00:00:19 what about using (defconstant +foo+ (if (boundp '+foo+) (symbol-value '+foo+) #.(..))) instead of defvar? 00:00:46 -!- davazp [n=user@56.Red-79-153-148.dynamicIP.rima-tde.net] has quit [Read error: 104 (Connection reset by peer)] 00:01:50 luis, if #'foo is only called at read/compile time, why does the eval-when include :execute? 00:02:03 does sbcl treat constants as constants? 00:02:59 stassats: As in "constants aren't, variables don't"? 00:03:43 seems that it does 00:03:54 (Yes, SBCL will happily inline constant values that have identity.) 00:04:04 -!- mcspiff [n=user@hlfxns0163w-142068210006.pppoe-dynamic.ns.aliant.net] has quit [Read error: 54 (Connection reset by peer)] 00:04:31 Adlai: well, what if you don't compile the file? 00:05:27 luis, true, your version also works without compilation. 00:05:43 nyef: what about putting them in some special place for GC? 00:05:58 what about just sticking the entire body of code for generating the hash table within the #. ? 00:05:58 stassats: That's a little less likely. 00:06:04 And the GC is scary anyway. 00:06:07 I think nyef suggested that already 00:06:24 nyef: heh, how so? 00:06:30 -!- Swords2 [n=kohii@pool-71-112-23-110.sttlwa.dsl-w.verizon.net] has quit [Read error: 110 (Connection timed out)] 00:06:53 luis: Have you looked at the code? 00:07:18 (02:55:23 AM) nyef: Build the table in a #.(let ((table (make-hash-table ...))) ...)? 00:08:21 nyef: is it too far from the state of the art? 00:09:00 nyef: is that a rhetorical question? :) 00:10:12 fph [i=joe@adsl-75-14-202-212.dsl.renocs.sbcglobal.net] has joined #lisp 00:10:37 stassats: I don't know about far from the state of the art, but the code itself is scary. 00:13:56 nyef: can you give me an example or two? I have read the code, but I haven't read other GCs. 00:14:07 antoszka: the mailing list (And #tech.coop) are the main discussion areas. Those of us who are actually 'core' members are simply working on our little piece of the puzzle independantly. 00:16:22 -!- kmcorbett1 [n=Keith@c-76-119-215-57.hsd1.ma.comcast.net] has quit [Read error: 110 (Connection timed out)] 00:16:33 homovitruvius [n=user@pool-151-201-33-139.pitt.east.verizon.net] has joined #lisp 00:19:45 Damnit, now I'm thinking about SBCL-os again. 00:20:25 give true power to the command line? 00:20:53 -!- mrSpec [n=NoOne@82.177.125.6] has quit [] 00:21:16 -!- krumholt__ [n=krumholt@port-92-193-60-163.dynamic.qsc.de] has quit [Read error: 110 (Connection timed out)] 00:23:01 nyef: That's good isn't it? 00:24:47 plage: No, it isn't. 00:24:51 I have enough else to do. 00:25:58 And SBCL-os isn't a project that I consider worthwhile at this time. 00:26:48 -!- chavo_ [n=user@c-66-41-11-10.hsd1.mn.comcast.net] has quit [Remote closed the connection] 00:27:08 -!- amnesiac [n=amnesiac@p3m/member/Amnesiac] has quit ["Leaving"] 00:27:34 chavo_ [n=user@c-66-41-11-10.hsd1.mn.comcast.net] has joined #lisp 00:27:35 -!- blackened` [n=blackene@ip-89-102-28-224.karneval.cz] has quit [] 00:31:58 -!- joast [n=rick@76.178.184.231] has quit [Read error: 110 (Connection timed out)] 00:33:01 -!- fph [i=joe@adsl-75-14-202-212.dsl.renocs.sbcglobal.net] has quit [" "] 00:33:53 joast [n=rick@76.178.184.231] has joined #lisp 00:35:11 -!- athos [n=philipp@92.250.250.68] has quit [Remote closed the connection] 00:35:52 -!- schoppenhauer [n=schoppen@unaffiliated/schoppenhauer] has quit [Read error: 104 (Connection reset by peer)] 00:36:10 -!- mattrepl [n=mattrepl@ip68-100-82-124.dc.dc.cox.net] has quit [] 00:37:37 mattrepl [n=mattrepl@ip68-100-82-124.dc.dc.cox.net] has joined #lisp 00:37:46 -!- mattrepl [n=mattrepl@ip68-100-82-124.dc.dc.cox.net] has quit [Client Quit] 00:38:58 quek [n=read_eva@router1.gpy1.ms246.net] has joined #lisp 00:40:29 schoppenhauer [n=schoppen@unaffiliated/schoppenhauer] has joined #lisp 00:45:09 -!- ygip [i=foobar@gateway.blegkilde.dk] has quit [Connection timed out] 00:46:47 ruepel0r [n=rue@203-97-49-162.dsl.clear.net.nz] has joined #lisp 00:50:04 -!- boyscared [n=bm3719@c-68-32-124-6.hsd1.va.comcast.net] has quit ["leaving"] 00:50:15 Edward_ [i=Ed@AAubervilliers-154-1-60-201.w90-3.abo.wanadoo.fr] has joined #lisp 00:51:41 -!- ruepel0r [n=rue@203-97-49-162.dsl.clear.net.nz] has quit [Remote closed the connection] 00:51:58 ruepel0r [n=rue@203-97-49-162.dsl.clear.net.nz] has joined #lisp 00:53:12 boyscared [n=bm3719@c-68-32-124-6.hsd1.va.comcast.net] has joined #lisp 00:54:16 -!- schoppenhauer [n=schoppen@unaffiliated/schoppenhauer] has quit [Remote closed the connection] 00:55:46 -!- plage [n=user@118.68.32.211] has quit [Read error: 110 (Connection timed out)] 00:56:03 WarWeasle [n=brad@c-98-220-168-14.hsd1.in.comcast.net] has joined #lisp 00:56:24 -!- ahaas [n=ahaas@c-71-59-145-183.hsd1.or.comcast.net] has quit [Read error: 110 (Connection timed out)] 00:57:25 schoppenhauer [n=schoppen@unaffiliated/schoppenhauer] has joined #lisp 00:57:34 CFFI question to anyone: What happens if I forget to dealloc a foreign array created by the CFFI? Does it get garbage collected? 00:58:30 -!- schoppenhauer [n=schoppen@unaffiliated/schoppenhauer] has quit [Remote closed the connection] 00:58:36 -!- sreeram [n=sreeram@122.174.65.3] has quit [Read error: 60 (Operation timed out)] 00:59:26 ahaas_ [n=ahaas@c-71-59-145-183.hsd1.or.comcast.net] has joined #lisp 01:00:13 nyef: now, if only OSkit wasn't borked due to changes in GCC... 01:00:51 p_l: Is OKkit still actively maintained? 01:01:10 schoppenhauer [n=schoppen@unaffiliated/schoppenhauer] has joined #lisp 01:02:07 -!- ruepel0r [n=rue@203-97-49-162.dsl.clear.net.nz] has quit [Remote closed the connection] 01:02:22 Well, my position at this point is that a Lisp userland is more interesting than a Lisp kernel. 01:05:04 *stassats* agrees 01:05:27 -!- dalton is now known as ausente 01:06:04 sp: OSKit 01:06:29 ruepel0r [n=rue@203-97-49-162.dsl.clear.net.nz] has joined #lisp 01:06:35 mrsolo_ [n=mrsolo@adsl-68-126-192-146.dsl.pltn13.pacbell.net] has joined #lisp 01:08:10 WarWeasle: no. 01:09:29 luis: thanx 01:10:37 -!- schoppenhauer [n=schoppen@unaffiliated/schoppenhauer] has quit [Remote closed the connection] 01:10:39 luis: Wait, to which question? OSKit being maintained or cffi-foreign objects being cleaned up? 01:11:01 (sorry, I haven't heard of OSKit in years.) 01:11:44 Let's talk... display technology. We were discussing keyboards earlier, and one of the examples used a laser or something to draw a keyboard on a desk surface and then had some way to track key events from there. Lasers surely aren't limited to just red... 01:12:27 nyef: there's one of those already, though. 01:12:38 Really? 01:12:40 yeah 01:12:51 nyef: http://www.virtual-laser-keyboard.com/ 01:13:33 Exactly! But why limit it to drawing a keyboard? 01:13:38 Can I get a color framebuffer on one of those? 01:14:32 looks like it's just a keyboard :P 01:14:46 You're missing my point here. 01:15:06 It's a projective display. 01:15:36 Can I get a full-color framebuffer of that? 01:16:01 technically, why not? 01:16:20 there are laser shows with all kinds of colours 01:16:51 Okay, so it's technically possible. Does anyone make one yet? 01:17:21 http://yuml.me/diagram/scruffy/class/samples < very nice site 01:17:32 Heck, go one better, does anyone make one that will take a standard SVGA input? 01:17:43 WarWeasle: I meant you have to manually free foreign memory. 01:17:54 Just paper over the wall of your cube and project your display in that direction. 01:18:01 nyef: there are instructions online on how to make your own table touchscreen display. 01:18:10 is that more what you're looking for? 01:18:21 It might be. 01:18:22 like that microsoft surface thing, but without the garbage OS. 01:18:30 Heh. 01:18:51 What I'm wanting is a decently large portable display for my laptop. 01:19:24 kewl. 01:19:35 Something that I can set up in a cube farm with no fuss, without taking up much desk space, yet providing a very good output resolution and size. 01:19:36 on a sheet of paper? 01:20:18 Paper makes for a great white backdrop. Some sticky tape and paper and you can convert one of the walls in a cell of a cube farm into a decent output surface. 01:21:24 nyef: that keyboard has one big problem 01:21:39 http://www.laserfx.com/Backstage.LaserFX.com/Standards/ILDAframes.html some laser image standard 01:21:48 just like all touchscreen-as-keyboard thingies 01:22:19 (good thing nokia knows not to fuckup like that with their future design) 01:22:53 -!- Soulman [n=kvirc@154.80-202-254.nextgentel.com] has quit [Read error: 110 (Connection timed out)] 01:23:08 p_l: are you an expert in multithreading algorithms? 01:23:36 Bigshot_: Nope. I do have some understanding on MT, but no algo, not to mention expert 01:23:53 -!- ahaas_ is now known as ahaas 01:23:57 p_l: The problem is, I hope, on the keyboard input tracking side of things? 01:24:25 nyef: no, it's because of lack of proper tactile feedback. plus, fingers will hurt 01:24:32 it's surprising how many programmers make their living in this multicore quad core age w/o understanding about MT 01:24:32 -!- mrsolo_ [n=mrsolo@adsl-68-126-192-146.dsl.pltn13.pacbell.net] has quit [Read error: 60 (Operation timed out)] 01:24:45 Right. So, not a problem, since I don't want the feedback side anyway. 01:25:40 nyef: display is quite good, but won't reach levels of µCRT, OLED, or LCD, not even CRT, I think 01:25:59 *stassats* would also like some better energy sources 01:26:12 sepisultrum [n=enigma@hcl-club.lu] has joined #lisp 01:26:19 stassats: maybe we should work on building dyson spheres? 01:26:52 sykopomp: for mobile devices 01:27:05 stassats: superconducting capacitors 01:27:15 transistors 01:27:29 PITA when they blow up, but hold more power than gas 01:27:33 p_l: "superconducting" doesn't smell like mobile to me 01:27:37 I have a superclass A and a subclass B. If A has a :before method and B a normal one. When I call (method b-object), will the :before method defined with A also be called? 01:27:40 PNP diode 01:28:06 sepisultrum: Typically, yes. 01:28:10 stassats: dyson spheres for individual mobile devices?! BRILLIANT! 01:28:18 nyef: unless you mess with MOP? 01:28:20 I mean, I'm sure we have enough stars :) 01:28:32 energy star efficient 01:28:35 stassats: we are getting better with that. And they have similar energy density per kilogram to car fuel 01:28:40 schoppenhauer [n=schoppen@unaffiliated/schoppenhauer] has joined #lisp 01:29:40 I built a strange gui hierarchy using CLOS and :before and :after methods, don't know if that's a good idea 01:29:59 p_l: Really? What's the recharge time on that, the decay while at minimum/no draw, and the consequences of a kinetic impact event? 01:30:15 -!- Nshag [i=user@Mix-Orleans-106-3-139.w193-248.abo.wanadoo.fr] has quit ["Quitte"] 01:30:24 Oh, and the cost. Can't forget the cost. 01:30:38 potential energy into kinetic energy yeah 01:30:46 recharge... depends on power source and how good dampener you've got to avoid getting a surge that will kill local power grid. 01:31:26 another way is to improve power efficiency of the mobile devices itself 01:31:32 decay should be neglible once we get nanofabs (or something close to that) to make them at sensible prices. 01:31:38 is there a telnet client library other than the one at http://www.cliki.net/telnetlib ? 01:32:10 consequences of breach... I once heard that superconducting capacitor fault is ... messy 01:32:50 the battery the size of the unremovable one in newest macbook should wipe out machine, user and cubicle ^^; 01:32:55 superintergalactictransistorophobs 01:33:19 Bigshot_: please, no Star Trek-like technobabble. 