.res.rr.com] has quit [Quit: leaving] 14:45:59 pjb, thanks. I'll report back on the results. It may take a few minutes. I'm checking to make sure I don't have corrupt data. 14:48:22 LiamH [~none@pdp8.nrl.navy.mil] has joined #lisp 14:50:24 tims [ae1dc0b6@gateway/web/freenode/ip.174.29.192.182] has joined #lisp 14:50:50 -!- tims is now known as Guest77279 14:50:56 beach [~user@ABordeaux-552-1-80-70.w90-38.abo.wanadoo.fr] has joined #lisp 14:51:02 Good afternoon everyone! 14:53:26 -!- Beetny [~Beetny@ppp118-208-8-91.lns20.bne1.internode.on.net] has quit [Ping timeout: 265 seconds] 14:54:21 hello beach! 14:54:23 setheus [~setheus@cpe-70-116-140-134.tx.res.rr.com] has joined #lisp 14:55:01 -!- nefo [~nefo@unaffiliated/nefo] has quit [Quit: Leaving] 14:56:13 CTP` [~user@talula.plus.com] has joined #lisp 14:56:14 aDuck [~aduck@bl5-3-53.dsl.telepac.pt] has joined #lisp 14:56:39 edlinde [~edlinde@90-227-7-243-no15.tbcn.telia.com] has joined #lisp 14:59:05 -!- CTP [~user@talula.plus.com] has quit [Ping timeout: 240 seconds] 15:01:25 -!- rTypo [~rTypo@pc54.clicknet.iasi.rdsnet.ro] has quit [Ping timeout: 250 seconds] 15:01:36 -!- Guest77279 [ae1dc0b6@gateway/web/freenode/ip.174.29.192.182] has quit [Quit: Page closed] 15:06:18 kclifton [~kclifton@s198-166-45-245.ab.hsia.telus.net] has joined #lisp 15:07:58 redline6561 [~redline@gate-20.spsu.edu] has joined #lisp 15:09:21 pavelludiq [~quassel@91.139.196.22] has joined #lisp 15:10:30 Does (- (nth-value 8 (decode-universal-time 0))) look like a reliable way to get a system's default timezone? 15:11:39 there's pretty much no portable way to get a system's default timezone 15:12:02 portable, reliable way, at least 15:12:19 dlowe: hah, somehow I'm not suprised 15:12:32 well, this isn't limited to lisp languages, either 15:12:38 the unix std locales stuff doesn't work? 15:12:48 Lycurgus: sure. On unix. 15:13:23 is this dos or mac or what? 15:13:38 Then you can just look up a unix-epoch-encoded time in /etc/localtime and get your offset from GMT 15:14:05 mac is unix, windows probably has an api, so how is this so hard to do? 15:14:25 or I see, you're asking about a hypothetical lisp cross platform std'zation 15:14:31 http://www.chronos-st.org/Discovering%20the%20Local%20Time%20Zone--Why%20It%27s%20a%20Hard%20Problem.html 15:14:32 *o i c 15:14:55 marijnjh` [~user@p5799E9D0.dip.t-dialin.net] has joined #lisp 15:16:04 -!- marijnjh [~user@p5799EE2F.dip.t-dialin.net] has quit [Disconnected by services] 15:16:06 I did say portable. 15:16:08 -!- marijnjh` is now known as marijnjh 15:16:46 marijnjh: the method your lisp implementation is probably "good enough," honestly 15:17:31 sbcl gives me 293/900 for my hack, which is not what my computer's locale says 15:17:32 grrr 15:17:36 works on allegro, though 15:20:05 marijnjh: define system default timezone when your computer is a cluster spaning 17 time zones? 15:20:52 (getenv "TZ") is as good an anwer as your form. Yours have the advantage of being conformant. Not that it will give better results, for any definition of better. 15:21:22 (because of the way clhs specified it). 15:21:41 -!- ziarkaen [~ziarkaen@stu301.queens.ox.ac.uk] has quit [Ping timeout: 255 seconds] 15:23:42 pjb: right, that's a problem I have *all the time*. damn those world-spanning-clusters abstracted as a single machine 15:25:30 *beach* detects some sarcasm there. 15:26:06 I think most people, when they think they care about exact terms, they only want a reasonable wall-clock approximation so they can stop thinking about timekeeping 15:26:37 rabbit holes look simple from above the ground :p 15:27:14 -!- rme [rme@clozure-AC1C217F.chi.dsl-w.verizon.net] has quit [Quit: rme] 15:27:14 -!- rme [~rme@pool-70-105-122-27.chi.dsl-w.verizon.net] has quit [Quit: rme] 15:28:25 Davidbrcz [~david@212-198-92-25.rev.numericable.fr] has joined #lisp 15:28:28 -!- petercoulton [~petercoul@cpc2-midd16-2-0-cust169.11-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 15:31:43 p_l|home [~pl@pp84.internetdsl.tpnet.pl] has joined #lisp 15:33:13 -!- sonnym [~evissecer@rrcs-184-74-137-167.nys.biz.rr.com] has quit [Quit: Leaving.] 15:33:26 -!- superflit [~superflit@c-24-9-162-197.hsd1.co.comcast.net] has quit [Quit: superflit] 15:37:06 b-man_ [~b-man@189.34.49.137] has joined #lisp 15:38:42 -!- c|mell [~cmell@188-220-238-74.zone11.bethere.co.uk] has quit [Quit: Leaving] 15:39:04 bobbysmith0071 [~russ@216.155.97.1] has joined #lisp 15:40:33 -!- bobbysmith007 [~russ@216.155.97.1] has quit [Ping timeout: 245 seconds] 15:40:50 -!- dnm_ [~dnm@c-68-34-57-169.hsd1.va.comcast.net] has quit [Ping timeout: 272 seconds] 15:43:20 beach: I'm looking at SICL's LOOP implementation 15:44:33 beach: would it be correct to generate the TAGBODY inside a BLOCK? 15:44:55 mvilleneuve: I think so, yes. 15:45:25 mvilleneuve: The block would have the name of the loop (or nil if no explicit name has been given). 15:45:30 -!- gravicappa [~gravicapp@ppp85-141-167-181.pppoe.mtu-net.ru] has quit [Ping timeout: 245 seconds] 15:45:33 beach: right 15:45:41 nyef [~nyef@pool-64-222-179-143.man.east.myfairpoint.net] has joined #lisp 15:47:43 -!- bobbysmith0071 [~russ@216.155.97.1] has quit [Quit: Leaving.] 15:47:59 -!- pocket_ [~pocket_@p4252-ipbf3210hodogaya.kanagawa.ocn.ne.jp] has quit [Remote host closed the connection] 15:50:16 -!- marijnjh [~user@p5799E9D0.dip.t-dialin.net] has left #lisp 15:50:52 CTP`` [~user@talula.plus.com] has joined #lisp 15:51:51 churib [~tg@dslb-088-071-150-055.pools.arcor-ip.net] has joined #lisp 15:52:04 rme [~rme@pool-70-105-122-27.chi.dsl-w.verizon.net] has joined #lisp 15:54:05 -!- CTP` [~user@talula.plus.com] has quit [Ping timeout: 240 seconds] 15:54:27 aurelien [~Libre@fsf/member/aurelien] has joined #lisp 15:57:04 -!- dean_ [~deanz@onshore-gw.logika.net] has quit [Remote host closed the connection] 15:58:54 sellout- [~greg@64.134.66.79] has joined #lisp 15:58:55 -!- sellout [~greg@64.134.66.79] has quit [Read error: Connection reset by peer] 15:59:00 -!- sellout- is now known as sellout 16:00:52 plediii_ [~plediii@adsl-99-23-105-181.dsl.hstntx.sbcglobal.net] has joined #lisp 16:01:27 mega1 [~quassel@catv4E5CABA2.pool.t-online.hu] has joined #lisp 16:01:39 morphling [~stefan@gssn-5f757ffa.pool.mediaWays.net] has joined #lisp 16:02:25 -!- Kovensky [~kovensky@abraxo.bluebottle.net.au] has quit [Ping timeout: 240 seconds] 16:02:31 phf [~user@38.104.111.94] has joined #lisp 16:02:53 -!- Hun [~Hun@80.81.19.29] has quit [Remote host closed the connection] 16:02:57 milanj [~milanj_@93-86-57-180.dynamic.isp.telekom.rs] has joined #lisp 16:03:13 For a computer program, the only time zone that should matter is UTC. For the user, the timezone that matters depends on too many parameters. Just ask the user what time zone he wants its data be displayed in. 16:03:49 -!- plediii [~plediii@nat-128-42-156-192.rice.edu] has quit [Ping timeout: 250 seconds] 16:03:49 -!- plediii_ is now known as plediii 16:04:03 -!- edlinde [~edlinde@90-227-7-243-no15.tbcn.telia.com] has quit [Quit: edlinde] 16:04:22 beach: I'm having a problem when parsing e.g. (loop with x = 1), the code (loop.lisp line 419) attempts to create an instance of a class with unknown slots 16:04:35 CTP``` [~user@talula.plus.com] has joined #lisp 16:04:37 -!- nikodemus [~nikodemus@cs181058025.pp.htv.fi] has quit [Quit: This computer has gone to sleep] 16:04:50 -!- redline6561 [~redline@gate-20.spsu.edu] has quit [Ping timeout: 264 seconds] 16:06:54 beach: I'll probably need to take more time and look at the parsing code... 16:07:05 -!- CTP`` [~user@talula.plus.com] has quit [Ping timeout: 240 seconds] 16:07:40 sonnym [~evissecer@singlebrookvpn.lightlink.com] has joined #lisp 16:08:07 -!- mvilleneuve [~mvilleneu@LLagny-156-36-4-214.w80-14.abo.wanadoo.fr] has left #lisp 16:08:49 -!- varjag [~eugene@122.62-97-226.bkkb.no] has quit [Remote host closed the connection] 16:09:20 dean [~deanz@64.241.87.62] has joined #lisp 16:10:13 -!- dean [~deanz@64.241.87.62] has quit [Remote host closed the connection] 16:10:17 -!- jdz [~jdz@193.206.22.97] has quit [Ping timeout: 255 seconds] 16:10:29 -!- humasect [~humasect@S01060018f870b75e.rd.shawcable.net] has quit [Remote host closed the connection] 16:11:39 dean [~deanz@64.241.87.62] has joined #lisp 16:12:46 -!- dean [~deanz@64.241.87.62] has quit [Remote host closed the connection] 16:13:14 dean [~deanz@onshore-gw.logika.net] has joined #lisp 16:14:11 superflit [~superflit@184-96-119-251.hlrn.qwest.net] has joined #lisp 16:15:14 Kovensky [~kovensky@abraxo.bluebottle.net.au] has joined #lisp 16:16:32 -!- plediii [~plediii@adsl-99-23-105-181.dsl.hstntx.sbcglobal.net] has quit [Quit: plediii] 16:16:56 kmwallio [~kmwallio@host113-70.brownlie.fit.edu] has joined #lisp 16:17:08 whee_ [~whee@smaertness.net] has joined #lisp 16:17:17 -!- kral [~kral@93.185.125.210] has quit [Quit: Sto andando via] 16:18:02 Xach [~xach@pdpc/supporter/professional/xach] has joined #lisp 16:19:05 -!- whee [~whee@cpe-69-207-148-170.rochester.res.rr.com] has quit [Ping timeout: 276 seconds] 16:19:06 cffi's incompatible cffi-uffi-compat breaks several uffi-using projects, like clsql. i opened a launchpad bug, but is there more I can do to help? 16:20:10 Xach: patches are welcome 16:20:14 -!- superflit [~superflit@184-96-119-251.hlrn.qwest.net] has quit [Ping timeout: 276 seconds] 16:22:53 Ok, I'll see about doing that. 16:24:00 *Xach* tries to lure Xof back to comp.lang.lisp with nil-specialized array discussion 16:24:47 nil-specialized? 16:25:03 is anyone aware of decent docs for 'cells'? the manifesto and the examples are pretty scamp on deatils 16:25:06 details* 16:25:18 -!- aerique [euqirea@xs2.xs4all.nl] has quit [Quit: ...] 16:25:54 yan_: as far as i can tell, that is by design. 16:27:04 Xach: so it's supposed to be intuitive and all the info in the manifesto should be enough? 16:28:20 yan_: Cells is ridiculously hard to get running as well 16:28:39 hm, define 'ridiculously hard'. 16:29:25 yan_: unless you know enough CL to get around the source, you can forget about compiling the official sources on anything other than ACL, or so it seems 16:29:47 p_l|home: kenny has fixed things up enough to at least build in quicklisp. 16:29:49 yan_: https://github.com/Ramarren/cells <--- have this fork, I recall it compiled and worked 16:29:57 Xach: oh? 16:29:59 nice 16:30:02 p_l|home: ouch. so not a good idea to try to get this to run as a total lisp noob? 16:30:16 yan_: I was in your place 2~3 years ago 16:30:35 i'm trying to model some data structures with relationships, which i think cells would be very good for (or at least from the superficial understanding i have of it) 16:30:45 I give Kenny credit for finally making me learn CL, but Cells wasn't that fun 16:30:46 nil-specialized being those arrays with an element-type of NIL, thus being strings? 