01:33:50 stassats: replacing all possible paths with ultraconductors might cut a little 01:34:03 Damn, so something big enough to power a car would be a -bad- thing to have in a car accident? 01:34:26 but who will power cryostat? 01:34:33 nyef: I suspect it would be similar to fuel tank exploding 01:34:55 make car powered by a mini nuclear reactor 01:35:02 inside a car 01:35:05 -!- homovitruvius [n=user@pool-151-201-33-139.pitt.east.verizon.net] has quit [Remote closed the connection] 01:35:22 goes on for atleast a decade 01:35:24 however, the tech necessary to make them in big quantities probably will provide apriopriate casing and cryo 01:35:28 Bigshot_: I remember reading fiction which had cars powered by fusion reactors. 01:35:42 mrsolo_ [n=mrsolo@adsl-68-126-202-41.dsl.pltn13.pacbell.net] has joined #lisp 01:36:14 hah, fusion. Singularity as power source, now that's a thing. Fuckup, and it eats the planet ;-) 01:37:05 how much uranium does it consume? 01:37:41 Why would fusion -consume- uranium? It classically takes hydrogen as an input. 01:37:59 Bigshot_: none. It runs of whatever you dump into the black hole 01:38:00 i thought nuclear plants use uranium 01:38:39 they put in that rod made up of what i forgot to control the splitting of atoms(?) 01:38:48 oopsy what am i saying lawl 01:38:50 Fission plants do. Fusion is a different process. 01:38:59 -!- WarWeasle [n=brad@c-98-220-168-14.hsd1.in.comcast.net] has left #lisp 01:39:21 ahh fission 01:39:35 Bigshot_: fission uses fissionable stuff, like uranium, thorium, other heavy materials, fusion uses light stuff and stops generating more power around iron 01:39:46 another 30 years, and we will have controllable fission reactors 01:39:58 stassats: s/fission/fusion/ 01:40:10 yes 01:40:22 and we do have controllable fusion reactors. It's just that they require more power to keep running than they produce 01:40:33 the smallest is the size of a toaster, iirc 01:40:52 Mmm... Get that problem sufficiently licked, and we'll be able to celebrate a bit. 01:41:05 (A toaster? Cool! Or not, given what happens inside.) 01:41:24 i thought (hehe) lisp programmers would be eagerly awaiting the 3rd edition of intro. to algorithms which now has multithreading algorithms. 01:41:37 p_l: can it toast? 01:41:49 stassats: with neutrons and gamma radiation, I guess 01:41:56 but they seem more interested in electronics and power plants 01:42:10 yummy 01:42:33 Bigshot_: yesterday (or more like two days ago?) there was a nice discussion on sex etc.... 01:42:43 can someone point me to a tutorial or an example of cl-pcre and named captures? i'm a perl guy tryign to learn lisp and wanted to convert some regexp code... 01:42:47 oh man i missed it 01:42:48 Bigshot_: This is nothing. A few years ago, the channel had semi-regular gun-control debates. 01:42:58 yes, LESS OT, MORE LISP 01:43:01 -!- anekos is now known as A_anekos 01:43:02 Demosthenes: you mean ppcre? 01:43:11 sykopomp: likely 01:43:21 -!- Odin- [n=sbkhh@adsl-2-92.du.snerpa.is] has quit [] 01:43:47 do they have multithreading in AI? 01:44:08 Bigshot_: depends on what color RAM it uses 01:44:32 Bigshot_: I prefer to avoid the term "multithreading". It's too cramped 01:44:34 Demosthenes: like this http://weitz.de/cl-ppcre/#*allow-named-registers* ? 01:44:37 color? hopefully it has enough TLB cache size 01:45:25 multithreading is gonna be a hot topic in techonology now 01:45:41 processors are getting more powerful :D 01:45:48 multcore 01:46:02 it's certainly been the most ignored topic in computers historically 01:46:07 better prepare ourselves 01:46:27 wanna make good living learn complicated MT algorithms 01:46:36 stassats: http://www.regex-engineer.org/slides/img18.html 01:46:39 stassats: effectively 01:47:12 Demosthenes: what's the moral of that picture? 01:48:19 -!- dlowe [n=dlowe@c-24-91-154-83.hsd1.ma.comcast.net] has quit ["Leaving."] 01:48:24 Tordek [n=tordek@host37.190-137-136.telecom.net.ar] has joined #lisp 01:48:53 Demosthenes: doesn't my link answer your question? 01:49:08 Bigshot_: There's more than one way to do it, but it's not Perl? XD 01:49:33 i don't get it you like MT in perl Adlai? 01:50:26 Bigshot_, I thought you were referring to that link which he posted. My comment had nothing to do with multithreading... 01:50:33 oh 01:51:16 -!- schoppenhauer [n=schoppen@unaffiliated/schoppenhauer] has quit [Remote closed the connection] 01:51:51 stassats: not so much ;] 01:52:05 stassats: i was looking for a more newb example :P 01:52:16 Bigshot_: learn lisp? 01:52:28 sreeram [n=sreeram@122.174.68.30] has joined #lisp 01:52:31 well, you're perl guy, it's almost identical 01:52:38 -!- joachifm [n=joachim@ti132110a340-3137.bb.online.no] has quit [Read error: 110 (Connection timed out)] 01:52:56 stassats: true, the syntax is, but i really don't see how the data is coming out 01:54:54 in perl you access %+, ie: $+{'a'} for '(?.*)' 01:55:12 -!- ignas [n=ignas@ctv-79-132-160-221.vinita.lt] has quit [Read error: 110 (Connection timed out)] 01:55:55 dialtone_ [n=dialtone@adsl-64-175-33-37.dsl.pltn13.pacbell.net] has joined #lisp 01:56:17 -!- dialtone [n=dialtone@unaffiliated/dialtone] has quit [Read error: 60 (Operation timed out)] 01:56:43 syntax?! burn burn burn 01:57:18 sykopomp: Even lisp has syntax, so don't even try that. 01:57:23 pah, give a new guy a break, at least i'm trying. 01:57:35 i have a regexp problem i want to convert, but again, that doc makes little sense to me 01:57:48 i'd hoped for another tutorial, but i haven't found anything 01:58:05 nyef: not when compared to perl :\ It just looks like a bunch of obfuscated line noise to me. 01:59:52 *nyef* is completely unfamiliar with that bit of regexp syntax, but has seen regexps used via cl-ppcre in the cl-irc cliki-bot example. 01:59:55 minion: Isn't that right? 01:59:56 you speak nonsense 02:00:01 Ugh. Bloody hotel internet. :-/ 02:00:03 At least I'm theoretically down to two weeks on that... 02:01:25 Demosthenes: it's useful only for backreferencing, for accessing you have to save names returned by CREATE-SCANNER and do the mapping by yourself 02:01:32 well, that may be true. but i'm extracting data from text using regexps and the named captures are handy 02:01:40 stassats: ick :P 02:01:49 -!- Edward_ [i=Ed@AAubervilliers-154-1-60-201.w90-3.abo.wanadoo.fr] has quit ["L'oignon fait la farce."] 02:02:12 *nyef* has this sudden horrible thought of "perl5 in cl". 02:02:23 5? 02:02:47 -!- thomas001 [n=thomas@p5B0F7BC1.dip.t-dialin.net] has quit [Remote closed the connection] 02:03:02 6 or 7, i can imagine, but 5 what for? 02:03:22 Err... Maybe perl6. 02:03:32 No, perl5. 02:03:46 There's a large body of established perl5 code out there. 02:03:52 schoppenhauer [n=schoppen@unaffiliated/schoppenhauer] has joined #lisp 02:04:12 Bigshot_: I wish people stopped squeaking about multi-threading and "multicore" like it was all that was there... 02:04:27 mikezor_ [n=mikael@c-36ed70d5.04-404-7570701.cust.bredbandsbolaget.se] has joined #lisp 02:04:34 what else is there enlighten me :D 02:04:34 it's the new OOP! 02:04:47 design patters? 02:05:01 Bigshot_: you shall not speak of dipshit patterns ;-) 02:05:03 nyef: for LispOS? 02:05:07 Machine learning! Inductive reasoning systems! 02:05:16 s/of dipshit patterns ;-)// 02:05:17 stassats: It's a possibility, isn't it? 02:05:27 hefner: Heh. 02:06:24 machine learning is a lot of fun. 02:06:24 -!- joast [n=rick@76.178.184.231] has quit [Read error: 110 (Connection timed out)] 02:07:05 A lot of the machine learning stuff I've been reading has essentially been toy problems, though. 02:07:08 p_l: well... it's true that CPU speeds are kind of capping off for the time being, and processors are getting more cores stuffed into them. 02:07:20 A lot of manual input massaging and whatnot going on. 02:07:31 so if you want to run an app that sucks up as much CPU as possible on a modern system, you need to thread stuff :\ 02:07:47 sykopomp: I meant that "multicore" isn't anything new, just more cpus stuffed into single package and multithreading isn't the only way to use said power 02:08:14 p_l: ah. How else do you use that power? :\ 02:08:16 We've had multiprocessing systems for -how- long? 02:08:33 (unless you mean multithreading in the sense of 'spawn thread + make locks') 02:08:42 sykopomp: traditional unix approach was to use multiple processes communicating with each other 02:08:44 joast [n=rick@76.178.184.231] has joined #lisp 02:08:44 -!- dreish [n=dreish@minus.dreish.org] has quit [] 02:09:07 which is even reflected in shell 02:09:11 hm 02:09:30 *hefner* would like to bring some more sophisticated algorithms to bear against the "play super mario bros." problem 02:09:57 -!- danlei` [n=user@pD9E2E33C.dip.t-dialin.net] has quit [Read error: 60 (Operation timed out)] 02:10:27 heh, design patterns are funny... 02:10:43 sykopomp: plus, somehow, you always endup with message passing because you can't always get STM ;-) 02:11:09 Tuple spaces? 02:11:14 p_l: hm. Maybe I should play around with message passing :| 02:11:38 yeah IPC 02:11:48 it's for small amounts of data only 02:12:07 joshe [n=aurum@opal.elsasser.org] has joined #lisp 02:12:15 whoops! 02:12:24 Bigshot_: My X server disagrees 02:13:00 *stassats* wouldn't trust an X server 02:13:08 hefner: Would you be offended if I called "playing super mario bros." a "toy problem"? 02:13:51 stassats: after you hammer HAL out, you can somehow trust X.Org, I guess 02:14:01 i thought of shared memory lo 02:14:09 p_l: HAL9000? 02:14:31 I thought HAL was going away, to be replaced with Something Better? 02:15:04 nyef: DeviceKit. Somehow better idea, except still braindamaged, like the whole "use D-Bus for everything" 02:15:07 message passing is for large amount of data and shared memory (faster) for small amounts of data 02:15:34 Bigshot_: it all depends how you actually do this and from where to where 02:15:59 -!- mikezor [n=mikael@c-36ed70d5.04-404-7570701.cust.bredbandsbolaget.se] has quit [Read error: 110 (Connection timed out)] 02:16:12 S11001001 [n=sirian@74-137-151-39.dhcp.insightbb.com] has joined #lisp 02:16:44 nyef: not really. there's plenty of room there to do interesting things, though. 02:17:18 No question about that, but where's the -application-? 02:18:17 I'll claim you can draw some parallels to robotics. 02:20:29 Maybe once I get a few things sorted out I'll start messing with learning systems and whatnot. 02:20:52 ruepelOr [n=rue@203-97-49-162.dsl.clear.net.nz] has joined #lisp 02:20:53 *nyef* glares at SBCL/ARMv5. 02:21:15 for instance, if you use it as a domain for doing reinforcement learning, it's considerably less "toyish" than grid worlds and other silliness usually used to demonstrate these algorithms 02:21:21 konr [n=konrad@201.82.132.33] has joined #lisp 02:21:48 I can see that. 02:23:07 But at the same time, okay, you have a reinforcement learning system... It still just plays super mario bros. 02:23:37 It's an impressive accomplishment, but I still don't quite see the point. 02:23:47 well sorry, but I haven't built my army of machine gun wielding robots yet. 02:24:25 Now, see, -that- would be cool. Dangerous, of dubious moral stance, but cool. 02:24:28 I don't see how it's any less legitimate than playing with tools to make tools 02:24:35 (porting lisp compilers and whatnot) 02:24:38 -!- legumbre [n=user@r190-135-16-172.dialup.adsl.anteldata.net.uy] has quit [Remote closed the connection] 02:24:46 Fair point. 