16:30:49 yan_: I think the author prefers to use lack of documentation and ease of use as a filter. If you're not motivated enough to figure it out despite the difficulty, maybe you're not his ideal Cells user. 16:31:02 nyef: yes 16:31:06 the link I gave compiles on SBCL, at least. 16:32:15 there's also at least one or two competing systems - computed-class and IIRC something called formulate 16:32:22 delYsid` [~user@chello084115136207.3.graz.surfer.at] has joined #lisp 16:32:28 Remember the recentish thing about SBCL failing to build on non-unicode systems? 16:32:44 symbolicweb's STM provided something similar to cells as well 16:33:11 nyef: yes 16:34:04 That was due to some logic around string types being built just wrongly enough that the NIL-strings fell through a case, and since base-char and character are the same type without unicode the array accessors got miscompiled. 16:34:38 nice! 16:35:18 -!- Bronsa [~bronsa@host165-180-dynamic.4-87-r.retail.telecomitalia.it] has quit [Quit: Lost terminal] 16:35:25 Took an entire day to sort out what was going on, even after finding the failing commit. 16:35:58 -!- delYsid [~user@chello084115136207.3.graz.surfer.at] has quit [Ping timeout: 245 seconds] 16:37:03 Xach: should the current quicklisp include cells? 16:39:35 schell [~schellsci@70-36-199-232.dsl.static.sonic.net] has joined #lisp 16:39:38 skaboy [~grant@c0144.aw.cl.cam.ac.uk] has joined #lisp 16:39:44 _s1gma [~herpderp@77.107.164.131] has joined #lisp 16:40:04 yan_: It should not. At the time I was compiling things, cells did not build. It will be in next month, though, assuming it doesn't break between now and then. 16:41:29 -!- sellout [~greg@64.134.66.79] has quit [Read error: Connection reset by peer] 16:41:51 sellout [~greg@64.134.66.79] has joined #lisp 16:42:58 -!- aurelien [~Libre@fsf/member/aurelien] has quit [Quit: be free!!] 16:44:21 redline6561 [~redline@c-66-56-55-169.hsd1.ga.comcast.net] has joined #lisp 16:45:14 -!- kmwallio [~kmwallio@host113-70.brownlie.fit.edu] has quit [Quit: have a test :'(] 16:45:24 vroufe [~omx@ppp91-77-243-60.pppoe.mtu-net.ru] has joined #lisp 16:46:23 -!- sellout [~greg@64.134.66.79] has quit [Ping timeout: 245 seconds] 16:47:02 -!- KalifG [~user@helios.isc.tamu.edu] has quit [Ping timeout: 276 seconds] 16:50:56 gravicappa [~gravicapp@ppp85-140-116-26.pppoe.mtu-net.ru] has joined #lisp 16:54:40 -!- nha [~prefect@imamac13.epfl.ch] has quit [Ping timeout: 264 seconds] 16:55:06 nefo [~nefo@unaffiliated/nefo] has joined #lisp 16:56:21 sellout [~greg@c-24-61-13-161.hsd1.ma.comcast.net] has joined #lisp 16:57:06 Xach: what do you think about possible automation for quicklisp to test if the packages build properly? i.e. something that could be put on the web, regularly building all possible packages using few different implementations and possibly running the tests 16:57:58 rpx__ [~rpx@138-180-109-153-rev.idiap.ch] has joined #lisp 16:58:00 -!- drl [~lat@125.167.140.159] has quit [Ping timeout: 245 seconds] 16:58:18 Hi, anyone here who have ever used lisp-magick (the bindings for image magick) 16:58:19 -!- Joreji [~thomas@85-113.eduroam.RWTH-Aachen.DE] has quit [Ping timeout: 240 seconds] 16:58:19 p_l|home: i have automation for that under sbcl, but i need to add more implementations and publish the results. 16:58:30 Saturnation [~dsouth@72.95.93.7] has joined #lisp 16:59:03 makks [~makks@p5DE8E68A.dip.t-dialin.net] has joined #lisp 16:59:07 hi 16:59:16 rpx__: I haven't used them, but maybe I can help if you're having trouble. What's up? 16:59:25 why does (eq :users (make-symbol "users")) evaluate to nil? 16:59:43 Xach: well I have problems with the filters. They seem to be missing :( 16:59:53 and what do I have to do to get :users out of "USERS"? 16:59:56 Xach: nice! 17:00:02 makks: Why would you expect it to? 17:00:15 Xach: sidenote, I have been starting to use quicklisp,, really great ! thanks for that ! 17:00:17 rpx__: Oh, sorry. I'm afraid in that case I don't know enough to help. 17:00:45 rpx__: glad you like it. now you can move on to new and exciting problems like lisp-magick not working properly :) 17:00:47 Xach: :foo is a symbol right? it gets printed to FOO, now I want to read it back into a symbol? 17:01:15 X-02 [~kohei@p1021-ipbf405kyoto.kyoto.ocn.ne.jp] has joined #lisp 17:01:20 makks: :foo is a keyword (which is a symbol) 17:01:26 -!- chp [~chp@114.113.65.235] has quit [Quit: leaving] 17:01:30 Xach: Exactly :), But seriously, I really like it. I think it is going to make the life of a lot of people a lot simpler 17:02:00 sykopomp: so how do I get the keyword? 17:02:01 edlinde [~edlinde@90-227-7-243-no15.tbcn.telia.com] has joined #lisp 17:02:06 dto [~dto@pool-96-252-62-25.bstnma.fios.verizon.net] has joined #lisp 17:02:10 -!- fmu_ [u89@gateway/web/irccloud.com/x-mxateypdumcvbmyp] has quit [Remote host closed the connection] 17:02:10 -!- lispmeister [u272@gateway/web/irccloud.com/x-olmuongyqnyxvvuu] has quit [Remote host closed the connection] 17:02:10 Alright guys, time to run for a train. I wish you a great week-end 17:02:17 makks: and if you intend to read it back in, you should print it with that intent. 17:02:21 clhs prin1 17:02:22 http://www.lispworks.com/reference/HyperSpec/Body/f_wr_pr.htm 17:02:32 makks: find-symbol is one way. you need to pass the right name (the name of :users is most likely not "users" but "USERS") and the right package. 17:02:36 -!- rpx__ [~rpx@138-180-109-153-rev.idiap.ch] has quit [Client Quit] 17:05:43 if you read it in in READable format, keywords will just be read into the KEYWORD package, neh? 17:05:52 returns nil, because I have not defined :users anywhere 17:06:16 fiveop [~fiveop@dslb-084-056-184-061.pools.arcor-ip.net] has joined #lisp 17:06:56 makks: (eq :users (find-symbol ...)) returns nil? 17:07:34 (eq :users (find-symbol "USERS")) 17:07:36 NIL 17:07:53 makks: You didn't provide the right package to FIND-SYMBOL. 17:08:04 what package could that be? 17:08:04 makks: in what package is :users interned? 17:08:10 in none 17:08:13 you can find out with symbol-package. 17:08:15 makks: incorrect. 17:08:54 ok 17:08:58 I got it 17:09:00 -!- vroufe [~omx@ppp91-77-243-60.pppoe.mtu-net.ru] has quit [Ping timeout: 260 seconds] 17:09:03 -!- REPLeffect [~REPLeffec@69.54.115.254] has quit [Ping timeout: 265 seconds] 17:09:14 -!- aDuck [~aduck@bl5-3-53.dsl.telepac.pt] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)] 17:09:43 -!- The_Fellow [~spider1@glida.mooo.com] has quit [Ping timeout: 245 seconds] 17:09:45 so everytime I use a symbol, it gets interned in the (in my case) "keywords" package? 17:10:00 makks: No. But keyword symbols do. 17:10:07 -!- lonstein [lonstein@ohno.mrbill.net] has quit [Ping timeout: 250 seconds] 17:10:07 -!- spacebat [~spacebat@ubermonkey.net] has quit [Ping timeout: 250 seconds] 17:10:08 -!- Kovensky [~kovensky@abraxo.bluebottle.net.au] has quit [Ping timeout: 255 seconds] 17:10:27 I dont understand the difference between keywords and symbols yet... 17:10:42 what is the killer feature of keywords? 17:11:08 Guthur [~Guthur@cpc11-belf9-2-0-cust855.2-1.cable.virginmedia.com] has joined #lisp 17:11:13 -!- cods [~cods@tuxee.net] has quit [Ping timeout: 252 seconds] 17:11:15 I just used keywords as a cheap alternative to strings so far... 17:11:16 makks: http://www.lispworks.com/documentation/HyperSpec/Body/t_kwd.htm has a list of three things. 17:11:38 The_Fellow [~spider1@glida.mooo.com] has joined #lisp 17:11:54 spacebat [~spacebat@ubermonkey.net] has joined #lisp 17:11:59 lonstein [lonstein@ohno.mrbill.net] has joined #lisp 17:12:07 -!- hlavaty [~user@77-22-100-178-dynip.superkabel.de] has quit [Remote host closed the connection] 17:12:17 cods [~cods@tuxee.net] has joined #lisp 17:12:37 instead of '("foo" "bar") I used (:foo :bar) to specify a path in a tree structure, is that bad? 17:12:47 makks: No. 17:12:52 Phantom_Hoover [~phantomho@cpc3-sgyl21-0-0-cust116.sgyl.cable.virginmedia.com] has joined #lisp 17:12:53 ok 17:13:24 makks: symbols have useful properties when it comes to comparing them for equality. but it helps to know what those properties are and how to get the symbols you want. 17:14:35 so what I am missing is the convenience function: (defun symbol (string) (find-symbol string :keyword)) ? 17:14:51 fmu_ [u89@gateway/web/irccloud.com/x-blqgxassysloqfry] has joined #lisp 17:15:36 makks: That is a pretty bad name for something that does that. 17:16:06 makks: INTERN is another handy function that will create a symbol if it doesn't already exist. 17:16:13 Xach: what about (pathname string) ? 17:16:31 clhs pathname 17:16:31 http://www.lispworks.com/reference/HyperSpec/Body/a_pn.htm 17:16:43 is that automatic? 17:16:55 rpg [~rpg@216.243.156.16.real-time.com] has joined #lisp 17:16:58 or is sykopomp breaking my balls right now? 17:17:20 no ball-breaking. Just pointing out that there's already a symbol bound to that name. 17:17:39 makks: where do the strings come from? 17:17:45 stop it, if you dont understand what I am saying ask. 17:17:54 okay 17:18:07 Xach: filesystem hierachy 17:18:33 I am mapping paths like (:foo :bar) to /FOO/BAR 17:18:43 sounds stupid 17:18:52 seangrove [~user@c-71-198-44-87.hsd1.ca.comcast.net] has joined #lisp 17:19:14 but, actually part of a caching mechanism that I dont want to have to compare strings. 17:19:31 kraison [~kraison@70.90.182.149] has joined #lisp 17:19:47 makks: Are you mapping backwards from pathnames to your list of keywords? 17:19:52 -!- kraison [~kraison@70.90.182.149] has left #lisp 17:20:21 Xach: Yes 17:20:29 right now, I am trying to :) 17:20:55 -!- benny [~benny@i577A3BE2.versanet.de] has quit [Ping timeout: 245 seconds] 17:21:09 makks: well, you could use alexandria:make-keyword 17:21:23 Kovensky [~kovensky@abraxo.bluebottle.net.au] has joined #lisp 17:21:27 makks: are you familiar with the structured pathnames of CL? 17:22:01 Xach: yes, I think so 17:22:21 not 100% but I should have reached 60-60% by now 17:22:24 yeah... why would you store lists of keywords when you can just store relative pathnames.. pathnames are not strings in common lisp 17:22:46 -!- pdo [~pdo@217.33.254.141] has left #lisp 17:22:46 so the result of (pathname-directory "/foo/bar/baz/object.txt") isn't a revelation to you 17:22:57 REPLeffect [~REPLeffec@69.54.115.254] has joined #lisp 17:23:03 drewc: because I want to say: (fsdb-get *db* '(:foo :bar)) 17:23:47 Xach: not really 17:24:02 I use it often, but its just a different thing 17:24:49 ikki [~ikki@201.144.87.42] has joined #lisp 17:25:02 pathnames are for filesystems, my database has a similar structure but no need for most pathname stuff 17:25:19 -!- nefo [~nefo@unaffiliated/nefo] has quit [Quit: Leaving] 17:25:29 is there a common lisp package available that implements a tree ADT? 17:25:32 makks: (cons :relative (mapcar #'symbol-name '(:foo bar))) 17:25:34 edlinde: more than one 17:25:59 Xach: could you direct me to one? 17:26:00 edlinde: like the obscurely-named TREES library 17:26:06 http://www.cliki.net/TREES 17:26:11 ok 17:26:12 thanks 17:26:15 will check it out 17:26:44 drewc: what is that supposed to do? 17:27:24 Xach: do you suggest writing them by myself? 