02:25:21 legumbre [n=user@r190-135-16-172.dialup.adsl.anteldata.net.uy] has joined #lisp 02:28:06 actually, I'm pondering a version control system as my next quick hack 02:28:54 nyef: (almost) speaking of ARM and embedded systems and whatnot, I wonder if this Openfirmware Forth system would make a nice basis for HobbyOSen at large. some quite nice neutral ground to share code for basic device drivers and whatnot. dunno if that's of any relevance to anything but it was a thought looking for a conversation :) 02:29:43 I started thinking about the index I'd need on disk to determine changed files, threw in Git's notion of storing everything indexed by hash, and when I realized I should store the index in there the same way, I think I'd basically derived git 02:30:05 -!- Urfin [i=foobar@85.65.93.74.dynamic.barak-online.net] has quit [Remote closed the connection] 02:30:40 -!- HET2 [i=diman@xover.htu.tuwien.ac.at] has quit ["This computer has gone to sleep"] 02:31:05 lukego: I was thinking of using a Forth based firmware, L4 kernel and some tools as basis for a weird OS ;-) 02:32:24 lukego: as in "Haskell, Forth, Lisp and Erlang meet and have groups sex" kind of weird 02:32:42 okay, that was bad mental image 02:33:34 well there are already a bunch of people writing toy OSen in lisp, smalltalk, haskell, python, etc. I feel like one common layer would be Forth to save them all writing a PS/2 keyboard driver, a USB driver, etc (or getting bored and moving on in the attempt) 02:33:58 without compromising the overall weirdo-language agenda with C or POSIX etc 02:34:59 Meh. PS/2 keyboard drivers are easy enough. 02:35:09 USB stack concerns me a little more. 02:35:16 And ACPI bloody terrifies me. 02:35:18 lukego: set of basic I/O drivers accessed through OpenFirmware would be good, but not really beneficient in long run (it's kinda like running on using BIOS) 02:35:37 bombshelter13p_ [n=bombshel@24.114.232.33] has joined #lisp 02:35:55 We have Linux. Why not leverage that for device support, at least? 02:36:19 nyef: L4Linux is a good way to do that 02:36:24 Especially with the KMS stuff landing. 02:36:32 If your OS uses L4, of course 02:36:54 Of course. 02:37:18 But where I was actually leading with that was why not just concentrate on a good userland instead of going the kernel route? 02:37:18 KMS? 02:37:21 though in my case, I'm not going to port Linux and do the work twice :D 02:37:30 Kernel Mode Setting. For video device support. 02:37:41 nyef: with L4, everything is userland ;-) 02:39:27 *p_l* is more interested in OS-level GC 02:40:00 nyef: I have no answer to the "why bother" argument. mine is an "if you're doing that anyway, what about" point :) 02:40:09 -!- schoppenhauer [n=schoppen@unaffiliated/schoppenhauer] has quit [Read error: 104 (Connection reset by peer)] 02:40:42 I do think that ultimately any reasoned decision to embrace e.g. Linux would tend to immediately follow a frustrated attempt to do without it. not a first choice :) 02:42:38 Heh. 02:43:49 Okay, I need to get up earlier tomorrow than I have the past couple of days, so I'm going to depart now. 02:43:53 -!- nyef [n=nyef@vcwl1-61.daktel.net] has quit ["G'night all."] 02:45:03 bombshelter13p__ [n=bombshel@24.114.232.33] has joined #lisp 02:45:48 aja [n=aja@S01060018f3ab066e.ed.shawcable.net] has joined #lisp 02:46:53 -!- quek [n=read_eva@router1.gpy1.ms246.net] has quit [Read error: 110 (Connection timed out)] 02:47:39 -!- bombshelter13p__ [n=bombshel@24.114.232.33] has quit [Client Quit] 02:48:51 drafael [n=tapio@ip-118-90-135-142.xdsl.xnet.co.nz] has joined #lisp 02:50:40 -!- dialtone_ [n=dialtone@unaffiliated/dialtone] has quit [Read error: 110 (Connection timed out)] 02:51:25 -!- bombshelter13p_ [n=bombshel@24.114.232.33] has quit [Read error: 60 (Operation timed out)] 02:53:15 I wonder if someone can help me with this simple function 02:53:17 http://paste.lisp.org/display/82268 02:53:31 The speed is exceptionally slow 02:53:53 there are 10 compiler notes and I haven't been able to get rid of any 02:53:57 first, your defun in let will be slower, I forget why 02:54:41 it is a closure 02:55:21 I'm having a problem with CXML... 02:55:29 why are you using let* ? 02:55:32 it's choking on   in the XML 02:55:55 is there anything I can do short of removing all   from the XML ahead of time? 02:56:18 (which would be a pain in the ass and I have no clue how to do, other than getting CL-PPCRE) 02:56:37 drafael: I forgot to change it 02:56:58 cky [n=cky@cpe-024-211-249-162.nc.res.rr.com] has joined #lisp 02:57:08 drafael: some people also like using let* even if they're not actually using the sequential binding thing. 02:57:30 oic 02:57:44 Adlai: how, as a rule, would you have CXML react to  ? 02:58:01 sykopomp: i consider it a bad style 02:58:04 hey leo2007 are you on sbcl? it doesn't like using setf so much -- rebind new variables with let instead 02:58:08 in this case, just treat it as a space. 02:58:24 stassats: me too 02:58:37 ilitirit: I am using sbcl 02:59:14 stassats: just pointing out that it's not universal. 02:59:21 here's the error it's throwing: 02:59:25 Adlai: you must add a DTD 02:59:30 sort of like how some people do #'(lambda ..) vs plain (lambda ..) 02:59:31 leo2007, it looks fun -- i'm going to lunch now, talk to you about it in a bit 02:59:32 Document not well-formed: Entity 'nbsp' is not defined. 02:59:47 ilitirit: ok. it works but too slow 02:59:51 S11001001, thanks, I'll look into that in the CXML manual. 03:00:01 p_l: how else can we make use of cores? 03:00:33 -!- cky [n=cky@cpe-024-211-249-162.nc.res.rr.com] has quit [Remote closed the connection] 03:00:51 it is meant to be 10 - 15 times faster. 03:01:47 why is sbcl using so much memory? 03:01:48 15491 web 19 0 522m 4400 3056 S 0 0.4 0:00.00 sbcl 03:02:01 (thats after a fresh start, nothing loaded/evaluated/etc) 03:02:33 cp2: you understand what VSZ means right? 03:03:26 S11001001: i guess not. 03:04:04 Is it possible to reload a package so that it restores to when it is first loaded 03:04:12 I am using sbcl 03:04:19 cky [n=cky@cpe-024-211-249-162.nc.res.rr.com] has joined #lisp 03:04:24 S11001001: but there's more to the problem: i am trying to install hunchentoot (via asdf yes, dont comment on that), but sbcl is complaining that it can't fork because it can't allocate any memory 03:04:43 this is on a vps with 768MB ram, 1024MB burst 03:04:49 it *should* be fine, no ? 03:05:10 that's your max RSS 03:05:26 see whether your provider limits VSZ 03:05:39 Bigshot_: scrollback buffer should have my response to that ;-) 03:05:59 S11001001: i will check 03:06:04 on my machine Clozure is "using" 512GB of VSZ :) 03:06:19 S11001001: and if they do, is there no other solution than to ask them to change it? 03:06:58 cp2: command line options 03:07:13 -!- mrsolo_ [n=mrsolo@adsl-68-126-202-41.dsl.pltn13.pacbell.net] has quit [Read error: 110 (Connection timed out)] 03:07:17 you might be able to play with the SBCL kernel to reduce the initial map, but that's where my knowledge ends, or you could use p_l's much simpler command-line solution :) 03:07:23 heh 03:07:32 p_l: what is your much simpler command-line solution then 03:07:43 using remote X server? 03:07:52 cp2: --dynamic-space-size 03:08:04 ill try it 03:08:16 S11001001, how do I create a DTD? I don't think I'm doing it right 03:08:20 also, what kind of VPS? Xen/KVM/VMware or some of that weird jail-like stuff? 03:08:40 Bigshot_: no, multiple processes etc. 03:08:47 p_l: i believe it is openvz 03:08:48 I tried using (setf (cxml:getdtd "nbsp" cxml:*dtd-cache*) " ") but it's still giving the same error. 03:09:20 multiple processes use more resource compared to threads 03:09:36 p_l: your much simpler command-line solution worked 03:09:37 Bigshot_: Depends on how you use them 03:09:38 thanks much 03:09:57 gimme an example p_l 03:09:59 cp2: don't make it too small 03:10:19 S11001001: of course 03:10:25 -!- ruepel0r [n=rue@203-97-49-162.dsl.clear.net.nz] has quit ["Leaving"] 03:10:35 Bigshot_: pipes, like | | etc 03:11:17 -!- sreeram [n=sreeram@122.174.68.30] has quit [] 03:11:23 Bigshot_: Or look into how Erlang does it. Sure, it uses one OS-thread per core, but one "image" spans multiple machines 03:11:53 image? 03:12:51 oh curse the day I decided to explain anything to anyone... 03:13:06 TDT pasted "Easier way to do this?" at http://paste.lisp.org/display/82269 03:13:37 p_l: hay, don't generalize Bigshot_ to everyone else 03:13:48 On that paste, I'm defining a hash of what error types I want to have tossed depending on regular expression matches. This works fine, but I want to do a define-condition on each of those to set stuff, is there a way to do it through a loop lik ei'm trying here? 03:13:55 Bigshot_: Think lisp-style image, with multiple nodes holding multiple processes passing messages 03:14:14 Ralith: I'm not generalizing, just complaining in general about explaining anything to anyone 03:14:17 Since (define-condition v) takes that literally as calling it "v", is there a way to use the variable v as I'm using it through the let? 03:14:18 How can I fix my CXML problem? 03:14:34 Adlai: It may be enough to prepend the standard !DOCTYPE header for XHTML1. 03:14:50 it's XML, not xhtml 03:15:38 TDT: write a macro 03:16:51 Adlai: in that case, write a short DTD that maps the entities you need, and stream it to the parser using the `entity-resolver' argument to the parser 03:17:06 stassats: k k, thanks. I thought about a macro, will have to take a look at how to do that again :) I guess this is my first real expample of the need for a macro which is kinda cool, took awhile but finally found a good example of a need for one. 03:17:11 S11001001, let me check that option out... 03:18:32 you may also have to pass some dummy as the DTD arg, it depends 03:18:41 ltriant [n=ltriant@lithium.mailguard.com.au] has joined #lisp 03:19:11 TDT: (gethash :errortype k) what's that? a poor man's classes? 03:20:35 stassats: I could have used classes here, but I decided as design decision to use a giant hash that has a listing of commands, and the error regexp to detect errors in the command (if there are any). I definitely welcome some opinions, as far as design, if you want to take a look at the full file (it's on github) 03:21:43 S11001001, I'm a bit confused... isn't there a way for me to just disable DTD? 03:21:46 mrsolo_ [n=mrsolo@adsl-68-126-202-41.dsl.pltn13.pacbell.net] has joined #lisp 03:22:10 stassats: http://github.com/TheDarkTrumpet/cl-gitinterface/blob/0b251dff160c1a90acde162163ebf9edfd4fcf8d/cl-gitinterface.lisp <-- current version. 03:22:24 Adlai: how else would the parser determine how to parse " "? 03:22:41 actually, I have no clue what DTD is... what I'd like to disable is the parser's attempts to parse things like that. 03:22:51 TDT: hash-table for command names is ok, but hashtables for commands itself isn't ok 03:23:12 I only want it to group things by their 03:23:30 DTD is a language for validating SGML (and incidentally XML) documents, and includes an entity definition syntax. That's where the &...; comes from in the first place, it's the SGML entity reference syntax, where entities are defined by DTDs 03:23:45 (deflcass git-command () ((error-type :initarg :error-type) ...)) 03:23:47 ok 03:23:53 defclass 03:24:06 hm 03:24:10 now,   doesn't "naturally" mean char 160, that's defined by the DTD 03:24:35 TDT: why not just define a package called GIT and stuff the commands into it as functions?.... O.o 03:24:36 the only XML-defined entities are apos,quot,lt,gt 03:24:52 (git:clone ...) and such 03:25:11 elliotstern [n=chatzill@cpe-67-240-162-85.rochester.res.rr.com] has joined #lisp 03:25:18 simplechat [n=simplech@unaffiliated/simplechat] has joined #lisp 03:25:44 sykopomp: This, I'm hoping, will just handle calls to git in the end. But - I see where you're going, so instead of passing a symbol, just have the function itself named that. 03:26:22 yes 03:26:32 right, git has an executable for each command, let's have a function for each command 03:27:51 Adlai1 [n=adlai@93-173-254-22.bb.netvision.net.il] has joined #lisp 03:27:53 Seems like a bit of duplication, would a macro work well for this too to define each command since the structure itself is pretty similar? 03:28:36 S11001001, sorry, did you answer my question? my internet died briefly 03:28:51 if you only need to store a regexp and a condition name, i still wouldn't make a bunch of functions 03:28:56 -!- Adlai [n=adlai@93-173-254-22.bb.netvision.net.il] has quit [Read error: 104 (Connection reset by peer)] 03:29:38   doesn't "naturally" mean char 160, that's defined by the DTD. the only XML-defined entities are apos,quot,lt,gt 03:29:39 the current idea seems ok, just implementation isn't great 03:30:25 writing a DTD is relatively simple, go look at the XML 1.0 spec and defining entities will probably be right there 03:30:29 so should my entity-resolver return (code-char 160) for   ? 03:30:37 stassats: Ah, thought you were for the idea of having functions for each. And yeah, I think I can do better. If you don't mind, I'll change up the way that you suggested earlier using classes - are there any other pointers you can make and I get your opinion on this in a few days? 03:32:00 no, you still have to write a DTD 03:32:53 tseug [n=tseug3@cpe-70-112-21-137.austin.res.rr.com] has joined #lisp 03:33:45 TDT: and make also a nice way to define all commands, something like (define-commands ((clone "*regexp*") (push "...") ...)) 03:34:03 out of curiosity why does your XML contain   if it isn't HTML? 03:34:09 condition names can be derived from the command names 03:34:11 -!- rullie [n=rullie@CPE00179acf86ed-CM000a735f8e51.cpe.net.cable.rogers.com] has quit [Read error: 60 (Operation timed out)] 03:34:38 hey leo2007 are you still around -- it seems to me you are shifting by 32 bits with a fp mul?! 03:35:06 I'm not sure why, but it occasionally does, and I'd like to be able to deal with that. 03:35:30 the part which contains   is an event log, which includes chat messages. 03:36:13 it seems as though every so often, a   creeps into the chat messages, and wreaks havoc on my cxml... 03:36:15 then you might want to copy even larger swaths of the XHTML DTD entity definitions 03:36:26 sqweek [n=none@203-206-65-102.dyn.iinet.net.au] has joined #lisp 03:36:43 stassats: k k, I think I understand what you mean - so a wrapper to define the commands a bit nicer than how I'm diong it wish hashes. 03:36:51 how do I define DTDs with CXML? it doesn't seem to work to (setf (getdtd ...) ...) 03:37:41 either prepend the DTD inline to the XML character stream, or define an entity-resolver that answers a stream encoding it back to ASCII codes 03:37:53 leo2007, wait a minute -- #xfffffffe is not 2^32-1 03:38:11 and give a fake DTD extid (described in CXML docs) so that your resolver will be called 03:38:47 2^32-2 03:39:06 leo2007, yes, but your rr is 1D/(ASH 1 32) 03:39:07 the way an XML document with a DTD prepended should look should be an example in the XML 1.0 spec 03:39:20 TDT: you wouldn't store anything besides regexps? 03:39:26 leo2007, *1d0/(ash 1 32) 03:39:34 S11001001, I'm reading this http://www.w3schools.com/DTD/dtd_intro.asp am I at the right place? 03:39:42 ok 03:39:51 if so, you can just have a HT with 'clone -> "regexp", 'push -> "push-regexp" 03:40:00 with no classes 03:40:24 actually, I'd rather make an entity resolver. What do you mean by giving a fake DTD extid? 03:40:38 Adlai1: that is basically how it looks but skip forward to the entity declarations 03:41:05 ilitirit: i've corrected it 03:41:40 w3schools is not part of the w3 consortium, so it's not an authoritative resource 03:41:49 but it has good stuff 03:42:48 read on make-extid, pass a systemid of NIL so it doesn't try to look on the filesystem 03:43:14 stassats: at this point, I don't feel - and hmm, so if i have a list of regexp like that, I can derive error type based off the key..then loop through the key in the macro for the define-condition. 03:43:17 if the two (cost 20) notes can be rid of, the speed will increase dramtically 03:44:13 it's "fake" because your entity resolver will ignore it and always return the same stream 03:44:16 TDT: you don't need to use hash tables at all with a proper macro 03:44:30 you let the implementation do the lookup 03:44:54 guaqua: How would I store the regexps in that case? 03:45:10 hm, so I'm basically just adding :dtd (make-extid nil nil) :entity-resolver #'(lambda ...) 03:45:23 TDT: doesn't git have more robust error reporting? because parsing its output seems to be too fragile 03:45:28 I think that would be fine 03:45:32 let me try it 03:45:48 you could just pass them as the arguments to the define-command 03:46:07 ilitirit: the function is my attempt to translate it from C as in http://school.anhb.uwa.edu.au/personalpages/kwessen/shared/Marsaglia03.html 03:46:45 stassats: I wish it did, actually..that's one of the problems I ran into is how to detect if it's an error, and the error messages aren't always similar 03:47:10 stassats: For example, a lot of commands have a "fatal:", which I could use to grab the message and send back..but not all of them have fatal: supported 03:47:26 Which...is kinda messed up in my opinion, made my job a bit more difficult. 03:47:43 why not just print error messages when it returns non-zero status? 03:48:39 S11001001, it's not working... 03:49:07 well, truth be told I haven't figured out how to do that quite yet, but some indication of what went wrong is probably helpful, unless I want to print everything returned...which I guess is possible 03:49:34 Would simplify matters a lot if I didn't have to store the regexp for this..just need to figur eout how to get the return code and I'll see how that works 03:49:55 Adlai pasted "DTD errors" at http://paste.lisp.org/display/82272 03:50:50 [Head|Rest] [n=win7@217.149.190.112] has joined #lisp 03:51:15 Adlai annotated #82272 "The call which generated these errors" at http://paste.lisp.org/display/82272#1 03:51:45 leo2007, just translate the C program! don't add bizarre long doubles (>> is ash) 03:52:13 Adlai1: make up a dummy URI for the systemid then 03:53:14 -!- Adlai1 is now known as Adlai 03:53:52 ilitirit: the original program returns uniform integers, I map them to the interval (0,1), uniform distribution 03:54:29 -!- ausente is now known as dalton 03:54:57 The bottleneck is the types. But I have very limited knowledge how to give the compiler more knowledge about them 03:57:07 leo2007, use the compiler more -- try to replace tmp with 'let's 03:57:17 sbcl's type inference doesn't like setf 03:58:15 Adlai: furthermore, entity-resolver yields the DTD, not an entity resolution. 03:58:45 so it should yield a DTD similar to those on the W3school page? 03:59:13 -!- chavo_ [n=user@c-66-41-11-10.hsd1.mn.comcast.net] has quit [Remote closed the connection] 04:00:24 Yes, except the external form. I really recommend you consult the W3C XML 1.0 spec instead 04:01:32 -!- nunb` [n=user@94.161.100.2] has quit [Read error: 60 (Operation timed out)] 04:04:49 -!- billstclair [n=billstcl@unaffiliated/billstclair] has quit ["Leaving..."] 04:08:58 -!- Bigshot_ [n=BIG_SHOT@CPE002129abc864-CM001ac35cd4d0.cpe.net.cable.rogers.com] has quit [Read error: 110 (Connection timed out)] 04:10:00 Alright, heading off to bed - got some of the changes already done, will make the rest later. Thanks again for your help stassats, guaqua, and sykopomp. 04:10:13 -!- TDT [n=TDT@126.91.248.216.dyn.southslope.net] has quit ["sleep"] 04:20:50 oudeis [n=oudeis@bzq-79-182-115-160.red.bezeqint.net] has joined #lisp 04:22:19 -!- pstickne [n=pstickne@c-24-21-76-57.hsd1.wa.comcast.net] has quit [Read error: 110 (Connection timed out)] 04:24:42 -!- dalton is now known as ausente 04:37:39 ilitirit: there? 04:40:58 -!- elliotstern [n=chatzill@cpe-67-240-162-85.rochester.res.rr.com] has quit [Read error: 113 (No route to host)] 04:44:18 amnesiac [n=amnesiac@p3m/member/Amnesiac] has joined #lisp 04:45:06 danlei [n=user@pD9E2C252.dip.t-dialin.net] has joined #lisp 04:53:55 leo2007 where? 04:54:16 -!- keithr [n=user@ip68-13-249-183.ok.ok.cox.net] has quit [Remote closed the connection] 04:58:32 keithr [n=user@ip68-13-249-183.ok.ok.cox.net] has joined #lisp 05:04:23 i wonder if i can change the slime hyperspec URL 05:13:14 sykopomp` [n=sykopomp@unaffiliated/sykopomp] has joined #lisp 05:13:39 -!- Raiford_ [n=rai@234.157-62-69.ftth.swbr.surewest.net] has quit [Read error: 60 (Operation timed out)] 05:13:46 Raiford [n=rai@234.157-62-69.ftth.swbr.surewest.net] has joined #lisp 05:15:03 zophy: Do you mean like this? (setq common-lisp-hyperspec-root "file:///usr/share/doc/hyperspec/") 05:17:51 ilitirit: is a static variable inside a C function persistent? 05:17:59 yes 05:18:24 -!- tsuru [n=user@c-69-245-36-64.hsd1.tn.comcast.net] has quit [Read error: 110 (Connection timed out)] 05:18:34 you are doing the right thing with the (let outside the defun, i don't think it's slow 05:19:27 ahaas, yes, i found where it is defined in hyperspec.el 05:19:38 zophy: You still owe me a beer. 05:20:03 -!- amnesiac [n=amnesiac@p3m/member/Amnesiac] has quit ["Leaving"] 05:20:15 where is your favorite virtual beer lounge ? 05:20:23 ha 05:21:02 -!- stassats [n=stassats@wikipedia/stassats] has quit [Read error: 113 (No route to host)] 05:23:42 ilitirit: so I think 'i' should also be moved to the outside let's 05:24:01 yes 05:25:10 blandest [n=blandest@softhouse.is.ew.ro] has joined #lisp 05:25:54 thanks 05:29:29 -!- sykopomp [n=sykopomp@c-66-41-62-126.hsd1.mn.comcast.net] has quit [Read error: 110 (Connection timed out)] 05:30:45 pstickne [n=pstickne@c-24-21-76-57.hsd1.wa.comcast.net] has joined #lisp 05:33:50 -!- sykopomp` is now known as sykopomp 05:34:12 S11001001, I realize this is a bit delayed, but I was taking a mental break (and vomiting) after reading the XML spec. 05:34:27 My parser works correctly now, thanks to your help, S11001001. 05:34:47 s/parser/usage of CXML 05:37:07 -!- oudeis [n=oudeis@bzq-79-182-115-160.red.bezeqint.net] has quit ["This computer has gone to sleep"] 05:39:34 KingNatoG5 [n=patrik@84-217-14-183.tn.glocalnet.net] has joined #lisp 05:40:43 -!- drafael [n=tapio@ip-118-90-135-142.xdsl.xnet.co.nz] has quit [Read error: 110 (Connection timed out)] 05:41:38 jlpeters [n=james@98.207.63.51] has joined #lisp 05:41:39 drafael [n=tapio@ip-118-90-138-62.xdsl.xnet.co.nz] has joined #lisp 05:41:59 -!- jlpeters [n=james@98.207.63.51] has quit [Remote closed the connection] 05:42:28 jlpeters [n=james@98.207.63.51] has joined #lisp 05:42:47 mrsolo__ [n=mrsolo@adsl-68-126-189-5.dsl.pltn13.pacbell.net] has joined #lisp 05:43:03 -!- jlpeters [n=james@98.207.63.51] has quit [Remote closed the connection] 05:43:15 -!- mrsolo_ [n=mrsolo@adsl-68-126-202-41.dsl.pltn13.pacbell.net] has quit [Read error: 113 (No route to host)] 05:46:02 *p_l* imagined a "send beer through internet" service 05:46:26 jmbr [n=jmbr@227.32.220.87.dynamic.jazztel.es] has joined #lisp 05:46:31 *Adlai* would send a beer to S11001001 as gratitude for helping him with CXML and DTD. 05:48:15 *p_l* actually remembers such a thing done by certain hacker 05:48:32 How? 05:48:34 nha [n=prefect@137-64.105-92.cust.bluewin.ch] has joined #lisp 05:48:56 clhs replace 05:48:57 http://www.lispworks.com/reference/HyperSpec/Body/f_replac.htm 05:49:16 *Adlai* apologizes for using specbot as a reference material, but was too lazy to Google. 