17:27:50 makks: map a list of (:foo :bar) to the pathname "FOO/BAR/" ... isn't that what you are trying to do? 17:27:51 Xach: the trees there are quite specialized 17:28:18 drewc: I already did that, right now I want to do it the other way around 17:28:27 euphidime [u178@gateway/web/irccloud.com/x-jbjtrqkimqylsswy] has joined #lisp 17:28:38 syntard_ [~chatzilla@74.112.63.251] has joined #lisp 17:28:39 edlinde: writing what by yourself? 17:28:46 the tree ADT 17:28:47 quicklisp doesn't have clsql-fluid ? 17:28:59 syntard_: No. 17:29:09 makks: (mapcar (lambda (n) (intern n :keyword)) (cdr (pathname-directory path))) 17:29:27 edlinde: Not really. Nathan Froyd already wrote something you can reuse. 17:29:41 syntard_: the build procedure for clsql-fluid is too difficult for me to handle right now. 17:30:19 Xach: just trying to use weblocks with odbc. So best way is to get fluid myself? 17:31:20 you need to run the git repo and checkout the fluid branch. 17:31:28 ok 17:31:34 syntard_: Not sure. 17:31:46 -!- REPLeffect [~REPLeffec@69.54.115.254] has quit [Ping timeout: 265 seconds] 17:32:42 nha [~prefect@imamac13.epfl.ch] has joined #lisp 17:34:58 -!- makks [~makks@p5DE8E68A.dip.t-dialin.net] has quit [Quit: makks] 17:36:02 gavv\w [~gavv@178.236.241.138] has joined #lisp 17:38:15 -!- nha [~prefect@imamac13.epfl.ch] has quit [Ping timeout: 265 seconds] 17:40:44 -!- kushal [~kdas@fedora/kushal] has quit [Ping timeout: 255 seconds] 17:40:53 -!- drforr [~drforr@pool-98-112-230-87.lsanca.fios.verizon.net] has quit [Ping timeout: 255 seconds] 17:42:48 drforr [~drforr@pool-98-112-230-87.lsanca.fios.verizon.net] has joined #lisp 17:43:07 lispmeister [u272@gateway/web/irccloud.com/x-legraimrcixoykpd] has joined #lisp 17:43:40 -!- schell [~schellsci@70-36-199-232.dsl.static.sonic.net] has quit [Quit: schell] 17:43:55 edlinde: To me, a tree is not an abstract data type, but a concrete one. There are abstract data types that can be implemented as trees, such as dictionaries and priority queues, but the only abstract data type you will find on a tree is: (make-node children) (children node) (setf (children node)) [I might be forgetting one or two]. 17:44:40 beach: ok 17:45:07 REPLeffect [~REPLeffec@69.54.115.254] has joined #lisp 17:45:13 beach: I am just planning on the operations that will be needed and accordingly thinking of some functions for it at the moment 17:45:30 edlinde: For a tree? 17:45:55 yeah just the common ones for my kind of a tree, which is like a prefix tree 17:46:04 edlinde: Then you are on the wrong track! You are most likely thinking of a special case of a tree. 17:46:07 like make-node, node-children 17:46:28 nah I am making it as generic as possible to start with 17:46:38 then I will try and do a tree traversal to see if it works 17:46:43 schell [~schellsci@70-36-199-232.dsl.static.sonic.net] has joined #lisp 17:46:52 edlinde: I am telling you; the generic version is going to have at most 5 operations. 17:47:15 edlinde: it is not worth reusing existing code, and it is not worth making it reusable. 17:47:31 ramkrsna [~ramkrsna@122.169.92.134] has joined #lisp 17:47:31 -!- ramkrsna [~ramkrsna@122.169.92.134] has quit [Changing host] 17:47:31 ramkrsna [~ramkrsna@unaffiliated/ramkrsna] has joined #lisp 17:47:31 edlinde: Think of the abstract data type that you want to implement instead. 17:48:04 ok 17:48:14 edlinde: Or think of the concrete one, if you are into speed, but don't try to make it abstract then. 17:48:26 Ok, I made a fluid branch of clsql, can I sort of nudge it into my quicklisp installs? 17:48:56 edlinde: You could have implemented the operations on the "tree" ADT in the time that you spent asking about it here. 17:49:01 -!- slyrus [~chatzilla@adsl-75-55-215-234.dsl.pltn13.sbcglobal.net] has quit [Remote host closed the connection] 17:49:31 -!- trigen [~MSX@87.209.144.213] has quit [Ping timeout: 240 seconds] 17:49:39 -!- REPLeffect [~REPLeffec@69.54.115.254] has quit [Ping timeout: 265 seconds] 17:50:02 -!- Kovensky [~kovensky@abraxo.bluebottle.net.au] has quit [Ping timeout: 272 seconds] 17:50:22 crudely, I could just replace the contents of clsql-20101107-git with my branch, is there a subtler way to substitute? 17:50:27 Kovensky [~kovensky@abraxo.bluebottle.net.au] has joined #lisp 17:50:49 edlinde: It's not very far from asking about a generic library for managing conses: cons, car, cdr, that's all there is to it. 17:51:00 stassats [~stassats@wikipedia/stassats] has joined #lisp 17:51:25 -!- whee_ [~whee@smaertness.net] has quit [Read error: Operation timed out] 17:52:05 Shaftoe [~Moe111@bas1-montreal02-1096728010.dsl.bell.ca] has joined #lisp 17:52:07 -!- clop [~jared@moat3.centtech.com] has quit [Ping timeout: 240 seconds] 17:52:10 Hi all. 17:52:17 hello Shaftoe 17:52:29 how goes the battle for lowered entropy? 17:52:32 whee [~whee@cpe-69-207-148-170.rochester.res.rr.com] has joined #lisp 17:52:44 jdz [~jdz@host106-104-dynamic.8-87-r.retail.telecomitalia.it] has joined #lisp 17:53:11 kmwallio [~kmwallio@host113-70.brownlie.fit.edu] has joined #lisp 17:53:19 I'm working with html-template, and have an idea for a semantic extension of the language... thought I'd confer with the community about this. 17:53:38 anyone familiar or interested to listen: I welcome your comments. 17:54:25 so the idea is this: variable tags are empty pseudo-entities, like so 17:54:37 I'd like to add an extension to make it have content, like so: 17:54:54 bar baz 17:55:19 the idea would be that it would be semantically equivalent to an if statement. 17:55:52 so the web designer could put code like so: Fax: $ 17:56:08 and if fax is nil, nothing gets displayed, if fax is not nill, "Fax: " gets displayed. 17:56:13 what think you all? 17:57:25 LaPingvino [~LaPingvin@187.87.224.70] has joined #lisp 17:58:00 An alternative and equally useful feature would be to do Lorem ipsum sit dolor ... which would display lorem ipsum text if foo is empty. 17:58:26 I realize that the two forms I've just shown are ambiguous, but the use case is the point I'm drawing attention to 17:58:32 -!- hugod [~hugod@bas1-montreal50-1279440432.dsl.bell.ca] has quit [Quit: hugod] 17:59:20 -!- rme [~rme@pool-70-105-122-27.chi.dsl-w.verizon.net] has left #lisp 18:03:25 Shaftoe: I'm not familiar with html-template, do you have a link? 18:03:50 That doesn't look like a very nice syntax - it seems like it's trying to be a kind of limited conditional 18:03:58 -!- jdz [~jdz@host106-104-dynamic.8-87-r.retail.telecomitalia.it] has quit [Ping timeout: 272 seconds] 18:05:12 clop [~jared@moat3.centtech.com] has joined #lisp 18:07:59 kushal [~kdas@fedora/kushal] has joined #lisp 18:10:27 -!- jajcloz [~jaj@pool-108-7-68-199.bstnma.fios.verizon.net] has quit [Quit: jajcloz] 18:12:46 trigen [~MSX@87.209.144.213] has joined #lisp 18:13:04 -!- V-ille [~ville@ville-hp.dhcp.fnal.gov] has quit [Read error: Operation timed out] 18:13:29 syntard_: no need to replace 18:13:44 syntard_: you could put your clsql in ~/src/clsql/ and add that directory to your asdf:*central-registry* 18:13:49 syntard_: that will shadow any systems in quicklisp 18:13:55 -!- Phantom_Hoover [~phantomho@cpc3-sgyl21-0-0-cust116.sgyl.cable.virginmedia.com] has quit [Ping timeout: 240 seconds] 18:14:08 oh, ok 18:16:36 Phantom_Hoover [~phantomho@cpc3-sgyl21-0-0-cust116.sgyl.cable.virginmedia.com] has joined #lisp 18:18:17 devinus [~devinus@65.107.181.222.ptr.us.xo.net] has joined #lisp 18:19:43 -!- phrixos [~miar1@bs0176.nerc-bas.ac.uk] has quit [Ping timeout: 245 seconds] 18:21:08 phrixos [~miar1@bs0176.nerc-bas.ac.uk] has joined #lisp 18:24:44 V-ille [~ville@ville-hp.dhcp.fnal.gov] has joined #lisp 18:26:00 -!- schell [~schellsci@70-36-199-232.dsl.static.sonic.net] has quit [Quit: schell] 18:26:44 seangrove: it is a limited conditional. 18:27:03 however, the alternative to what I propose is much more verbose and prone to mistakes when the designer uses it. 18:27:04 -!- ignas [~ignas@78-60-36-123.static.zebra.lt] has quit [Ping timeout: 265 seconds] 18:27:48 Shaftoe: maybe you want TAL? comes with yaclml, and is extensible. 18:28:15 -!- delYsid` is now known as delYsid 18:28:26 -!- phrixos [~miar1@bs0176.nerc-bas.ac.uk] has quit [Ping timeout: 276 seconds] 18:28:27 -!- delYsid [~user@chello084115136207.3.graz.surfer.at] has quit [Changing host] 18:28:27 delYsid [~user@debian/developer/mlang] has joined #lisp 18:28:37 Hexstream [~hexstream@modemcable075.97-200-24.mc.videotron.ca] has joined #lisp 18:28:42 drewc: you got a link for TAL? 18:29:06 minion: yaclml 18:29:09 phrixos [~miar1@bs0176.nerc-bas.ac.uk] has joined #lisp 18:29:16 ??? 18:29:21 What's the state of support for the MOP in the various implementations? Are there any "reasonable", "notable" implementations that don't support the MOP? 18:29:29 drewc: although, to answer you, this extension I'm thinking of is in response to our system being used by designers. 18:29:46 ak70 [~ak70@46.11.10.227] has joined #lisp 18:29:52 drewc: so there's already a bunch of usage going on and changing the html templates themselves would be counter productive. 18:30:08 Shaftoe: http://common-lisp.net/project/bese/docs/yaclml/TAL_0020-_0020Dynamic_0020HTML_0020Templating.html 18:30:18 drewc: thanks 18:30:30 and (ql:quickload "yaclml") 18:30:50 drewc: still liking yaclml over other things? 18:31:08 Hexstream: abcl's mop is not complete 18:31:34 but that's considered a bug, not a feature. 18:31:46 most implementations don't adhere 100% to the specification from AMOP 18:32:03 closer-mop is the portability layer to use 18:32:08 But I guess if more libraries relied on it then it would be supported better ;) 18:32:12 sykopomp: yup 18:32:13 Hexstream: and, if people want MOP, we're happy to accept patches to ABCL. 18:32:20 Hexstream: many libraries rely on closer-mop 18:32:20 drewc: Actually, abcl does consider a bit a bug. 18:32:37 easyE: which is what i said, non? 18:32:38 slyrus [~chatzilla@adsl-75-55-215-234.dsl.pltn13.sbcglobal.net] has joined #lisp 18:32:43 CLeaning things up is "in progress", but as ehu sez, we'd welcome elp. 18:32:51 drewc: Yes. 18:33:00 -!- urandom__ [~user@p548A2E28.dip.t-dialin.net] has quit [Ping timeout: 245 seconds] 18:33:13 Just trying to explain we (abcl community) isn't happy with our current state of affairs. 18:33:23 -!- milanj [~milanj_@93-86-57-180.dynamic.isp.telekom.rs] has quit [Ping timeout: 255 seconds] 18:33:35 Hexstream: the MOP is quite usable across implementations, especially for the more basic things. It's only the darker corners where things get messy 18:33:43 easyE: of course :) 18:33:51 Ok, so that's awesome. Because I have at least one library that I just realized could greatly benefit from a "reimplementation" in terms of the MOP and I was wondering if I could just go all out and drop support for the "old" implementation that didn't require the MOP. Apparently the answer is yes :) 18:34:07 MOPless? 