05:49:36 Adlai: he posted his bank account number and instructions how to send him a beer (or something else) this way 05:49:45 ah 05:49:46 h 05:49:47 m 05:49:48 hm 05:50:09 give me your bank account number and i will happily send you a beer 05:50:27 i will also need your SSN, address and DOB 05:50:34 is there any way that I can replace all occurrences of a certain character by another one? 05:50:39 wait, why would you need a bank account number to send a beer? 05:50:44 *Adlai* also wants to know your mother's maiden name... 05:50:49 hahaha 05:50:51 Adlai: yeah that too 05:51:00 ehehe 05:51:07 that always struck me as ridiculously easy to look up 05:51:24 in poland, you don't have to keep your bank account number secret, as the only thing this knowledge allows is to send money to said account :D 05:52:01 well, even if you get a credit card number, you still need either a PIN, well-forged signature, or a digital signature 05:52:38 Adlai: Not if you get the whole number, including the last three digits, at least in case of VISA 05:53:10 well, you'd be able to use it for a bit, but if the person at the other end had any brains, he'd get the police on your ass pronto. 05:53:36 btw, said bank account # looks like this: 00-1111-2222-3333-4444-5555-6666 05:54:06 lol that's either a nice bank account number, or a country that's planning on way too much world domination. 05:54:33 10^26 potential citizens? holy crap. 05:54:34 Adlai: that's polish account numbering system 05:54:43 *Adlai* does naught but jest. 05:54:51 and it refers to account, not account holder 05:55:01 -!- ASau [n=user@193.138.70.52] has quit [Remote closed the connection] 05:55:14 ASau [n=user@193.138.70.52] has joined #lisp 05:55:28 I remember that it got extended few years ago into current form 05:56:08 -!- KingNatoG5_ [n=patrik@84-217-3-27.tn.glocalnet.net] has quit [Read error: 110 (Connection timed out)] 05:56:34 -!- ecraven [n=nex@140.78.42.103] has quit [Remote closed the connection] 05:59:09 Adlai1 [n=adlai@93-173-254-22.bb.netvision.net.il] has joined #lisp 05:59:09 -!- Adlai [n=adlai@93-173-254-22.bb.netvision.net.il] has quit [Read error: 104 (Connection reset by peer)] 05:59:29 -!- Adlai1 is now known as Adlai 05:59:35 ach, stupid internet. 06:00:32 -!- Lycurgus [n=Ren@pool-71-186-203-3.bflony.east.verizon.net] has quit [Remote closed the connection] 06:01:50 seejay [n=seejay@unaffiliated/seejay] has joined #lisp 06:03:00 SandGorgon [n=OmNomNom@122.160.140.244] has joined #lisp 06:06:42 -!- SandGorgon [n=OmNomNom@122.160.140.244] has quit [Read error: 104 (Connection reset by peer)] 06:11:18 -!- seejay [n=seejay@unaffiliated/seejay] has quit ["Leaving"] 06:12:45 Athas [n=athas@0x50a157d6.alb2nxx15.dynamic.dsl.tele.dk] has joined #lisp 06:13:06 eno_ [n=eno@adsl-70-137-150-39.dsl.snfc21.sbcglobal.net] has joined #lisp 06:13:17 -!- xan-afk is now known as xan 06:16:26 -!- ASau [n=user@193.138.70.52] has quit ["off"] 06:19:17 Davse_Bamse [n=davse@130.226.210.2] has joined #lisp 06:20:03 -!- Ash [n=aaron@outofband.org] has quit ["leaving"] 06:20:55 seejay [n=seejay@unaffiliated/seejay] has joined #lisp 06:23:44 -!- eno [n=eno@nslu2-linux/eno] has quit [Read error: 110 (Connection timed out)] 06:26:30 -!- eno_ is now known as eno 06:26:53 SandGorgon [n=OmNomNom@122.160.140.244] has joined #lisp 06:30:56 -!- seejay [n=seejay@unaffiliated/seejay] has quit ["Leaving"] 06:32:34 Soulman [n=kvirc@154.80-202-254.nextgentel.com] has joined #lisp 06:35:32 Adlai: you're welcome 06:35:51 It was a bit harsh of you to tell me to read the XML spec. 06:36:22 Let's just say that it's a bit dense. 06:36:31 Really? 06:36:51 well, actually, no more so than the Hyperspec 06:37:25 I find many specs to be far more useful than tutorial-like alternatives. 06:38:02 I've been doing some Plone stuff, and Plone's got nothing but tutorials, and it's maddening trying to figure out specific things without aggregating bits mentioned in passing and implied from multiple sources. 06:38:05 sometimes, I agree, although I think that when you know nothing about a subject, a tutorial is good. After the tutorial, you can read the spec to see where the tutorial was wrong :) 06:38:28 minion: tell me about Plone 06:38:29 Sorry, I couldn't find anything in the database for ``Plone''. 06:38:32 hm 06:38:34 brb 06:38:35 a Python/Zope CMS 06:38:36 google :) 06:38:38 ah, ok 06:39:05 I can't even remember what I read to learn how to write DTDs, probably some book on XML from the library. Nowadays I would just pull "XML The Definitive Guide" off my shelf and look it up :) 06:39:17 oudeis [n=oudeis@bzq-82-81-26-68.red.bezeqint.net] has joined #lisp 06:39:55 I ended up writing a DTD, and just linking straight to it in my parse-xml function 06:40:02 But one spec other than CLHS I consult frequently is CSS 2.1. I find everything else to be abjectly useless, except when I need browser-specific hacking. 06:40:03 (a wrapper around cxml:parse) 06:40:30 How much control do you have over the XML generator? 06:40:37 the source of your input logs 06:40:41 he's a friend of mine :D 06:41:21 maybe you should give the DTD to him and have him maintain it. Without it, after all, the XML he's generating isn't even well-formed, much less valid 06:41:23 I've actually sent him a message specifying exactly what he should put at the top of any XML files that might have the   bit in them. 06:41:57 none of it is valid, because it doesn't have doctype or or anything like that 06:41:58 -!- jmbr [n=jmbr@227.32.220.87.dynamic.jazztel.es] has quit ["ERC Version 5.3 (IRC client for Emacs)"] 06:42:07 however, up until now, he's been the only one dealing with it 06:42:28 he's been writing his own server/client app 06:42:32 this isn't by chance coming from PHP? 06:42:49 I think he has done some PHP hacking, but I'm not sure 06:43:29 the server itself might be in PHP. I don't know how that's implemented. 06:43:36 the current clients are in Javascript though 06:44:27 Now is the first time that somebody else has to use his API, so it's almost a certainty that some things like invalid XML will turn up. 06:47:20 If he's hand-writing the XML in PHP, he might be using the htmlentities function, which assumes that the entities in HTML DTD are available. Changing these calls to use htmlspecialchars would make it only convert those that are also allowed by XML, so the XML output would be well-formed even without a DTD. 06:47:43 -!- ruepelOr [n=rue@203-97-49-162.dsl.clear.net.nz] has quit [Read error: 110 (Connection timed out)] 06:47:56 nikodemus [n=nikodemu@cs181229041.pp.htv.fi] has joined #lisp 06:49:02 I'm not sure, but I'll ask him when we next talk. 06:49:13 thx again for the pointers 06:51:44 alinp [n=alinp@86.122.9.2] has joined #lisp 06:53:09 -!- SandGorgon [n=OmNomNom@122.160.140.244] has quit [Read error: 54 (Connection reset by peer)] 06:53:51 seejay [n=seejay@202.69.200.5] has joined #lisp 06:53:58 -!- Foofie [n=innocent@86.80-203-225.nextgentel.com] has quit ["Leaving"] 06:54:12 -!- oudeis [n=oudeis@bzq-82-81-26-68.red.bezeqint.net] has quit ["This computer has gone to sleep"] 06:55:45 ruepel0r [n=rue@203-97-49-162.dsl.clear.net.nz] has joined #lisp 06:56:32 ecraven [n=nex@140.78.42.103] has joined #lisp 07:02:31 -!- aja [n=aja@S01060018f3ab066e.ed.shawcable.net] has quit [""I am going to a commune in Vermont and will deal with no unit of time shorter than a season.""] 07:02:44 mrSpec [n=NoOne@82.177.125.6] has joined #lisp 07:02:58 aerique [i=euqirea@xs2.xs4all.nl] has joined #lisp 07:07:04 -!- plutonas [n=plutonas@c-83-233-152-13.cust.bredband2.com] has quit [Remote closed the connection] 07:07:51 mega1 [n=mega@pool-01811.externet.hu] has joined #lisp 07:12:25 -!- alinp [n=alinp@86.122.9.2] has quit ["Leaving."] 07:12:33 legumbre_ [n=user@r190-135-4-18.dialup.adsl.anteldata.net.uy] has joined #lisp 07:14:34 SandGorgon [n=OmNomNom@122.160.140.244] has joined #lisp 07:16:18 hello 07:16:45 alinp [n=alinp@86.122.9.2] has joined #lisp 07:17:29 ASau [n=user@host169-231-msk.microtest.ru] has joined #lisp 07:17:33 -!- alinp [n=alinp@86.122.9.2] has quit [Client Quit] 07:18:16 alinp [n=alinp@86.122.9.2] has joined #lisp 07:33:14 tcr [n=tcr@host145.natpool.mwn.de] has joined #lisp 07:34:04 -!- legumbre [n=user@r190-135-16-172.dialup.adsl.anteldata.net.uy] has quit [Read error: 110 (Connection timed out)] 07:35:21 ejs [n=eugen@36-90-135-95.pool.ukrtel.net] has joined #lisp 07:35:24 asksol [n=ask@pat-tdc.opera.com] has joined #lisp 07:36:51 athos [n=philipp@92.250.250.68] has joined #lisp 07:37:51 -!- alinp [n=alinp@86.122.9.2] has quit ["Leaving."] 07:38:04 alinp [n=alinp@86.122.9.2] has joined #lisp 07:40:12 -!- ltriant [n=ltriant@lithium.mailguard.com.au] has left #lisp 07:43:25 S11001001: How are you finding Plone? I haven't played with Zope/Plone in long time 07:46:56 -!- asksol [n=ask@pat-tdc.opera.com] has quit [] 07:50:31 -!- pstickne [n=pstickne@c-24-21-76-57.hsd1.wa.comcast.net] has quit [Read error: 110 (Connection timed out)] 07:50:44 asksol [n=ask@213.236.208.247] has joined #lisp 07:51:18 -!- ejs [n=eugen@36-90-135-95.pool.ukrtel.net] has quit ["This computer has gone to sleep"] 07:52:05 -!- SandGorgon [n=OmNomNom@122.160.140.244] has quit [Read error: 104 (Connection reset by peer)] 07:54:53 gemelen [n=shelta@shpd-92-101-154-182.vologda.ru] has joined #lisp 08:01:29 Fufie [n=poff@Gatekeeper.vizrt.com] has joined #lisp 08:07:11 mvilleneuve [n=mvillene@ABordeaux-253-1-5-135.w82-125.abo.wanadoo.fr] has joined #lisp 08:07:30 -!- Gertm [n=user@mail.dzine.be] has quit [Remote closed the connection] 08:07:41 Gertm [n=user@mail.dzine.be] has joined #lisp 08:08:15 oudeis [n=oudeis@bzq-84-109-137-43.red.bezeqint.net] has joined #lisp 08:10:22 oudeis_ [n=oudeis@bzq-84-109-137-43.red.bezeqint.net] has joined #lisp 08:12:49 -!- oudeis [n=oudeis@bzq-84-109-137-43.red.bezeqint.net] has quit [Read error: 60 (Operation timed out)] 08:13:15 SandGorgon [n=OmNomNom@122.160.140.244] has joined #lisp 08:17:02 -!- ace4016 [i=ace4016@cpe-76-168-248-118.socal.res.rr.com] has quit ["night"] 08:17:07 good morning 08:17:12 -!- oudeis_ [n=oudeis@bzq-84-109-137-43.red.bezeqint.net] has quit ["This computer has gone to sleep"] 08:17:37 p_l: delightfully complex and configuration-happy in its implementation 08:18:04 kiuma [n=kiuma@85-18-55-37.ip.fastwebnet.it] has joined #lisp 08:18:53 a great example of XML making things more confusing than they would be by just loading Python scripts. When your configuration format is constantly referencing classes by their package path, and regularly uses the Python packaging system itself, it's time to consider a metacircular configuration format 08:19:14 Otherwise it's fine I guess 08:25:34 asksol_ [n=ask@pat-tdc.opera.com] has joined #lisp 08:26:43 ejs [n=eugen@77.222.151.102] has joined #lisp 08:28:47 fe[nl]ix [n=algidus@88-149-210-231.dynamic.ngi.it] has joined #lisp 08:31:32 plage [n=user@118.68.38.106] has joined #lisp 08:31:36 Good afternoon. 08:32:13 -!- ejs [n=eugen@77.222.151.102] has quit [Read error: 60 (Operation timed out)] 08:32:36 ejs [n=eugen@nat.ironport.com] has joined #lisp 08:33:47 -!- Adlai [n=adlai@93-173-254-22.bb.netvision.net.il] has quit [Read error: 60 (Operation timed out)] 08:33:55 ThomasI [n=thomas@unaffiliated/thomasi] has joined #lisp 08:33:56 hej plage 08:41:53 -!- asksol [n=ask@213.236.208.247] has quit [Read error: 110 (Connection timed out)] 08:42:59 Adlai [n=adlai@93-173-254-22.bb.netvision.net.il] has joined #lisp 08:43:19 blackened` [n=blackene@ip-89-102-28-224.karneval.cz] has joined #lisp 08:44:52 There is a C function: void foo(double *x, int n) that takes an array of size N and writes stuff to it. What's the nicest way with cffi to wrap it? 08:45:37 -!- Athas [n=athas@0x50a157d6.alb2nxx15.dynamic.dsl.tele.dk] has quit [Remote closed the connection] 08:46:58 antoni [n=user@109.pool85-53-21.dynamic.orange.es] has joined #lisp 08:49:35 Sandeep` [n=Agent@122.172.124.215] has joined #lisp 08:50:37 yango [n=yango@unaffiliated/yango] has joined #lisp 08:50:38 Simple ASDF question: Can I have circular dependencies within files? ie, a few files that depend on eachother? 08:51:02 I realize this sounds like a stupid question. 08:51:41 -!- antoni [n=user@109.pool85-53-21.dynamic.orange.es] has quit [Remote closed the connection] 08:51:46 However, you can have distinct parts of a file which are mutually dependent... so I'm wondering if I can do this somehow with ASDF too. 08:52:46 matimago [n=user@cac94-10-88-170-236-224.fbx.proxad.net] has joined #lisp 08:53:22 fe[nl]ix pasted "wrap foo with cffi" at http://paste.lisp.org/display/82277 08:53:24 drafael1 [n=tapio@ip-118-90-139-33.xdsl.xnet.co.nz] has joined #lisp 08:56:48 fe[nl]ix annotated #82277 "err" at http://paste.lisp.org/display/82277#1 08:56:53 -!- kiuma [n=kiuma@85-18-55-37.ip.fastwebnet.it] has quit ["Bye bye ppl"] 08:57:22 fe[nl]ix: thanks. Isn't using with-pointer-to-vector-data controversial? 08:58:06 mega1: it pins the vector. AFAIK it's necessary 08:58:17 oudeis [n=oudeis@89-138-239-170.bb.netvision.net.il] has joined #lisp 08:58:30 either that or you copy it 08:58:45 -!- drafael [n=tapio@ip-118-90-138-62.xdsl.xnet.co.nz] has quit [Read error: 60 (Operation timed out)] 08:58:48 yes 08:59:19 what's controversial about it? 08:59:43 that the implementation is using without-gcing on some platforms 08:59:59 I see. 09:00:01 and that it doesn't seem to be implemented for allegro 09:00:35 Unrelated bit: Does anyone have/know real-life code with code-deletion notes? 09:01:30 tcr: I do 09:01:43 mega1: or you could allocate the vector in static memory 09:01:55 nenorbot [n=nix@bzq-219-145-16.static.bezeqint.net] has joined #lisp 09:02:17 fe[nl]ix: In IOlib? 09:02:22 fe[nl]ix: is that possible to do portably? 09:02:29 tcr: yes 09:04:35 mega1: it works in allegro, lispworks and ccl. I've got one for sbcl 09:04:44 *one implementation 09:05:36 I'll stick to hand translating then. 09:05:44 -!- SandGorgon [n=OmNomNom@122.160.140.244] has quit [Read error: 104 (Connection reset by peer)] 09:07:31 Swordsman [n=kohii@pool-71-112-23-110.sttlwa.dsl-w.verizon.net] has joined #lisp 09:11:26 stassats [n=stassats@wikipedia/stassats] has joined #lisp 09:12:05 -!- mrsolo__ [n=mrsolo@adsl-68-126-189-5.dsl.pltn13.pacbell.net] has quit [Read error: 110 (Connection timed out)] 09:15:33 tcr: they're in src/streams/zeta/ffi-functions-unix.lisp, the two uses of %sys-ioctl 09:23:17 -!- DeusExPikachu [n=DeusExPi@pool-173-58-94-2.lsanca.fios.verizon.net] has quit ["Leaving"] 09:25:13 -!- fe[nl]ix [n=algidus@88-149-210-231.dynamic.ngi.it] has quit ["Valete!"] 09:26:36 thom_ [n=thom@pool-173-51-224-238.lsanca.fios.verizon.net] has joined #lisp 09:28:42 -!- athos [n=philipp@92.250.250.68] has quit ["leaving"] 09:33:28 -!- sohail [n=Sohail@unaffiliated/sohail] has quit [Read error: 60 (Operation timed out)] 09:34:41 S11001001: After asking, I've got this strange idea of making a "Lope", aka Lisp Zope ;-) 09:35:06 especially since *this* approach haven't yet appeared in known to me lisp apps 09:37:33 ignas [n=ignas@78-60-73-85.static.zebra.lt] has joined #lisp 09:39:40 ugh don't 09:40:02 HET2 [i=diman@xover.htu.tuwien.ac.at] has joined #lisp 09:40:16 most of those platforms just reinvent the wheel 09:40:24 What would that be like? Just a Lisp that compiled natively on a mobile phone? 09:41:16 p_l, Zope is a horrible plaform. 09:41:20 S11001001: why not? Of course with something better than XML... 09:41:22 it was terribly sad that it was easier to pick up Weblocks with its almost complete lack of docs outside the source than Zope with its manuals and tons of tutorials etc 09:41:30 it's not even very Pythonic. 09:41:43 I don't mean remaking it, but making "similar" 09:41:57 I kinda liked Zope back when I tried it 09:41:59 there are great web frameworks for Lisp already, and if you're looking outside the Lisp world you should start with the web frameworks for Smalltalk 09:42:04 there's some real innovation there 09:42:06 -!- Ringo48 [n=Ringo48@97-122-180-214.hlrn.qwest.net] has quit ["Leaving"] 09:42:25 oh, weblocks has rounded corners 09:42:31 i'm up for it! 09:42:37 -!- [Head|Rest] [n=win7@217.149.190.112] has quit [Read error: 110 (Connection timed out)] 09:42:47 i found ruby on rails quite easy to pick up 09:42:48 -!- oudeis [n=oudeis@89-138-239-170.bb.netvision.net.il] has quit ["This computer has gone to sleep"] 09:42:51 -!- Sandeep` [n=Agent@122.172.124.215] has quit [Client Quit] 09:43:36 -!- ThomasI [n=thomas@unaffiliated/thomasi] has quit ["Bye Bye!"] 09:43:42 jewel [n=jewel@dsl-242-129-65.telkomadsl.co.za] has joined #lisp 09:44:50 fe[nl]ix [n=algidus@88-149-210-231.dynamic.ngi.it] has joined #lisp 09:45:33 ZabaQ [n=johnc@playboxgames.com] has joined #lisp 09:45:49 -!- pem [n=pem@159.226.35.246] has quit [Remote closed the connection] 09:46:54 morning fellow lambdaphiles 09:47:35 HET2: The nicest way to try a web framework is to do a project in one you never tried before, with deadline "yesterday" 09:48:08 p_l: unfortunately i've sworn to myself never to do web projects again - when i was 16 09:48:32 HET2: lucky you! 09:49:14 HET2: what didn't you like about web projects? 09:49:39 ( I did something similar, but about COBOL :-) ) 09:49:48 matimago: the flair of sweat^H^H^H^H^Hweb shops 09:50:39 Right, it's highly competitive. But I hope that with meta-programming and automatic programming we could compete. 09:51:01 matimago: That's what I'm hoping right now, after sneaking my way into webapp 09:51:21 i'm sticking with academia for the time being, thank you very much 09:51:29 HET2: but mostly, there are web sites that are really remote applications with a web interface. When the server side code is sophisticated we can still compete easily. 09:51:41 and some days I feel like trying COBOL over PHP ^_^ 09:52:08 Ah, if we must compare COBOL and PHP, of course, COBOL may look much better... 09:52:36 it's not the competitiveness that bothers me, it's the live to work as opposed to a work to live attitude that annoys me 09:53:14 well, at least I'm allowed to use Lisp ^_^ 09:53:24 that's always a plus 09:53:33 i think i'd do web dev if i was allowed to use lisp 09:53:45 at least for 1-2 years 09:54:00 current app is in Rails, but it needs some new features and I'm considering doing those in lisp 09:54:40 zomg, mixing web frameworks? 09:55:14 HET2: only data would be shared, with all parts being basically separate parts using single database 09:56:31 p_l, that actually sounds like a really good solution to integrating two systems in different languages... and everybody nowadays passes XML back and forth :-/ 09:56:50 argh the pain! 09:56:51 change topic! 09:57:11 HET2, I'm with you there... 09:57:42 I have to deal with mountains of XML in my current app. Luckily, I can use cxml-xmls to make it all go away... 09:59:01 fiveop [n=fiveop@pD9E6BE97.dip.t-dialin.net] has joined #lisp 09:59:50 Adlai: how do you avoid encoding hell 10:00:13 do you mean utf-8 vs others? 10:00:42 i mean that when receiving xml from different sources you rarely get them in a uniform encoding 10:01:07 luckily for me, it's all from one server 10:01:42 also, all the data which is relevant for my app is simple ASCII stuff -- latin characters and numerals. 10:02:17 now, there are chat logs involved, which could be in god-knows-what encoding... but my app doesn't have to deal with those. it can just throw them away. 10:02:43 xml'ed chat-logs? 10:02:58 yep. 10:02:58 *HET2* cringes at the thought of overhead 10:03:05 hehe 10:03:31 billstclair [n=billstcl@unaffiliated/billstclair] has joined #lisp 10:06:47 Swords2 [n=kohii@pool-71-112-23-110.sttlwa.dsl-w.verizon.net] has joined #lisp 10:07:07 -!- seejay [n=seejay@unaffiliated/seejay] has quit [Remote closed the connection] 10:07:19 btw, would you say it's better to use CXML-DOM or CXML-XMLS? 10:08:08 I'm using XMLS just for the convenience (assoc, remove, etc), but my cousin (who might be biased because he's never used Lisp) said that DOM is very neat and convenient. 10:08:33 frozsyn [n=FrozSyn@wafer.futurs.inria.fr] has joined #lisp 10:08:38 oh i've used neither 10:09:12 then how do you handle the XML? 10:10:02 nikodemus: I get a "could not stack-allocate" note in alexandria:map-derangements with 1.0.29.6 in, I believe, default optimization settings. Is is deliberate that dx-declarations are trusted when (= speed safety) ? 10:13:01 threading question re: locks 10:13:03 mattrepl [n=mattrepl@hq-users.caci.com] has joined #lisp 10:13:11 should I keep each lock in a dynamic variable, 10:13:31 or keep them all in one dynamic collection (probably an alist, since I don't have many)? 10:13:40 s/dynamic/global 10:13:59 -!- Swordsman [n=kohii@pool-71-112-23-110.sttlwa.dsl-w.verizon.net] has quit [Read error: 110 (Connection timed out)] 10:14:36 the latter introduces the overhead of another symbol, but I think it's conceptually a lot cleaner. 10:14:53 any thoughts? 10:23:09 Edward_ [i=Ed@AAubervilliers-154-1-43-234.w90-3.abo.wanadoo.fr] has joined #lisp 10:23:25 dlowe [n=dlowe@c-24-91-154-83.hsd1.ma.comcast.net] has joined #lisp 10:24:43 -!- hober [n=ted@unaffiliated/hober] has left #lisp 10:26:44 -!- cracki [n=cracki@sglty.kawo2.RWTH-Aachen.DE] has quit ["If technology is distinguishable from magic, it is insufficiently advanced."] 10:31:01 Athas [n=athas@0x50a157d6.alb2nxx15.dynamic.dsl.tele.dk] has joined #lisp 10:33:02 tcr: dx declarations for vectors of unknown size are ignored unless safety is 0, because they can subvert stack overflow protection 10:33:41 Adlai: depends what you are locking 10:33:53 access to a bunch of variables 10:34:02 if you are locking a structure, you should keep the lock in a structure 10:34:19 one is a list, the rest are CLOS objects 10:34:30 so I'm guessing what you said applies to them too? 10:34:34 if you are locking a variable, i would use variable for the lock as well 10:35:00 -!- drafael1 [n=tapio@ip-118-90-139-33.xdsl.xnet.co.nz] has quit [Read error: 110 (Connection timed out)] 10:35:33 well, they're in variables... but the values themselves are a list, and a bunch of CLOS objects 10:35:48 seems to me like I should use a separate var for the list, and a slot for each of the objects 10:35:50 yes? 10:35:56 drafael [n=tapio@ip-118-90-139-108.xdsl.xnet.co.nz] has joined #lisp 10:38:40 [Blay] [n=blaay@BSN-143-124-250.dial-up.dsl.siol.net] has joined #lisp 10:45:31 adityo_ [n=adityo@122.169.38.55] has joined #lisp 10:46:28 Yuuhi [n=user@p5483E711.dip.t-dialin.net] has joined #lisp 10:49:38 nikodemus: it should be possible to dynamically check for stack overflow, no? 10:50:07 shmho [n=user@58.142.15.103] has joined #lisp 10:51:26 kiuma [n=kiuma@proxy.emea.fedex.com] has joined #lisp 10:51:40 hello lispers 10:51:57 hi kluma 10:52:51 gzip4 [n=xxx@ws102.zone134.zaural.ru] has joined #lisp 10:52:58 porting my application from old hunchentoot to the new one, has been pretty easy, I'm very happy :) 10:55:27 *rsynnott* really couldn't be bothered right now 10:55:31 *rsynnott* is living in the past :) 10:56:12 :) 10:56:39 I just pulled something together 10:56:43 this might have been done already 10:56:48 but if not, here it is. 10:57:04 -!