18:34:11 lol 18:34:17 lol. 18:34:36 type-maker, which is the next library I'm cleaning up + documentating. 18:34:59 What I really want to release is linkval, but it relies on type-maker and I'm a bit of a depth-first guy ;P 18:35:17 -!- BrandLeeJones [~BrandLeeJ@chello062178064156.22.11.vie.surfer.at] has quit [Quit: BrandLeeJones] 18:35:40 And I really want to be able to say linkval is "comprehensively documented", and it's cheating a bit if the transitive closure of its dependencies aren't all themselves comprehensively documented. 18:36:03 documenting* 18:36:31 -!- PissedNu1lock is now known as PissedNumlock 18:36:46 dreamcode [~Wistful@unaffiliated/wistful] has joined #lisp 18:37:05 Hexstream: I don't think you can do this "using the MOP," but you /can/ using Closer-MOP. 18:37:26 Yeah, well that's what I mean. 18:37:28 Using the MOP directly is a real nuisance, if only because of the differences in package names across implementations. 18:37:59 Hexstream: Oh. Yeah, then. Closer-MOP was a huge service. Hats off to Pascal. 18:38:13 milanj [~milanj_@93-87-167-170.dynamic.isp.telekom.rs] has joined #lisp 18:38:27 Zephyrus [~emanuele@unaffiliated/zephyrus] has joined #lisp 18:39:34 drewc: any reason why yaclml isn't listed anywhere in the projects? 18:40:00 drewc: scratch that. 18:41:19 -!- Zephyrus [~emanuele@unaffiliated/zephyrus] has quit [Client Quit] 18:45:36 Thanks for the info ;) 18:45:37 -!- Hexstream [~hexstream@modemcable075.97-200-24.mc.videotron.ca] has left #lisp 18:45:37 -!- _s1gma [~herpderp@77.107.164.131] has quit [Quit: format t "motherfuckers"] 18:48:02 petercoulton [~petercoul@cpc2-midd16-2-0-cust169.11-1.cable.virginmedia.com] has joined #lisp 18:49:35 -!- setheus [~setheus@cpe-70-116-140-134.tx.res.rr.com] has quit [Quit: leaving] 18:51:44 I need a consult .... 18:52:10 I need a consultant... 18:52:16 I've written a package that implements Google's protocol buffers. 18:52:22 -!- CTP``` [~user@talula.plus.com] has quit [Remote host closed the connection] 18:52:37 reb`: yes! 18:52:44 -!- dreamcode [~Wistful@unaffiliated/wistful] has quit [Quit: Leaving] 18:53:09 Currently, when someone defines a protocol buffer, the definition file gets translated into Lisp and all the code for all the different kinds of protocol buffers lives in the same package PROTOCOL-BUFFER. 18:53:32 benny` [~benny@i577A2DCB.versanet.de] has joined #lisp 18:53:55 I'm considering changing things so that each proto definition file can indicate what package its code should live in. 18:54:37 superflit [~superflit@140.226.49.160] has joined #lisp 18:55:23 So a protocol buffer definition file called people.proto might cause its code to live in package PEOPLE. 18:55:32 reb`: What is a "protocol buffer", a "definition file", and a "proto definition file"? 18:55:57 It's a scheme for encoding data compactly. 18:56:17 beach: protocol buffers is an IPC system from Google 18:56:28 the other two things are part of it 18:56:29 You define "messages" (a.k.a. protocol buffers) in .proto definition files. 18:56:55 They are translated into C++, Java, Python classes with fast encoders and decoders. 18:57:14 Google uses them for data storage and RPC, because they are very compact. 18:57:23 (the wire format is compact) 18:57:24 -!- echo-area [~user@114.251.86.0] has quit [Read error: Connection reset by peer] 18:58:11 *beach* didn't realize that knowing about a google IPC system was a requirement for participating in #lisp 18:58:27 Here is an example .proto file: 18:58:29 https://github.com/brown/protobuf/blob/master/example/addressbook.proto 18:58:44 *beach* thus vanishes to pursue other activities 18:59:31 -!- kushal [~kdas@fedora/kushal] has quit [Ping timeout: 240 seconds] 19:00:25 The question is: What package should contain the machine-generated Lisp code that implements a particular protocol buffer message? 19:02:00 reb`: is it feasible to specify it at generation time? e.g. (translate "foo.proto" "foo.lisp" :package :foo)? 19:02:36 rickmode [~rickmode@cpe-75-85-93-10.socal.res.rr.com] has joined #lisp 19:02:38 urandom__ [~user@p548A689B.dip.t-dialin.net] has joined #lisp 19:02:46 yes 19:02:50 I get There is no package named "NIL" when I try to load clsql-odbc 19:03:28 (INTERN "ATOL64" NIL) 19:03:40 elliot johnson recently wrote about that 19:03:43 Then if I have .proto files people.proto and dogs.proto that define Person and Dog messages, the generated code might live in packages PEOPLE and DOGS. 19:04:10 So to access the name field of either one you'd need PEOPLE:NAME and DOGS:NAME. 19:04:12 syntard_: http://elliottjohnson.net/blog/?p=182 19:04:46 ... I'm leaning toward this now, but my initial implementation placed all the field accessor code in one package. 19:05:05 ... so you could access the NAME field with PB:NAME. 19:05:21 ... and the NAME function is generic. 19:05:50 If separate packages are used, then the accessors don't have to be generic, but users have to get the package right. 19:06:24 *syntard_* considers whining like a little girl 19:07:29 reb`: generic sounds like a good idea and package choice (optionally the same package) sounds like a good idea 19:07:30 (cl:defpackage :nick (:export "NAME")) ? 19:07:55 -!- benny` is now known as benny 19:09:34 nyef: Yes, you can export the field accessor functions. 19:11:02 Maybe functions common to all generated classes, such as OCTET-SIZE should live in one package, while field names and message classes should live where the user wants them. 19:11:07 adf [~user@188.149.41.215] has joined #lisp 19:13:11 Jabberwockey [~Jens@f050066129.adsl.alicedsl.de] has joined #lisp 19:13:23 -!- lanthan_afh [~ze@p50992b91.dip0.t-ipconnect.de] has quit [Quit: Ex-Chat] 19:13:28 -!- petercoulton [~petercoul@cpc2-midd16-2-0-cust169.11-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 19:14:02 schoppenhauer [~christoph@unaffiliated/schoppenhauer] has joined #lisp 19:14:43 -!- ramkrsna [~ramkrsna@unaffiliated/ramkrsna] has quit [Ping timeout: 240 seconds] 19:14:50 Xach: Generic accessors allow users to subcass the generated code. Non-generic functions appear to have better type checking of arguments and are probably faster. 19:14:53 -!- schoppenhauer [~christoph@unaffiliated/schoppenhauer] has quit [Client Quit] 19:15:17 vroufe [~omx@ppp91-77-243-60.pppoe.mtu-net.ru] has joined #lisp 19:15:22 More intended to play on NICK:NAME than anything else, really. 19:16:15 -!- adf [~user@188.149.41.215] has quit [Quit: leaving] 19:19:30 nowl [~nowl@c-71-233-2-216.hsd1.ma.comcast.net] has joined #lisp 19:20:12 -!- redline6561 [~redline@c-66-56-55-169.hsd1.ga.comcast.net] has quit [Remote host closed the connection] 19:21:16 redline6561 [~redline@c-66-56-55-169.hsd1.ga.comcast.net] has joined #lisp 19:21:59 -!- Saturnation [~dsouth@72.95.93.7] has quit [Ping timeout: 255 seconds] 19:23:47 -!- churib [~tg@dslb-088-071-150-055.pools.arcor-ip.net] has quit [Ping timeout: 255 seconds] 19:24:30 -!- b-man_ [~b-man@189.34.49.137] has quit [Remote host closed the connection] 19:32:55 Hi can I please get some help with this stack overflow error - I am trying to read in space separated strings into a tree. there seems to be a problem with my function on line 26 - fp-insert http://paste.lisp.org/display/116551 19:33:25 edlinde: You should paste the error you get, too, and perhaps a backtrace. 19:33:31 ok 19:33:57 Xach: on slime is there an easy way to get out the backtrace in format that it expands all the node data? 19:34:18 edlinde: I don't know. 19:34:45 http://paste.lisp.org/display/116551#1 19:35:09 the thing is that in my function, if the list is nil I don't want to do anything and just return 19:35:33 Saturnation [~dsouth@71.169.180.213] has joined #lisp 19:35:49 and later on I have to add in a check saying that if the new node I am trying to exist already exists in the tree then just update the count of that node, don't insert.... if not then create a new node 19:36:50 edlinde: What does (split-sequence:split-sequence #\Space nil) do? 19:36:55 nyef beat me to it. 19:37:29 CL-USER> (split-sequence:SPLIT-SEQUENCE #\Space "A stitch in time saves nine.") 19:37:29 ("A" "stitch" "in" "time" "saves" "nine.") 19:38:05 edlinde: And (split-sequence:split-sequence #\Space '("stitch" "in" "time" "saves" "nine")) ? 19:38:07 so the string I read in with read-line gets converted to a list of the strings i need to insert into my tree... 19:38:09 edlinde: That was a fucking annoying misreading of the question. 19:38:20 edlinde: Why don't you answer nyef's question instead of what you want to answer? 19:38:39 i pasted what it does 19:38:43 what do you want me to say? 19:38:50 edlinde: I asked about a specific boundary condition. 19:38:59 smanek [~smanek@dsl081-028-042.sfo1.dsl.speakeasy.net] has joined #lisp 19:38:59 ah sorry 19:39:01 didn't see nil 19:39:04 my bad 19:39:21 edlinde: Try harder. 19:39:23 -!- rread [~rread@c-24-5-127-116.hsd1.ca.comcast.net] has quit [Quit: rread] 19:39:28 returns (nil) 19:39:35 *drewc* blows the whistle, shows edlinde a yellow card, and motions for a direct free kick. 19:39:49 and actually its showing a "0" too 19:39:55 edlinde: And is (nil) null? 19:39:55 edlinde: Do you see why (nil) is a problem? 19:39:56 -!- rntz [~rntz@ASPHODEL.RES.CMU.EDU] has left #lisp 19:39:57 so it seems to return the number of items 19:40:21 -!- Saturnation [~dsouth@71.169.180.213] has quit [Client Quit] 19:40:43 hm yeah I think I know what you mean.. just not sure how to get around it at the moment 19:40:58 My second question was also germane. 19:40:59 its still in list context yeah 19:41:05 edlinde: You might change when you test for a nil. 19:41:22 Since it covers what happens on the /second/ iteration. 19:42:10 sorry guys you lost me there 19:42:25 how do I change "when" to test for a nil? 19:42:29 ejohnson [~elliott@elliottjohnson.net] has joined #lisp 19:42:58 You could check for NIL before calling split-sequence. 19:43:12 hmm 19:43:16 But I get the distinct impression that there's more wrong with fp-insert than just the infinite recursion. 19:43:32 could be :) 19:43:55 So, (split-sequence:split-sequence #\Space '("saves" "nine.")) => what? 19:44:01 pchrist_ [~spirit@gentoo/developer/pchrist] has joined #lisp 19:44:27 CL-USER> (split-sequence:split-sequence #\Space '("saves" "nine.")) 19:44:27 (("saves" "nine.")) 19:44:28 2 19:44:51 the thing is that "path" will always be a string like "a b c d e" or something 19:45:04 No, it won't! 19:45:09 so I wanted to first convert this string vector to a list which I could process 19:45:22 That's a good start, but your intuition is wrong. 19:45:24 nyef: ah crap.. I think I get you 19:45:38 its the recursive call that will stuff up :( 19:46:19 -!