- fe[nl]ix [n=algidus@88-149-210-231.dynamic.ngi.it] has quit ["Valete!"] 10:59:21 oudeis [n=oudeis@p11811120.orange.net.il] has joined #lisp 11:00:09 zu22 [n=ubuntu@unaffiliated/zu22] has joined #lisp 11:00:13 any debian slime users here? 11:00:35 having trouble installing slime on squeexe: http://zu22.pastebin.com/fef3e71e 11:01:01 don't install slime from debian 11:01:34 in fact, avoid debian packages? :-D 11:01:34 use clbuild instead 11:01:47 Adlai pasted "alref" at http://paste.lisp.org/display/82279 11:02:10 debian _lisp_ packages 11:02:17 arch is the best ;) 11:02:31 the title is deceptively simple 11:02:52 drafael: sbcl from arch is quite good, true. And then you get clbuild :> 11:03:06 -!- jthing [n=jthing@165.244.251.212.customer.cdi.no] has quit [Remote closed the connection] 11:03:08 that paste actually involved (slightly) more work than (cdr (assoc foo bar)) 11:03:22 jthing [n=jthing@212.251.244.165] has joined #lisp 11:03:25 p_l: yep, I'm using clbuild :) but slime-cvs etc are in community anyway 11:03:35 Adlai: why EQ? 11:03:37 the distributions should really just stop including lisp packages 11:03:40 asking for trouble 11:03:45 stassats, as the default? 11:03:51 Adlai: yes 11:04:09 I was under the impression that was the default for assoc... 11:04:17 rsynnott: just like with Haskell, though Arch actually has properly working packaging for GHC and packages 11:04:41 (well, people maintain arch-packaged mirror of nearly whole hackage) 11:04:41 I was on the verge of including key-included-p and test-included-p parameters, and providing the keys or not based on those, but decided not to. 11:04:58 Adlai: it's EQL 11:05:01 really the only bit about debian packages is tat they are a bit outdated 11:05:02 I chose #'eq as the default because a) I thought it was the default for assoc, and b) it's the "simplest" one 11:05:20 unless you mix them with new libraries you will be fine 11:05:23 jthing: what about CLC? 11:05:31 choose EQL, because that's what everyone would expect 11:06:08 Adlai annotated #82279 "EQL as default parameter" at http://paste.lisp.org/display/82279#1 11:06:18 p_i: old stuff again. old tarballs 11:06:55 *jthing* is on ubuntu, a cleaned up debian distro 11:07:13 cleaned up? oh please 11:07:14 *Adlai* <3 ubuntu 11:07:35 well, I wouldn't say it's cleaned up, I'd just say it's more user-friendly 11:07:36 <_3b> ubuntu still uses clc 11:07:47 Adlai, FSVO user-friendly. 11:07:55 stassats: Debian contains a lot of 'junc' most people never use.. 11:08:01 most stuff that gives you OS info tells me that I'm running Debian 11:08:52 tic, true 11:09:38 ah 11:09:53 i also cannot get mod_lisp working 11:10:16 don't use it either 11:10:25 stassats, other than your comment about the default :test, what do you think of my hacklet? 11:10:57 i installed mod_lisp and set it up with: http://zu22.pastebin.com/f5334914d 11:11:09 so i made directory /var/www/lisp and put simple lisp file in there 11:11:17 when i go to it in browser nothing happens 11:11:20 it just lists the file contents 11:11:24 am i missing some key step? 11:12:00 use hunchentoot instead, unless you have a reason not to 11:12:09 -!- HET2 [i=diman@xover.htu.tuwien.ac.at] has quit [Remote closed the connection] 11:12:26 jthing: buahahaha. At least Debian doesn't force you to install it nor depends on it by default 11:12:49 nikodemus: My point is why the note is emitted 11:13:23 <_3b> zu22: did you miss a line while pasting that? 11:13:55 _3b: lemme check 11:14:37 _3b: nopee 11:14:52 to chime in, zu22 .. don't use mod_lisp .. use the mod_proxy found in pretty much any http front-end 11:15:24 <_3b> looks like it should have a at the end (and yeah, running a lisp web server through a proxy, apache or othewise would probably be better) 11:15:52 tcr: because it's not stack allocated 11:16:45 cracki [n=cracki@41-144.eduroam.RWTH-Aachen.DE] has joined #lisp 11:17:28 nikodemus: But the dynamic-extent is ignored in this optimization settings, so I don't care 11:17:33 stassats: i tried install hutchentoot but it fails :( i cannot install sbcl for some odd reason (get error i don't fathom), clisp and cmucl are working at least. my sbcl install error: http://zu22.pastebin.com/f6a57f57e and my hunchentoot install error: http://zu22.pastebin.com/ffa889d1 11:18:07 lnostdal: if i use mod_proxy it has no awareness of a lisp server though, i mean i still need a lisp web server right? 11:18:20 <_3b> zu22 : that's why we suggest avoiding distro packages :) 11:18:33 Adlai: have you tested it? 11:18:42 yes 11:18:42 with e.g. empty list 11:18:43 _3b: ok so lets say i will now install from source, can you point me at the source packages i should build? :) 11:18:46 well 11:18:49 I've done a few tests 11:19:05 it creates and deletes pairs as expected 11:19:06 if you wrap the make-array in (locally (declare (optimize (safety 0)) ...) or replace SIZE (the (integer 0 #.(* 31 1024)) size) 11:19:10 <_3b> zu22: get a binary from sbcl.org, optionally get newer source and build using that binary 11:19:21 I'm gonna test the &keys now 11:19:27 nikodemus: Related, I think there should be an optimization flag, trust-dx-declarations, so I can see which settings enable it using describe-compiler-policy 11:19:36 zu22, of course .. you need "something" at the lisp end, anyway .. oh, and yeah - don't bother with distro packages for lisp stuff; i've found none of it to work very well at all 11:19:42 <_3b> zu22: then get clbuild, and use that to install libraries/applications (and purge debian stuff to get rid of clc) 11:19:53 tcr: it's not ignored, but because the compiler cannot prove it is safe to stack allocate it is not _trusted_ 11:19:58 (I added an eval-when that asdf:load-ops anaphora unless (find-package :anaphora)) 11:20:03 that's one of the more important dx notes, imo 11:20:09 nikodemus: Ah,ok, that makes a bit more sense 11:20:41 tcr: see sb-ext:*stack-allocate-dynamic-extent* 11:20:45 minion, tell zu22 about clbuild 11:20:46 zu22: please look at clbuild: clbuild [common-lisp.net] is a shell script helping with the download, compilation, and invocation of Common Lisp applications. http://www.cliki.net/clbuild 11:21:03 nikodemus: Yes, I saw that, but I did not understand why it's a global variable, not a optimization flag 11:21:03 Adlai: test with an empty list first 11:21:03 _3b: ok thanks will do that, that debian common-lisp-controller is infamous for causing problems it seems! 11:21:13 stassats, what do you mean? 11:21:25 minion: ok thanks, i want to do lisp the Right Way, i am buying the gigamonkey book today! cant wait for it to arrive! :) 11:21:26 i like lisp... i'm written in it 11:21:31 tcr: because the important use case is setting it to NIL 11:21:37 <_3b> zu22: it is probably better if you can stay with entirely debian stuff, or know how to add to it, but most of us don't use it so would rather not support it :) 11:21:42 ah 11:21:43 i've read so many nice reviews of it 11:21:44 hm 11:21:54 hope a lisp novice like me can learn from it 11:21:56 if that makes the bug go away, you know you probably have a bad dx declaration 11:22:01 -!- ZabaQ [n=johnc@playboxgames.com] has quit [Read error: 110 (Connection timed out)] 11:22:11 sepult [n=user@xdsl-87-78-75-85.netcologne.de] has joined #lisp 11:22:21 _3b: ok thanks, i will just avoid it :) 11:22:38 if it was a policy there'd be all sort of trouble with local trust-dx-declarations 11:22:38 do you guys also strong recommend the gigamonkey book for lisp newbies? 11:22:43 nikodemus: You could get the same effect with (restrict-compiler-policy '(sb-ext:trust-dx-declarations nil)) 11:22:46 zu22, yes. 11:22:51 tic: great 11:23:13 <_3b> zu22: it even more or less says so on the cover :) 11:23:38 tcr: matter of taste, i guess -- plus i didn't want two depend on an experimental feature for something this basic 11:23:44 to, even 11:23:52 mcspiff [n=user@DC7CB.WPA.Dal.Ca] has joined #lisp 11:25:07 we used to have a declaration, but it proved to be harder to use 11:25:43 also, i don't think this is a particularly good match for what declarations mean 11:25:53 (which is the matter of taste bit) 11:26:18 nikodemus: It's also what allegro does, and I like it a lot. I think it's perfect use case for declarations. 11:26:23 CyberBlue [n=yong@60.26.103.13] has joined #lisp 11:26:37 declaration change the behaviour of the compiler 11:27:40 tcr: see here: http://thread.gmane.org/gmane.lisp.steel-bank.devel/11405 11:28:19 _3b: heh 11:28:38 nikodemus: Thanks. I missed that, or didn't particularly interest me back then. I'll read through it later. 11:30:25 dismantr [n=jean-bap@ptr25-1-78-228-5-94.fbx.proxad.net] has joined #lisp 11:31:31 tcr: i dislike having it as a declaration because first you say "stack allocate this", and then you need to make sure you say "trust what i say", and then you need to make sure you don't say to trust things outside your lexical scope, which makes it a lot harder to provide DX bindings from WITH-FOO style macros without unconditionally trusting all dx declarations in their bodies 11:31:47 if we can't get that right consistently, i don't expect our users to 11:31:54 stassats, is there any way to fix that? I've been looking at get-setf-expansion, but it doesn't seem to do the trick. 11:32:00 hence i believe it is a broken mechanism 11:32:29 (by fix, I mean get the right behavior... the obvious "fix" if that's not possible is to throw an appropriate error) 11:32:31 clhs d-s-e 11:32:31 DEFINE-SETF-EXPANDER: http://www.lispworks.com/reference/HyperSpec/Body/m_defi_3.htm 11:32:41 thx 11:33:26 that said, i can guess why ACL may have done that: if they used to never stack allocate stuff, and later added support for it, their customers may have had broken DX declarations, at it may have been deemed less painful to enable/disable things locally then to tell people to fix their code... 11:33:27 clhs acons 11:33:27 http://www.lispworks.com/reference/HyperSpec/Body/f_acons.htm 11:33:34 nikodemus: Uh, well, many macros are broken with respect to declarations. They make macro writing hard, but it's what there. 11:33:55 right, but no need to make it even harder 11:34:57 nikodemus: The macro-writer could use truly-dynamic-extent 11:35:05 tcr: they should not 11:35:09 alright 11:35:32 it is internal for a good reason, because it enables really dangerous stuff that can break the compiler when used with arbitrary code 11:35:34 anyway, it was just not intuitive to me why not make it a declaration. It's a matter of taste, surely 11:35:57 if someone has a large codebase with broken DX declarations that they want to port to sbcl, i'll be happy to provide them with a patch/contrib that enables local control 11:36:31 but i really don't think the local control should be a normal programming tool -- get the declarations right, or delete them 11:36:46 Another thing; do you know the whereabouts of with-ir1-environment-from-node? I added something to with-ir1-environment, and I'm now wondering whethere I have to care for that, too. So long everything seems to work out fine not doing it. 11:37:10 stassats, thx wrt acons as well 11:37:17 *Adlai* will be back shortly. 11:37:48 Basically, when is with-ir1-environment-from-node used? Surely, after macro macro expansion, right? 11:38:13 -!- dismantr [n=jean-bap@ptr25-1-78-228-5-94.fbx.proxad.net] has quit ["ERC Version 5.3 (IRC client for Emacs)"] 11:38:38 during optimization 11:38:52 -!