- pchrist [~spirit@gentoo/developer/pchrist] has quit [Ping timeout: 240 seconds] 19:46:22 maybe its best I just do this split sequence before even calling fp-insert and then pass it the list yeah? 19:46:31 nyef: does that sound like a better plan? 19:46:49 Cowmoo [~Cowmoo@cambridge-vxty.basistech.com] has joined #lisp 19:47:07 Actually, I'd (defun fp-insert (...) (flet ((build-node-structure (path-list parent) ...)) (build-node-structure (split-sequence ...) nil))). 19:48:04 Umm... And I'd be obscenely clever and arrange for the node structure to be tail-recursive, but that's for the advanced class. 19:48:43 nyef: I am still a newbie , started reading cl a week or so ago 19:49:01 I understand that flet lets me declare a local function 19:49:18 but I am not super sure I get what you are doing with the "build-node-structure" 19:49:35 -!- Phantom_Hoover [~phantomho@cpc3-sgyl21-0-0-cust116.sgyl.cable.virginmedia.com] has left #lisp 19:49:40 would this work? (defmacro named-lambda (name args &body body) `(flet ((,name ,args ,@body)) #',name)) 19:49:43 is that a wrapper around "make-node" for the struct? 19:50:45 another quick one, if I have two forms and I wanted to make a new temp variable in the first form and use this in the other form... do I have to wrap them both in a let? 19:50:49 edlinde: Yeah, it's a wrapper around the make-node and... Oh. It has to be LABELS, not FLET, because it needs to be the recursive bit. 19:50:52 or is there some better construct? 19:51:26 You would probably need a LET in some form, yes. 19:51:41 hmm ok 19:51:55 -!- vroufe [~omx@ppp91-77-243-60.pppoe.mtu-net.ru] has quit [Ping timeout: 240 seconds] 19:52:08 sentry [~sentry@ec2-72-44-49-66.compute-1.amazonaws.com] has joined #lisp 19:52:39 Umm... You should probably be using defclass instead of defstruct, defclass is a lot easier to work with during development. 19:52:40 pers [~user@221.sub-75-196-166.myvzw.com] has joined #lisp 19:53:29 what is the reason for structs being so picky about being redefined? 19:53:48 Guthur: The accessors and typechecks and whatnot are permitted to be inlined. 19:54:10 nyef: ok 19:54:12 ah, fair enough 19:55:02 And there's still something wrong with fp-insert: it doesn't actually modify the parent node or check for existing nodes with the same DATA slot. 19:55:06 -!- sentry [~sentry@ec2-72-44-49-66.compute-1.amazonaws.com] has quit [Read error: Connection reset by peer] 19:56:30 sentry [~sentry@ec2-72-44-49-66.compute-1.amazonaws.com] has joined #lisp 19:57:27 -!- katesmith [~katesmith@unaffiliated/costume] has quit [Ping timeout: 240 seconds] 19:57:39 katesmith_ [~katesmith@unaffiliated/costume] has joined #lisp 19:58:47 thanks Xach and nyef 19:59:10 nyef: you mentioned I should read up about using defclass instead of defstruct? is there a special reason why? 19:59:34 edlinde: defclass is friendlier for interactive development. 19:59:41 nyef: I liked the defstruct because it automatically offered me the make-struct function + all the accessor fns 19:59:55 ziarkaen [~ziarkaen@stu301.queens.ox.ac.uk] has joined #lisp 20:00:17 sykopomp: as in I can see values within the class while debugging etc? 20:00:22 petercoulton [~petercoul@cpc2-midd16-2-0-cust169.11-1.cable.virginmedia.com] has joined #lisp 20:00:30 you can change the slots in a class in a running image 20:00:54 -!- smanek [~smanek@dsl081-028-042.sfo1.dsl.speakeasy.net] has quit [Quit: This computer has gone to sleep] 20:01:07 ok will check it out 20:01:10 edlinde: if you need all the accessors by default, you're doing it wrong anyway. 20:01:12 thanks for the tip 20:01:16 or at least most likely 20:01:29 -!- Guthur [~Guthur@cpc11-belf9-2-0-cust855.2-1.cable.virginmedia.com] has quit [Quit: Leaving] 20:01:36 sykopomp: huh? 20:01:38 sykopomp: why do you say that? 20:01:55 why shouldn't I be able to get to the slots in my struct? 20:02:00 sykopomp: that's a controversial statement :p I'd rather not be using slot-value for everything 20:02:21 because you should be writing code around interfaces, not crapping up your package with a bunch of auto-generated genfuns. 20:02:50 dlowe: I'm not advocating for spamming slot-value either. I'm saying that the issue becomes more or less moot when you start writing your programs around protocols. 20:02:54 ok i see what you mean, like add my own layer of abstraction 20:03:13 i shall do that later on as I go along 20:03:13 defclass with :reader/:writer/:accessor becomes more a matter of implementing a protocol, not a "I have this slot, I should have this accessor" 20:03:14 christoph_debian [christoph@cl-1281.dus-01.de.sixxs.net] has joined #lisp 20:03:21 which is what defclass* does (shudder) 20:03:47 sykopomp: how do you presume to implement this protocol without accessor functions? 20:04:01 -!- scgilardi [~scgilardi@c-71-192-85-172.hsd1.ma.comcast.net] has quit [Quit: ZNC - http://znc.sourceforge.net] 20:04:26 Bronsa [~bronsa@host165-180-dynamic.4-87-r.retail.telecomitalia.it] has joined #lisp 20:04:41 I find that I tend to use SLOT-VALUE quite a bit when I work with CLOS. 20:04:56 dlowe: I'm not saying to not use :accessor. I'm saying that I found its use to be much less common once I actually started with genfuns. 20:04:57 that seems not so clear 20:05:22 and I'm using :reader and :writer a lot more. 20:05:39 Ragnaroek [~chatzilla@pD951ED35.dip.t-dialin.net] has joined #lisp 20:06:17 it's a different kind of design. I think defclass*-style blanket-definitions are the wrong way to go. 20:06:47 and I agree that :accessor should at least warn if the genfun doesn't already exist. 20:06:57 -!- edlinde [~edlinde@90-227-7-243-no15.tbcn.telia.com] has quit [Quit: edlinde] 20:12:43 -!- pers [~user@221.sub-75-196-166.myvzw.com] has left #lisp 20:13:26 Hmm, does lispworks support arrays with element-type nil? 20:13:29 specialized that is? 20:14:41 pjb: can you paste a clall of (type-of (make-array 0 :element-type nil))? 20:15:19 schell [~schellsci@70-36-199-232.dsl.static.sonic.net] has joined #lisp 20:15:36 -!- ehu [~ehuels@ip118-64-212-87.adsl2.static.versatel.nl] has quit [] 20:15:51 hi all 20:15:53 fisxoj [~fisxoj@cpe-24-59-205-231.twcny.res.rr.com] has joined #lisp 20:16:03 ehu [~ehuels@ip118-64-212-87.adsl2.static.versatel.nl] has joined #lisp 20:16:41 scgilardi [~scgilardi@c-71-192-85-172.hsd1.ma.comcast.net] has joined #lisp 20:18:29 anybody had luck using ccl and clsql-odbc on windows? 20:19:40 smanek [~smanek@206.111.142.135.ptr.us.xo.net] has joined #lisp 20:24:39 redline6561: that tweet of yours got me to publish zpb-aws even though it's nowhere near done. 20:25:02 Xach: Whoa, whoa. Hang on now. It was on github *when* you tweeted it. I didn't do that. ;) 20:25:07 -!- HG` [~HG@xdsl-92-252-105-72.dip.osnanet.de] has quit [Quit: Leaving.] 20:25:27 redline6561: I only put it up last night! Your tweet was days ago. Or a day ago. Or something. 20:26:01 Xach: Huh...my system clock got screwed up at some point recently. That might be why. Sorry about that. I can try to be more discreet about retweets if you like. 20:26:57 redline6561: No, i mean your tweet about build something & release it! 20:27:09 varjag [~eugene@4.169.249.62.customer.cdi.no] has joined #lisp 20:27:12 I have a lot of half-finished things I need to push out and share. 20:27:27 Xach: Oh! Wow. I'm really flattered. Just wrote a blog article about that and posted it to lisp reddit actually. 20:27:41 Xach: How funny. I wrote that largely in response to reading ILC Part 04. 20:28:02 Dang it, *I* wanted the karma. 20:28:04 *redline6561* marvels at the irony 20:28:10 *Xach* will have to submit it to HN or something instead 20:28:24 Xach: Hahaha. Either you're really watching lisp reddit or have insane timing. But sure, throw it up! :D 20:28:49 redline6561: google reader to the rescue 20:29:36 Xach: Indeed. Now to do write up my first Weblocks introductory app... it's a long day of blogging. On the bright side I get to be useful late tonight or tomorrow and write more code... 20:31:35 *Xach* was working on a project to make the Naggum usenet search system generic earlier this year when he got derailed by quicklisp 20:31:56 -!- fisxoj [~fisxoj@cpe-24-59-205-231.twcny.res.rr.com] has quit [Ping timeout: 276 seconds] 20:31:56 *Xach* must finish it enough to be useful and publish it, too, and get KMP and RPW's archives up and searchable 20:32:40 Man, I'm such a young lisper. Kent Pitman and who? 20:32:59 -!- morphling [~stefan@gssn-5f757ffa.pool.mediaWays.net] has quit [Remote host closed the connection] 20:33:00 Rob Warnock 20:33:53 http://xach.livejournal.com/139017.html has some links to articles of his 20:33:55 not nearly enough 20:35:12 Awesome. One of my professors actually encouraged me to get in touch with him... 20:35:18 -!- Bronsa [~bronsa@host165-180-dynamic.4-87-r.retail.telecomitalia.it] has quit [Quit: Lost terminal] 20:35:22 Apparently they were college roommates. 20:35:30 Or something like that.. 20:35:39 Small world. 20:36:07 And getting smaller all the time. 20:40:16 Xach: Wow. Using FORMAT for C data tables is awesome and nutty and terrifying all at once. Good articles as usual thouhg. Thanks. 20:40:21 *though 20:42:32 GrayMagiker [~steve@c-174-56-26-27.hsd1.nm.comcast.net] has joined #lisp 20:43:48 *Xach* has to update the buildapp documentation to mention Quicklisp 20:45:00 vroufe [~omx@ppp94-29-9-58.pppoe.spdop.ru] has joined #lisp 20:45:33 *redline6561* needs to update his CL Web Primer Pt. 1 to use quicklisp... 20:45:38 but not right now... 20:46:06 redline6561: where is your blog? 20:46:41 Fade: http://redlinernotes.com/blog/ but it gets mirrored at http://redline6561.livejournal.com/ 20:51:06 <[df]> 'common lisp is a programmer amplifier' - I like it :) 20:51:32 i just read that 20:51:39 you typed it as i read it 20:52:08 Marc Feeley had a cute jab at the ball of mud thing 20:52:22 Scheme is like a ball of snow. You can add any amount of snow to it and it still looks like a ball of snow. 20:52:26 Moreover, snow is cleaner than mud. 20:53:00 Snow turns to brown slushy crap after a week or two. 20:53:01 ("snow" is the name of his Scheme project that does something similar to quicklisp) 20:53:05 Ha. 20:53:37 he has some nice wordplay, too. the archives are called "snowballs" and the central repo is the "snowfort" 20:54:01 I wonder if Mudballs took any inspiration from it. 20:54:27 Xach: When I had my "is there an acceptable scheme" research phase, Snow looked really cool and I had high hopes for gambit. Doesn't seem to have worked. :( 20:54:38 boiantz [~boiantz@92-247-214-154.spectrumnet.bg] has joined #lisp 20:55:20 redline6561: when i talked to him at ILC, he lamented he had very little uptake. everyone wanted to do their own thing. 20:55:20 Xach: The libraries are still balkanized by implementation but Chicken and Racket/PLT both have fairly nice collections. But these days I'm way too lost in slime and quicklisp to notice. :P 20:55:43 Xach: Yeah, it really is sad. It *seemed* when I looked that he'd thought through most of the issues. I guess that's scheme for you? 20:55:49 he worked hard to make it work across 16 or so implementations 20:56:01 Xach: Exactly. 20:56:28 -!- Faed is now known as Fade 20:57:21 weblocks question: can I use simple clsql-odbc instead of clsql-fluid? 20:57:24 wbooze` [~user@xdsl-87-79-140-119.netcologne.de] has joined #lisp 20:57:29 16 implementations? 20:57:31 wow 20:58:13 -!- wbooze` [~user@xdsl-87-79-140-119.netcologne.de] has quit [Read error: Connection reset by peer] 20:58:31 holycow: i might have the number wrong, but it was a lot of different implementations. 20:58:48 anywhere up there is pretty amazing 20:59:23 <[df]> yeah, I've grumbled about having to make things work with 4 CL implementations 20:59:25 syntard_: Give it a shot. 20:59:27 -!- homie [~user@xdsl-195-14-197-168.netcologne.de] has quit [Ping timeout: 240 seconds] 20:59:46 syntard_: I haven't tried the clsql parts of weblocks yet and I haven't tried it on Windows. I doubt many people have, really. 21:00:02 -!- wbooze [~user@xdsl-195-14-197-168.netcologne.de] has quit [Ping timeout: 264 seconds] 21:00:07 -!- freiksenet [~freiksene@cs181130165.pp.htv.fi] has quit [Ping timeout: 240 seconds] 21:00:09 it would have been pretty hard to try it in the dark BQ days... 21:00:10 syntard_: But if you get weird errors, paste them and ask here or in #weblocks. I'll do my best to help. 21:00:34 -!- kmwallio [~kmwallio@host113-70.brownlie.fit.edu] has quit [Remote host closed the connection] 21:00:34 <[df]> plus I guess CL has more standardised stuff than scheme 21:00:58 [df]: Definitely. 21:01:16 has anyone played around with ecl on iOS? 21:01:37 apples phone OS? 21:01:42 yes 21:02:01 I don't even know how you'd get a lisp onto an iphone. is it possible? 21:02:18 I saw CCL on a jailbroken iPad. Not quite the same thing. 21:02:19 s0ber_ [~s0ber@111-240-215-136.dynamic.hinet.net] has joined #lisp 21:02:19 patched version of ecl, patched version of slime: https://github.com/kriyative/ecl-iphone-builder 21:02:24 no jailbreak 21:02:31 though i can't get a repl 21:02:38 -!- petercoulton [~petercoul@cpc2-midd16-2-0-cust169.11-1.cable.virginmedia.com] has quit [Ping timeout: 245 seconds] 21:02:55 i've bought my last apple hardware, I think. :) 21:02:58 i think it might be because the slime versions (in emacs and on the iphone) are different, was wondering what your take would be on it 21:03:41 -!- vroufe [~omx@ppp94-29-9-58.pppoe.spdop.ru] has quit [Ping timeout: 255 seconds] 21:03:48 Fade: This isn't CL, but: http://jlongster.com/blog/2009/06/17/write-apps-iphone-scheme/ 21:04:26 -!- s0ber [~s0ber@111-240-205-159.dynamic.hinet.net] has quit [Ping timeout: 265 seconds] 21:04:36 -!- s0ber_ is now known as s0ber 21:05:12 vroufe [~omx@ppp94-29-9-58.pppoe.spdop.ru] has joined #lisp 21:05:50 -!- symbole [~user@216.214.176.130] has quit [Remote host closed the connection] 21:06:38 -!- seangrove [~user@c-71-198-44-87.hsd1.ca.comcast.net] has quit [Ping timeout: 264 seconds] 21:07:00 -!- sonnym [~evissecer@singlebrookvpn.lightlink.com] has quit [Quit: Leaving.] 21:07:14 redline6561: clsql-demo needs clsql-fluid 21:07:15 fogus_ [c6970d0f@gateway/web/freenode/ip.198.151.13.15] has joined #lisp 21:08:04 SpitfireWP_ [~Spitfire@wikipedia/spitfire] has joined #lisp 21:08:35 -!- dlowe [~dlowe@ita4fw1.itasoftware.com] has quit [Quit: Leaving.] 21:08:58 syntard_: Hmm...give me a moment to pull the source up. 21:09:07 fogus_: Hey there. Thanks for the HN post! :) 21:09:29 setheus [~setheus@cpe-70-116-140-134.tx.res.rr.com] has joined #lisp 21:09:39 redline6561: My pleasure. Thank you for writing the post. I enjoyed it very much. 21:09:51 -!- SpitfireWP [~Spitfire@wikipedia/spitfire] has quit [Disconnected by services] 21:09:55 -!- SpitfireWP_ is now known as SpitfireWP 21:10:07 fogus_: Absolutely. Glad you did! I was a bit nervous. Didn't want to try and speak for anybody but so far this reception is a relief. :) 21:10:11 redline6561: I was hoping you would be here so I could commend you "in person". 21:10:50 fogus_: No kidding? Man. Thanks. I'm flattered. Oh, hey! I just picked up "The Joy of Clojure" yesterday, by the way. 21:11:05 *fogus_* blushes 21:11:13 -!- boiantz [~boiantz@92-247-214-154.spectrumnet.bg] has left #lisp 21:11:14 I hope you like it 21:11:38 fogus_: If I can't get a consultancy going in CL before I graduate, I would be happy to be able to write Clojure somewhere as an alternative. I'll keep you posted. ;) 21:11:55 Deal 21:12:19 redline6561: I'll try to make fluid work I guess 21:12:24 edlinde [~edlinde@90-227-7-243-no15.tbcn.telia.com] has joined #lisp 21:12:24 chiguire [~chiguire@gentoo/developer/chiguire] has joined #lisp 21:12:25 syntard_: Did you get weblocks through quicklisp? And where is this *.asd? 21:12:53 syntard_: AH. I see. It's not a part of weblocks. It's clsql itself. 21:12:59 the clsql-fluid thing vexes me a bit 21:13:06 syntard_: Sorry. I've never used clsql myself. 21:13:09 Xach: What is it? 21:13:23 if you follow the directions here, fluid worked, at least it did when I tried it last month: http://common-lisp.net/project/clsql-fluid/ 21:13:49 it isn't clear to me why the fluid thing isn't merged into clsql. it vexes me, too. 21:13:52 redline6561: it's something weblocks uses. to use it, you have to check out clsql, then add a new git repo to it and check out a particular branch from *that* 21:13:55 syntard_: See above ^^ I'm hoping to clean up a postmodern backend for weblocks that never got merged and get it in by Xmas. If not much sooner. 21:14:03 been there 21:14:10 syntard_: Got it. 21:14:15 if it was part of clsql it would be easy to include, or if it was fully separate it would be easy to include 21:14:19 this hybrid repo thing is the pits 21:14:26 Xach: WTF? *sigh* Oh, lisp. haha. I love you so. 21:14:28 syntard_: check out my blog from yesterday if you like: http://elliottjohnson.net/blog/?p=182 specifically the diff at the end 21:14:30 problem is, i got quicklisps odbc to work, the git version was too hard 21:14:33 -!- metasyntax` [~taylor@12.132.219.7] has quit [Quit: Be seeing you.] 21:14:43 *redline6561* thanks God again for quicklisp 21:14:48 -!- vroufe [~omx@ppp94-29-9-58.pppoe.spdop.ru] has quit [Quit: Lost terminal] 21:14:48 ... it's like I don't even have to *think* 21:14:49 ejohnson: will you ever moderate my comment? :~( 21:14:50 so I'll try to patch quicklisp's clsql with fluid 21:15:05 -!- stassats [~stassats@wikipedia/stassats] has quit [Ping timeout: 255 seconds] 21:15:07 Xach: hmm ok :) Let me check. 21:15:09 ejohnson: also, it seems like half the post is inadvertently in a fixed-width font. 21:15:19 syntard_: Can you paste the specific error you're getting? I see in weblocks-clsql.asd where we depend on it but don't know how or why you'd patch that. 21:15:22 sdsds [~sdsds@dyn-16.sub2.cmts01.cable.TORON10.iasl.com] has joined #lisp 21:15:29 zeroish [~zeroish@135.207.174.50] has joined #lisp 21:15:44 syntard_: ejohnson may be better help here. 21:15:50 ejohnso'n's tips are good, but I use odbc, not mysql 21:15:54 Xach: I put it in a pre tag. Do you suggest something better? 21:16:16 ejohnson: is that intentional? 21:16:30 Xach: it is a diff, so I wanted it to stand out a bit. 21:16:39 -!- mbohun [~mbohun@ppp115-156.static.internode.on.net] has quit [Remote host closed the connection] 21:16:41 -!- Demosthenex [~demo@206.180.155.43.adsl.hal-pc.org] has quit [Ping timeout: 250 seconds] 21:16:44 ejohnson: everything after your example "quickload" form is in a fixed-width font for me. paragraphs of prose, not code. 21:16:47 -!- Cowmoo [~Cowmoo@cambridge-vxty.basistech.com] has quit [Ping timeout: 276 seconds] 21:16:50 i'll share a picture... 21:17:22 Xach: interesting. I don't see that on firefox. phoowee 21:18:06 http://xach.com/tmp/tt.png 21:18:07 ejohnson: I see it on Chrome/Linux64 21:18:13 Ah missing [/code] tag that should fix it. 21:18:24 *redline6561* has done that so many times 21:18:32 much better 21:18:35 Demosthenes [~demo@206.180.155.43.adsl.hal-pc.org] has joined #lisp 21:18:49 I think my fonts are fixed format, so really didn't notice it. Thanks! :) 21:19:01 Yup fixed. 21:20:09 -!- billstclair [~billstcla@unaffiliated/billstclair] has quit [Ping timeout: 250 seconds] 21:20:23 Xach: I'll answer your question on the blog to keep it all in the same context. 21:20:32 -!- LaPingvino [~LaPingvin@187.87.224.70] has quit [Ping timeout: 276 seconds] 21:20:58 ejohnson: Thanks. I also filed https://bugs.launchpad.net/cffi/+bug/672586 for the compatibility bug. 21:21:40 that's one of the reasons i don't use cffi-uffi-compat in quicklisp. 21:21:41 billstclair [~billstcla@unaffiliated/billstclair] has joined #lisp 21:22:33 Xach: Very cool, thanks again. 21:23:23 -!- fogus_ [c6970d0f@gateway/web/freenode/ip.198.151.13.15] has quit [Quit: Page closed] 21:23:40 hugod [~hugod@bas1-montreal50-1279440432.dsl.bell.ca] has joined #lisp 21:29:00 bmel [~bmel@li236-165.members.linode.com] has joined #lisp 21:29:38 Adamant [~Adamant@unaffiliated/adamant] has joined #lisp 21:34:15 -!- Jasko [~tjasko@c-174-59-223-208.hsd1.pa.comcast.net] has quit [Ping timeout: 245 seconds] 21:34:35 bgs100 [~ian@unaffiliated/bgs100] has joined #lisp 21:34:42 dlowe [~dlowe@c-66-30-112-95.hsd1.ma.comcast.net] has joined #lisp 21:35:17 Jasko [~tjasko@c-174-59-223-208.hsd1.pa.comcast.net] has joined #lisp 21:35:45 -!- tychoish [~tychoish@foucault.cyborginstitute.net] has quit [Ping timeout: 250 seconds] 21:35:46 *Xach* has a feeling this will be an epic weekend for lisp hacking, globally 21:36:19 *sykopomp* is planning on a massive hackathon, personally. 21:37:44 Xach: oh why so? 21:37:51 Krystof [~csr21@csrhodes.plus.com] has joined #lisp 21:40:46 -!- rickmode [~rickmode@cpe-75-85-93-10.socal.res.rr.com] has quit [Quit: rickmode] 21:40:47 ejohnson: Something in the air. Smells like autumn, and parentheses. 21:41:33 francogrex [~user@109.130.117.141] has joined #lisp 21:41:55 *redline6561* laughs. Hard. 21:42:00 -!- Davidbrcz [~david@212-198-92-25.rev.numericable.fr] has quit [Ping timeout: 260 seconds] 21:42:00 Xach: Please tweet that. 21:42:27 Xach: (The autumn and parens bit, that is). 21:42:49 kriyative [~user@ip65-44-141-106.z141-44-65.customer.algx.net] has joined #lisp 21:42:50 hah, is your lisp barometer registering a change? If I wasn't going mountain biking in SC this weekend I know I'd be. 21:42:52 bsod1 [~osa1@78.173.124.33] has joined #lisp 21:43:27 redline6561: I can't bring myself to. Feel free to quote, though. 21:43:40 Xach: Will do. 21:43:48 ejohnson: Fair enough. Enjoy the warmth while you can. I've been out skateboarding half the day. 21:43:57 I'm learning common lisp from the book called "land of lisp" but I can't find any IDE or useful editor for it, I don't know anything about emacs so I can't use slime, do you have any advices? 21:44:04 redline6561: nice! 21:44:24 srolls [~user@c-76-126-221-48.hsd1.ca.comcast.net] has joined #lisp 21:44:30 bsod1: Emacs and slime are really lovely and some here (including myself) would probably be happy to help you get going. 21:44:35 bsod1: On the other hand... 21:44:45 bsod1: I found it worth the effort to learn emacs, and I think that's what most people here use. 21:44:50 bsod1: I think there is some prepackaged development environment.. lispbox maybe? 21:45:24 bsod1: If you really don't want to learn them, there's slimv for vim, a textmate bundle if you use that and cusp for eclipse...though I'm not sure what state that is in. 21:45:34 bsod1: Realistically, you'll need to learn slime/emacs eventually. If you really want an IDE you can try LispWorks personal (http://www.lispworks.com/downloads/index.html) 21:45:43 smanek! 21:45:50 Hi Xach! 21:46:03 smanek: I tried lispworks personal edition but it's really bad, 21:46:14 smanek: Can you tell me more about the late postabon? 21:46:21 Congrats on all the traction Quicklisp seems to be getting 21:46:26 bsod1: What bit did you find bad? 21:46:31 Postabon is still around, but rebranded to Signpost.com 21:46:49 Xach: perhaps a haiku-tweet would be appropriate. 21:46:51 smanek: I mean more postabon.lisp 21:47:00 Xach: it can't evet do proper indentation *sorry for my english* 21:47:06 bsod1: Oh, but it can. 21:47:12 bsod1: you can download the acl trial, but I might be a bit biased to recommend it. The express version has an ide worth trying: http://franz.com/downloads/clp/survey 21:47:17 Xach: is there a secret option about it? 21:47:18 I'm no longer with the company (co-founder drama). Unfortunately, a lot of it is Java these days (part of the reason why I left ...) 21:47:26 seangrove [~user@c-71-198-44-87.hsd1.ca.comcast.net] has joined #lisp 21:47:38 ejohnson: thanks I'm looking at it right now 21:47:46 smanek: any chance of a postscript to the postabon lisp story? 21:47:46 np 21:47:49 'winter approaches/I can smell it in the air/oh, Parentheses!' 21:48:26 bsod1: I don't know, sorry, but some of the hard-corest Lisp nerds I know use LispWorks and they have a lot of things automated in the editor. There's also a really friendly lispworks mailing list. 21:48:29 -!- SpitfireWP [~Spitfire@wikipedia/spitfire] has quit [Ping timeout: 276 seconds] 21:48:54 bsod1: yeah the lispworks ide is top quality. 21:48:58 -!- X-02 [~kohei@p1021-ipbf405kyoto.kyoto.ocn.ne.jp] has left #lisp 21:49:07 sykopomp: Eh, just broke it up between two tweets... 21:49:15 :P 21:49:26 ok, I'll try again but last time I tried lispworks personal edition, it can't even do proper indentation and I had to use tabs and spaces 21:49:50 Hmm, I actually do have a few weeks free right now. I'm writing a small product in RoR, but if I finish that up soon I can try to provide some closure ;-) What sense of postscript do you want - a blog post mortem, or factoring out some of my code for the beginnings of a lisp web framework? 21:50:09 stis [~stis@1-1-1-39a.veo.vs.bostream.se] has joined #lisp 21:50:58 -!- rtoym [~chatzilla@user-0c99ag2.cable.mindspring.com] has quit [Ping timeout: 245 seconds] 21:50:58 Couldn't load foreign library "clsql_uffi" 21:51:23 smanek: all of the above? though i'm more interested in the post mortem, kinda. 21:51:25 ok, that's because it didn't compile 21:52:12 rtoym [~chatzilla@user-0c99ag2.cable.mindspring.com] has joined #lisp 21:52:29 redline6561: 'fall of my humor/how sad the limitation/one-hundred fourty' 21:52:46 sykopomp: Hahaha. I did lose the lyricism of the exchange, I'm afraid. 21:52:53 Xach: Fair enough (and the post-mortem is a bit quicker for me). I can promise you the post-mortem before the end of the month :-D I'll post it to HN/Lisp-reddit/etc 21:53:02 s/fourty/forty/ 21:53:30 bsod1: if you get stuck, ask a lispworker for help. they are so friendly and helpful it's not funny. 21:53:59 Xach, ok thanks 21:54:17 The short story is Lisp is cool in production. Elephant (with BDB) not so much. It is a bit annoying to keep explaining 'why lisp' though 21:54:21 homie [~user@xdsl-87-79-140-119.netcologne.de] has joined #lisp 21:54:25 they don't hang out here, though. 21:55:13 -!- francogrex [~user@109.130.117.141] has quit [Remote host closed the connection] 21:55:32 http://paste.lisp.org/display/116565 21:55:44 erred while invoking # on 21:55:44 -!- chiguire [~chiguire@gentoo/developer/chiguire] has quit [Quit: chao chiguire ...] 21:55:46 -!- milanj [~milanj_@93-87-167-170.dynamic.isp.telekom.rs] has quit [Quit: Leaving] 21:57:39 wbooze [~user@xdsl-87-79-140-119.netcologne.de] has joined #lisp 21:58:17 -!- jweiss_ [~jweiss@cpe-069-134-009-048.nc.res.rr.com] has quit [Quit: Ex-Chat] 22:00:30 -!- ikki [~ikki@201.144.87.42] has quit [Ping timeout: 245 seconds] 22:00:43 -!- Ragnaroek [~chatzilla@pD951ED35.dip.t-dialin.net] has quit [Ping timeout: 240 seconds] 22:00:45 are there any other recommended ways to connect ms sql? this odbc stuff seems fragile 22:01:48 syntard_: have you tried going into the c:/LISP/clsql/uffi/ and running make by hand? Does it create the library? 22:01:50 -!- dfkjjkfd [~paulh@236-9-ftth.onsnetstudenten.nl] has quit [Quit: Lost terminal] 22:02:08 into that directory that is. 22:02:21 ejohnson: cygwin make? 22:02:47 syntard_: outside of my area since I've never built on windows before. cygwin make would be my choice though. 22:02:54 adf [~user@130.67.202.84.customer.cdi.no] has joined #lisp 22:03:53 ejohnson: cygwin make succeeded 22:04:25 ejohnson: ah, it made an .so 22:04:54 syntard_: ah, so this is outside of my area at this point. 22:05:08 ejohnson: but there is .dll there too 22:05:15 from git 22:05:37 syntard_: hmm can you load the foreign dll in lisp? 22:05:44 Couldn't load foreign library "clsql_uffi". (searched CLSQL-SYS:*FOREIGN-LIBRARY-SEARCH-PATHS*) 22:06:23 Serious errors encountered during compilation of "c:/LISP/clsql/uffi/clsql-uffi.lisp" 22:06:32 syntard_: there should be a list of suffixes that uffi accepts.. does it have ".dll" in there? 22:06:47 syntard_: let me find the name of that list. 22:07:28 ejohnson: look, I just pressed continue on breaks, and it created the table! 22:07:52 -!- adf [~user@130.67.202.84.customer.cdi.no] has quit [Quit: leaving] 22:08:12 syntard_: hmm weird! I guess I'd have to be there to see what's going on :) 22:08:49 -!- Efthymios [~efthymios@c-98-207-146-68.hsd1.ca.comcast.net] has quit [Quit: Leaving] 22:11:07 here's the paste http://paste.lisp.org/display/116565#1 22:12:54 Guthur [~Guthur@host86-136-51-90.range86-136.btcentralplus.com] has joined #lisp 22:13:30 -!- superflit [~superflit@140.226.49.160] has quit [Ping timeout: 272 seconds] 22:15:41 -!- setheus [~setheus@cpe-70-116-140-134.tx.res.rr.com] has quit [Quit: leaving] 22:17:38 ikki [~ikki@200.95.162.219] has joined #lisp 22:18:18 weblocks tried to use LIMIT in select to ms sql 22:18:50 -!- bsod1 [~osa1@78.173.124.33] has quit [Ping timeout: 245 seconds] 22:19:31 syntard_: what is your store configuration like? 22:20:42 ejohnson: (defstore *sql-store* :clsql '("demodb" "" "") :database-type :odbc) 22:22:49 -!- ak70 [~ak70@46.11.10.227] has left #lisp 22:23:57 sonnym [~evissecer@rrcs-184-74-137-167.nys.biz.rr.com] has joined #lisp 22:25:26 kaehi [~dsa@ext02.gsp.se] has joined #lisp 22:26:00 fgump [~gauthamg@188.74.82.177] has joined #lisp 22:26:17 syntard_: yeah, looks correct. I think that is how it is when you use a db that doesn't support limit. I think you'll need to customize it to use TOP or something like that. This MS stuff is so far away from anything I use, sorry I can't be more of a help. 22:27:39 ejohnson: is there a way to disable weblocks paging somehow? 22:30:22 Set the page size to infinity and beyond? 22:30:34 syntard_: I think you can write a method that doesn't :limit it's results 22:31:08 b-man_ [~b-man@189.34.49.137] has joined #lisp 22:32:00 it's ok 22:32:11 HG` [~HG@xdsl-92-252-105-72.dip.osnanet.de] has joined #lisp 22:32:57 syntard_: are you calling find-persistent-objects and is that what is issuing the SQL with LIMIT? 22:33:14 ejohnson: I'm just running clsql-demo 22:33:23 i don't know what it tries to do 22:33:41 -!- araujo [~araujo@gentoo/developer/araujo] has quit [Ping timeout: 255 seconds] 22:34:07 Just trying to get off the ground here, maybe not worth it 22:34:11 syntard_: ah ok, then you can comment out the ":limit (range-to-limit range)" line in src/store/clsql/clsql.lisp and recompile it. 22:34:18 -!- hargettp [~hargettp@dhcp-174.mirrorimage.net] has quit [Ping timeout: 245 seconds] 22:34:24 err recompile the function that is. 22:35:38 -!- b-man_ [~b-man@189.34.49.137] has quit [Read error: Connection reset by peer] 22:35:40 -!- ehu [~ehuels@ip118-64-212-87.adsl2.static.versatel.nl] has quit [Ping timeout: 260 seconds] 22:36:15 i'll do same thing for offset 22:36:16 lolsuper_ [~super_@unaffiliated/lolsuper-/x-9881387] has joined #lisp 22:37:19 ejohnson: umm, how do I recompile it then? 22:38:34 araujo [~araujo@gentoo/developer/araujo] has joined #lisp 22:40:03 syntard_: well I use slime, so usually a C-c C-c on a form. That differs from environment to environment.. If you just have a shell you can copy and paste in then (compile ..) the function name 22:40:22 kmwallio [~kmwallio@host113-70.brownlie.fit.edu] has joined #lisp 22:44:35 -!- lhz [~shrekz@c-dba672d5.021-158-73746f34.cust.bredbandsbolaget.se] has quit [Quit: Leaving] 22:45:00 I wonder what a videoclip for planet-lisp would look like... http://www.youtube.com/watch?v=1JZnj4eNHXE&feature=related http://sacurativo.blogspot.com/2010/08/javazone-by-lady-java.html 22:45:25 pjb: ! 22:46:19 -!- sdsds [~sdsds@dyn-16.sub2.cmts01.cable.TORON10.iasl.com] has quit [Read error: Connection reset by peer] 22:46:28 superflit [~superflit@140.226.49.160] has joined #lisp 22:46:38 Well, we have the Land-of-Lisp video, but it's not the same. 22:46:52 Steven_ [~sdsds@206.210.103.16] has joined #lisp 22:46:54 i had a clall request earlier. do you ahve time to try it? 22:46:57 Haha I love the number one comment on the youtube page for that video ;) 22:47:01 -!- Amadiro [~whoppix@ti0021a380-dhcp0225.bb.online.no] has quit [Remote host closed the connection] 22:48:52 http://paste.lisp.org/display/116572 22:49:05 Xach: ^ 22:49:35 pjb: Okay. That's cool. Where can you get a copy of this magical script? 22:49:40 By the way: Apple<->Oracle: http://www.informationweek.com/news/software/open_source/showArticle.jhtml?articleID=228200849&cid=RSSfeed_IWK_All 22:50:05 It's in http://git.informatimago.com/viewgit/?a=summary&p=public/bin http://tinyurl.com/327zd4u git clone http://git.informatimago.com/srv/git/public/bin 22:50:31 pjb: Thanks. 22:51:24 Amadiro [~whoppix@ti0021a380-dhcp0225.bb.online.no] has joined #lisp 22:53:05 mbohun [~mbohun@ppp115-156.static.internode.on.net] has joined #lisp 22:53:08 -!- phf [~user@38.104.111.94] has left #lisp 22:53:18 -!- LiamH [~none@pdp8.nrl.navy.mil] has quit [Quit: Leaving.] 22:53:19 redline6561: http://paste.lisp.org/display/116573 what does that mean? 22:53:34 syntard_: Checking... 22:53:56 That's when I press ADd 22:54:49 dfox [~dfox@ip-89-176-11-19.net.upcbroadband.cz] has joined #lisp 22:55:57 -!- Steven_ [~sdsds@206.210.103.16] has quit [Ping timeout: 240 seconds] 22:56:15 syntard_: I think weblocks is trying to use an initform to set the slot of an employee object in the demo or otherwise derive the type of one of its slots. 22:57:18 syntard_: I wish I could tell you more but I haven't used weblocks-clsql or clsql before. If you're just interested in trying weblocks out, my advice would be to use the cl-prevalence backend. If you are planning to deploy it in production/work environments, that's a different story. 22:57:29 syntard_: And either way I'd post this bug on the google group. 22:57:46 syntard_: http://groups.google.com/group/weblocks It would be most appreciated! Leslie would have a better idea than me. 22:58:06 redline6561: Ok, will do 22:59:47 redline6561: I'm evaluating its fitness for production, right now I'm scared 23:00:44 of course, it's not weblocks fault 23:00:49 -!- GrayMagiker [~steve@c-174-56-26-27.hsd1.nm.comcast.net] has quit [Quit: Leaving] 23:01:04 syntard_: I can appreciate that. 23:01:50 -!- ikki [~ikki@200.95.162.219] has quit [Ping timeout: 264 seconds] 23:02:03 syntard_: If you post to the google group with a few details on your needs you'll get an informed response, that I can assure you. 23:02:46 -!- Jabberwockey [~Jens@f050066129.adsl.alicedsl.de] has quit [Quit: Verlassend] 23:02:57 syntard_: It sounds like you have a strong need/desire to use MSSQL as a backend. If that's the case, clsql is your only option right now and I'm sure leslie or someone could figure out this issue quickly. Someone also has some patches for a cl-perec backend which (I think) would let you use MSSQL but I'm not sure. 23:03:47 redline6561: yeah, company is on it 23:04:32 redline6561: if weblocks used some REST, I could write an adapter 23:06:29 Steven_ [~sdsds@dyn-16.sub2.cmts01.cable.TORON10.iasl.com] has joined #lisp 23:07:10 syntard_: Leslie was recently telling me how RESTful URLs could be done with weblocks, it sounded fairly straightforward but I've forgotten the details at the moment. It might be worth checking out RESTAS though. 23:07:16 redline6561: are stores extensible, what if I wanted to write my own store interface 23:07:49 syntard_: They are. You just need to write methods for all the generic functions in the store api here: http://bitbucket.org/S11001001/weblocks-dev/src/tip/src/store/store-api.lisp 23:09:09 syntard_: And really, I suppose you could just *avoid* using the Store API altogether and do your own thing on the side. Wouldn't be my first choice though. At that point I'd be looking for more natural solutions. 23:11:03 -!- seangrove [~user@c-71-198-44-87.hsd1.ca.comcast.net] has quit [Remote host closed the connection] 23:11:22 seangrove [~user@c-71-198-44-87.hsd1.ca.comcast.net] has joined #lisp 23:11:23 -!- lolsuper_ [~super_@unaffiliated/lolsuper-/x-9881387] has quit [Ping timeout: 245 seconds] 23:12:12 redline6561: no I do want to play ball. I'll just probably create another store. Don't want to fight clsql 23:13:13 -!- fgump [~gauthamg@188.74.82.177] has quit [Quit: Ex-Chat] 23:14:20 -!- kriyative [~user@ip65-44-141-106.z141-44-65.customer.algx.net] has quit [Ping timeout: 255 seconds] 23:14:33 syntard_: Awesome! I've heard mixed things about clsql but it seems to be the only way to get MSSQL access at the moment. cl-rdbms doesn't support MS. 23:15:56 redline6561: plain-odbc works 23:16:27 -!- bgs100 [~ian@unaffiliated/bgs100] has quit [Read error: Connection reset by peer] 23:16:28 jpanenst: Good to know! syntard_ ^^ 23:16:38 *jpanest, Thanks. 23:16:57 redline6561: using it in production, no known issues. np. 23:16:59 bgs100 [~ian@unaffiliated/bgs100] has joined #lisp 23:17:11 jpanest: ok! redline6561: so it's just a matter of writing interface for it in weblocks? 23:17:12 ajmorgan [~ajmorgan@70.102.51.102] has joined #lisp 23:18:19 I not sure why clsql gets such a bad wrap 23:18:41 ikki [~ikki@201.144.87.40] has joined #lisp 23:18:58 syntard_: As I understand it. I'd pull it down with quicklisp and try to get a feel for using it, then print out the store-api.lisp file and use it as a guide (maybe with the clsql store code up for reference) and write some defmethods. :) 23:18:59 It can access quite a wide variety of databases 23:19:25 Guthur: As I said, I've heard mixed things. I should have been more explicit or just kept my mouth shut, I suppose. 23:19:30 Guthur: I'm sure it's great, just not for MS Sql 23:20:25 syntard_, Does it not work for MSSQL? 23:20:57 davertron [~david@c-24-218-166-166.hsd1.vt.comcast.net] has joined #lisp 23:20:57 LaPingvino [~LaPingvin@187.87.224.70] has joined #lisp 23:21:34 Guthur: It does, but it doesn't completely fit. Examples are LIMIT and OFFSET. Ms sql doesn't support them, yet clsql tries to query with them 23:21:40 Guthur: do you use clsql's caching layer or did you write one yourself? 23:22:16 ejohnson, I only used it at a very superficial level to be perfectly honest 23:22:45 I'd imagine caching is actually best left to the database in question 23:23:21 Guthur: you would think, but I've never liked how it's done in clsql and always cache myself. 23:23:39 *syntard_* is away 23:24:19 ejohnson, What would your recommendation be? 23:25:08 nefo [~nefo@2402:f000:5:8f01::143:81] has joined #lisp 23:25:08 -!- nefo [~nefo@2402:f000:5:8f01::143:81] has quit [Changing host] 23:25:08 nefo [~nefo@unaffiliated/nefo] has joined #lisp 23:25:09 Guthur: I'm not sure what you are asking. 23:26:44 ejohnson, Doesn't really matter much I suppose, if clsql is the only one that offers MSSQL support the choices are limited anyway 23:27:07 -!- fiveop [~fiveop@dslb-084-056-184-061.pools.arcor-ip.net] has quit [Quit: humhum] 23:27:07 SpitfireWP [~Spitfire@wikipedia/spitfire] has joined #lisp 23:27:10 -!- xinming [~hyy@218.73.128.97] has quit [Ping timeout: 245 seconds] 23:29:07 -!- trigen [~MSX@87.209.144.213] has quit [Ping timeout: 240 seconds] 23:29:22 -!- stis [~stis@1-1-1-39a.veo.vs.bostream.se] has quit [Remote host closed the connection] 23:30:04 Guthur: Ah you were thinking I was suggesting that he not use clsql.. Just the opposite, I expect that he would use clsql. Just asking how you delt with caching when using clsql, since the caching interface is the most lacking part of clsql in my opinion. 23:30:32 jao [~jao@83.32.170.229] has joined #lisp 23:30:42 lolsuper_ [~super_@unaffiliated/lolsuper-/x-9881387] has joined #lisp 23:30:59 ejohnson, Your experience goes a lot deeper than mine 23:31:51 Guthur: heh ok :) 23:32:24 I only used it quite superficially, if I was devising a strategy I would probably put it in the database system rather than try to build a solution in CL 23:32:38 -!- ikki [~ikki@201.144.87.40] has quit [Ping timeout: 255 seconds] 23:32:52 foo_ [44aa127b@gateway/web/freenode/ip.68.170.18.123] has joined #lisp 23:33:09 The DDL and DML are usually not great but there is often a wealth of expertise invested into most popular databases 23:33:28 they are pretty good at what they do 23:33:44 -!- varjag [~eugene@4.169.249.62.customer.cdi.no] has quit [Quit: Ex-Chat] 23:33:45 Guthur: actually what I'm talking about is caching on the lisp side, so you don't need to hit the database as often. 23:34:26 hargettp [~anonymous@pool-71-184-181-150.bstnma.east.verizon.net] has joined #lisp 23:34:29 umm I'd have that that would just come from good software design anyway...hopefully 23:34:35 that/thought 23:35:00 Guthur: well yes, good software doesn't just materialize ;) 23:35:34 there would be no fun if it did, hehe 23:36:15 Guthur: yeah, we'd all just be sitting on the beach enjoying a cool drink ;) 23:36:37 I'm not sure beach would like that. 23:36:58 hehe, I live in Ireland the time you can feasibly spend on the beach is quite limited 23:37:11 haha, I would enjoy the sauna though ;) 23:38:38 assuming that beach is in his sauna ;) 23:40:19 HET2 [~diman@cpc1-cdif12-2-0-cust125.5-1.cable.virginmedia.com] has joined #lisp 23:40:27 -!- foo_ [44aa127b@gateway/web/freenode/ip.68.170.18.123] has quit [Quit: Page closed] 23:42:20 sykopomp, How's you Lisp coming along? 23:42:21 dfkjjkfd [~paulh@236-9-ftth.onsnetstudenten.nl] has joined #lisp 23:42:27 you/your 23:45:11 -!- ejohnson is now known as ejohnson_afk 23:46:12 -!- schell [~schellsci@70-36-199-232.dsl.static.sonic.net] has quit [Quit: schell] 23:47:29 ikki [~ikki@200.95.162.219] has joined #lisp 23:47:48 I'm trying to load lispbuilder-sdl using quicklisp, but I get "The name "CFFI-UTILS" does not designate any package." this was working before I did an upgrade from ubuntu 9.10 to 10.04 last night, any ideas? 23:47:50 -!- edlinde [~edlinde@90-227-7-243-no15.tbcn.telia.com] has quit [Quit: edlinde] 23:50:08 davertron: what does (asdf:system-source-directory 'lispbuilder-sdl) tell you? 23:51:03 #P"/home/david/quicklisp/dists/quicklisp/software/lispbuilder-20101107-svn/lispbuilder-sdl/" 23:51:21 -!- V-ille [~ville@ville-hp.dhcp.fnal.gov] has quit [Quit: Leaving.] 23:51:39 -!- katesmith_ is now known as katesmith 23:51:48 hmmmm 23:51:55 davertron: what does a backtrace look like? 23:52:05 Xach: how do i get a backtrace? 23:52:14 davertron: do you use slime? 23:52:19 Xach: nope 23:52:26 davertron: do you use sbcl? 23:52:26 Xach: vim user :) 23:52:31 Xach: yeah sbcl 23:52:41 :backtrace in the debugger should show you a backtrace 23:53:07 -!- devinus [~devinus@65.107.181.222.ptr.us.xo.net] has quit [Ping timeout: 250 seconds] 23:54:00 0: (SB-INT:%FIND-PACKAGE-OR-LOSE "CFFI-UTILS") 23:54:00 1: (SB-INT:FIND-UNDELETED-PACKAGE-OR-LOSE "CFFI-UTILS") 23:54:00 2: (SB-IMPL::USE-LIST-PACKAGES NIL ("COMMON-LISP" "CFFI-SYS" "CFFI-UTILS")) 23:54:00 3: (SB-IMPL::%DEFPACKAGE 23:54:00 "CFFI" 23:54:00 -!- davertron [~david@c-24-218-166-166.hsd1.vt.comcast.net] has quit [Excess Flood] 23:54:20 davertron [~david@c-24-218-166-166.hsd1.vt.comcast.net] has joined #lisp 23:54:48 err, please use paste.lisp.org 23:55:03 Xach: sorry, didn't mean to paste there 23:55:48 -!- bmel [~bmel@li236-165.members.linode.com] has quit [Quit: leaving] 23:57:21 jeti [~user@p548EAF4C.dip.t-dialin.net] has joined #lisp 23:58:23 -!- hargettp [~anonymous@pool-71-184-181-150.bstnma.east.verizon.net] has quit [Quit: hargettp] 23:58:57 -!- superflit [~superflit@140.226.49.160] has quit [Quit: superflit] 23:59:10 tychoish [~tychoish@foucault.cyborginstitute.net] has joined #lisp 23:59:24 -!- araujo [~araujo@gentoo/developer/araujo] has quit [Ping timeout: 265 seconds] 23:59:33 davertron: the suspense is killing me!