- zu22 [n=ubuntu@unaffiliated/zu22] has left #lisp 11:38:54 but optimization may involve macro-expansion if new lambdas are introduced 11:39:20 dismantr [n=jean-bap@ptr25-1-78-228-5-94.fbx.proxad.net] has joined #lisp 11:39:22 for consistency if nothing else you should also do it in the -NODE version, but possibly you should be hanging it onto something else entirely -- hard to say 11:39:34 rakista [n=rakista@c-71-59-128-255.hsd1.wa.comcast.net] has joined #lisp 11:39:43 It's a hashtable which contains the macroexpansions from user macros 11:39:46 -!- rakista [n=rakista@c-71-59-128-255.hsd1.wa.comcast.net] has quit [Read error: 104 (Connection reset by peer)] 11:39:52 -!- dismantr [n=jean-bap@ptr25-1-78-228-5-94.fbx.proxad.net] has quit [Client Quit] 11:40:06 where is it used? 11:40:36 in get-source-path, and the code-deletion note heuristic 11:41:30 when I added source-paths for atoms, the problem is that for example, FORMAT expands to dead code referencing the format-string 11:42:08 node sure what TRT would be 11:42:13 s/node/not/ 11:42:13 daniel_ [i=daniel@unaffiliated/daniel] has joined #lisp 11:42:35 previously I only looked at the actual form-path of a source-path (what comes after the 'original-source sentinel) 11:43:25 I know also look at what comes before, and only accepts if what becomes before 'original-source is from the original-source, or expansion of user macros 11:43:31 s/know/now/ 11:44:22 -!- daniel [i=daniel@unaffiliated/daniel] has quit [Read error: 60 (Operation timed out)] 11:44:37 -!- blackwolf [n=blackwol@ool-45763541.dyn.optonline.net] has quit [Remote closed the connection] 11:44:52 merimus [n=wroth@nat-49.laurelnetworks.com] has joined #lisp 11:46:38 -!- oudeis [n=oudeis@p11811120.orange.net.il] has quit [Read error: 110 (Connection timed out)] 11:47:53 -!- daniel_ is now known as daniel 11:49:00 uh, package-data-list is a nightmare for the fontification code 11:49:54 ejs1 [n=eugen@77.222.151.102] has joined #lisp 11:50:39 -!- sellout [n=greg@c-24-128-50-176.hsd1.ma.comcast.net] has quit [] 11:50:42 ruediger [n=ruediger@62-47-150-223.adsl.highway.telekom.at] has joined #lisp 11:51:10 -!- mcspiff [n=user@DC7CB.WPA.Dal.Ca] has quit [Remote closed the connection] 11:51:57 -!- sepult [n=user@xdsl-87-78-75-85.netcologne.de] has quit ["ERC Version 5.3 (IRC client for Emacs)"] 11:54:34 -!- plage [n=user@118.68.38.106] has left #lisp 11:55:27 -!- udzinari [n=udzinari@94-43-70-11.dsl.utg.ge] has left #lisp 11:58:15 mcspiff [n=user@DC7CB.WPA.Dal.Ca] has joined #lisp 11:58:24 tsuru [n=user@c-69-245-36-64.hsd1.tn.comcast.net] has joined #lisp 11:59:48 -!- ejs [n=eugen@nat.ironport.com] has quit [Read error: 110 (Connection timed out)] 12:00:06 *Adlai* has to eat, but promises a bug-free ALREF soon. 12:01:16 -!- CyberBlue [n=yong@60.26.103.13] has quit ["Leaving"] 12:02:45 -!- Jasko [n=tjasko@c-98-235-105-148.hsd1.pa.comcast.net] has quit ["Leaving"] 12:03:53 -!- blandest [n=blandest@softhouse.is.ew.ro] has quit [Read error: 110 (Connection timed out)] 12:04:33 Jasko [n=tjasko@c-98-235-105-148.hsd1.pa.comcast.net] has joined #lisp 12:06:41 -!- meingbg [n=user@173-45-238-108.slicehost.net] has quit [Remote closed the connection] 12:06:50 meingbg [n=user@173-45-238-108.slicehost.net] has joined #lisp 12:08:22 aunwork [n=aunwork@mail.ionicsoft.com] has joined #lisp 12:08:50 blandest [n=blandest@softhouse.is.ew.ro] has joined #lisp 12:11:05 -!- meingbg [n=user@173-45-238-108.slicehost.net] has quit [Read error: 54 (Connection reset by peer)] 12:11:07 meingbg [n=user@173-45-238-108.slicehost.net] has joined #lisp 12:13:00 Odin- [n=sbkhh@adsl-2-92.du.snerpa.is] has joined #lisp 12:13:04 -!- jewel [n=jewel@dsl-242-129-65.telkomadsl.co.za] has quit [Read error: 113 (No route to host)] 12:13:19 amnesiac [n=amnesiac@p3m/member/Amnesiac] has joined #lisp 12:13:33 vy [n=user@nbvyazici.cs.bilkent.edu.tr] has joined #lisp 12:14:40 -!- ruediger [n=ruediger@62-47-150-223.adsl.highway.telekom.at] has quit ["Leaving"] 12:19:17 -!- ken-p [n=ken-p@84.92.70.37] has quit [Success] 12:22:29 -!- Athas [n=athas@0x50a157d6.alb2nxx15.dynamic.dsl.tele.dk] has quit [Connection timed out] 12:22:51 -!- ejs1 [n=eugen@77.222.151.102] has quit ["This computer has gone to sleep"] 12:24:11 nikodemus: I've sent another bug report to -devel ml about profiling threads stuff. FYI. 12:25:16 -!- amnesiac [n=amnesiac@p3m/member/Amnesiac] has quit [Remote closed the connection] 12:25:21 pjb: are you there? 12:25:24 Yep. 12:25:30 (eq matimago pjb) 12:25:59 amnesiac [n=amnesiac@p3m/member/Amnesiac] has joined #lisp 12:26:06 lol 12:26:31 -!- ausente is now known as dalton 12:27:09 -!- tseug [n=tseug3@cpe-70-112-21-137.austin.res.rr.com] has quit [Read error: 110 (Connection timed out)] 12:27:10 quek [n=read_eva@router1.gpy1.ms246.net] has joined #lisp 12:28:11 can someone help me what pjb was trying to say in http://www.rhinocerus.net/forum/lang-lisp/572467-improve-efficiency-simple-function.html 12:28:46 leo2007: do you know what happens in C when you add beyond MAX_UINT ? 12:29:13 Not exactly sure 12:29:21 become negative? 12:29:48 More or less. Depends on the variable MAX_UINT needs a unsigned long, so it would always mean positive, but it would indeed roll over. 12:29:58 It's modulo 2^32 arithmetic. 12:30:32 Actually, C falls back to what the processor provides, and the modern processors provide modulo 2^n arithmetic. 12:30:36 but in my function c has a upper bould that prevents it from going beyond that 12:30:48 That's not the point. 12:30:58 leo2007: (integer-length 4294967295) => 32 12:31:18 The point is that when you write c=a+b; in C, it's actually equivalent to (setf c (logand #xffffff (+ a b))) in lisp. 12:31:42 and not (setf c (+ a b)) ; in lisp, the arithmetic is much closer to normal integer arithmetic. 12:32:21 (setf c (logand #xffffffff (+ a b))) or (setf c (mod (+ a b) #x100000000)). 12:32:25 C specifies it to be modulo ULONG_MAX+1, which is all you need to care about. 12:32:27 (I missed some fs first. 12:32:30 sellout [n=greg@guest-fw.dc4.itasoftware.com] has joined #lisp 12:33:21 leo2007: so when you only write (setf c (+ a b)) in lisp you are asking more than what the modern processors provide, you're asking to test for processor ALU overflow, and switch to bignum arithmetic. This is more work, therefore slower. 12:34:05 matimago: which line in the code I need to change? 12:34:12 leo2007: all :-) 12:34:17 -!- cky [n=cky@cpe-024-211-249-162.nc.res.rr.com] has quit ["Restarting IRC client :-)"] 12:34:21 -!- mcspiff [n=user@DC7CB.WPA.Dal.Ca] has quit [Read error: 60 (Operation timed out)] 12:34:45 cky [n=cky@cpe-024-211-249-162.nc.res.rr.com] has joined #lisp 12:35:33 The second thing you missed, is that a type declaration in lisp doesn't change the type of the values, and doesn't prevent you or the program to bind values out of that type to the declared variables. The treatment of a type declaration is implementation dependant, and it's more like a promise YOU give to the compiler than anything else. However, sbcl may be more helpful there, check it's manual. 12:36:12 (defun +/32 (a b) (logand #xffffffff (+ a b))) (defun -/32 ...) and use +/32 instead of +, etc. 12:36:15 ok, that's good tip 12:36:52 Notice that sbcl understand this logand, and compiles the expresion to a single processor ADD instruction. 12:38:05 matimago: are you serious all needs to change in the code? 12:38:19 leo2007: no, you can be smart. 12:38:38 no I'm not 12:38:53 leo2007: you may shadow +, - ,etc and (defun + (&rest args) (logand #xffffffff (apply (function cl:+) args))) 12:39:15 But you have to be careful if you really need modulo arithmetic in your whole package. 12:40:07 oudeis [n=oudeis@bzq-79-182-115-160.red.bezeqint.net] has joined #lisp 12:40:17 there are 4 places that could happen. it seems to me only the assignment to x needs some care 12:40:38 There. 12:40:45 Smarts growing. 12:41:13 Athas [n=athas@0x50a157d6.alb2nxx15.dynamic.dsl.tele.dk] has joined #lisp 12:42:37 -!- oudeis [n=oudeis@bzq-79-182-115-160.red.bezeqint.net] has quit [Client Quit] 12:43:00 could you help annotate this paste http://paste.lisp.org/display/82281 ? 12:43:13 What's the point of that exercise anyway? 12:43:23 I still have a lot to learn about cl 12:43:45 tcr: I want a few rngs for a simulation 12:43:52 oudeis [n=oudeis@bzq-79-182-115-160.red.bezeqint.net] has joined #lisp 12:44:02 leo2007: It's probably better to declare things directly in the scope of declaration. Move the declare for c and i up... 12:44:44 -!- blandest [n=blandest@softhouse.is.ew.ro] has quit [Read error: 110 (Connection timed out)] 12:44:49 done 12:44:50 Replace incf by a logand 1+ 12:45:17 and logand t2 too. 12:45:55 r is larger than x 12:45:56 existentialmonk [n=carcdr@64-252-41-17.adsl.snet.net] has joined #lisp 12:46:05 -!- A_anekos is now known as anekos 12:46:20 i could be declared (unsigned-byte 16); in some processors it may give better code, indexing only 16-bit. 12:46:39 leo2007: are you sure the compile can deduce that itself? 12:46:42 compiler 12:47:04 something like this (setf c (logand #xffffffff (1+ c))) ? 12:47:06 -!- Edward_ [i=Ed@AAubervilliers-154-1-43-234.w90-3.abo.wanadoo.fr] has quit ["L'oignon fait la farce."] 12:47:10 Yes. 12:47:22 Of course, the problem is that (+ i 1) wouldn't be (unsigned-byte 16) :) 12:47:34 Then check the results with (disassemble (compile 'random-cmcw)) 12:47:36 Edward_ [n=Ed@AAubervilliers-154-1-43-234.w90-3.abo.wanadoo.fr] has joined #lisp 12:47:51 Since i<4096, we can safely logand #xffff that increment. 12:48:00 So you'd probably want a lot of (the ...)s in there to make promises about the results of the operations. 12:48:16 Well, it's already logand'ed by 4095. 12:48:51 *matimago* bbl 12:49:04 disassemble output http://paste.lisp.org/display/82281#1 12:49:45 ryepup [n=ryepup@one.firewall.gnv.acceleration.net] has joined #lisp 12:50:32 -!- amnesiac [n=amnesiac@p3m/member/Amnesiac] has quit ["This computer has gone to sleep"] 12:50:57 Zhivago: Could you help me with that? i.e. adding the ... 12:51:29 for example, there's one note in (setf t1 (logand #xffffffffffffffff (+ (* (aref Q i) a) c))) 12:51:42 Well, (the type foo) is a promise that foo will satisfy the type 'type'. 12:51:58 (the (unsigned-byte 16) i), for example. 12:53:29 I can handle for simple case like that but not the one in t1 12:53:51 actually there's no inefficiency in the i setting 12:54:10 stassats` [n=stassats@wikipedia/stassats] has joined #lisp 12:55:57 envi^home [n=envi@220.121.234.156] has joined #lisp 12:56:56 the compiler output is here http://paste.lisp.org/display/82281#2 12:57:29 amnesiac [n=amnesiac@p3m/member/Amnesiac] has joined #lisp 12:58:06 mcspiff [n=user@wifi-c102.EE.Dal.Ca] has joined #lisp 12:59:29 So, why is the result a (MOD 80668075734691)? 12:59:53 because t1 is 64-bit 13:00:08 the number is a*2^32 13:00:17 a is 18782 13:00:47 So, what is the problem with that compiler note? 13:01:18 there a list of beginner sbcl bugs somewhere that softfoots might be able to tackle? 13:01:45 I hope not :) 13:01:45 Zhivago: I can't figure it out 13:01:57 leo: Ok -- what surprises you about that compiler note? 13:02:20 many costs that slow things down 13:02:30 leo: Do you understand that note? 13:02:33 features/projects? 13:02:53 documentation *shudder* 13:02:55 I can not figure out by reading the notes where to improve the code 13:03:28 Start by trying to understand what that note means ... 13:03:34 that skill is crucial to me 13:04:30 Zhivago: could you help me understand one and I think I can survive thereafter 13:04:55 for the one started with " forced to do GENERIC-* (cost 30) ..." 13:05:26 merimus: look at the bug tracker 13:05:31 merimus: there are many 13:05:41 some are even already tagged "easy" 13:05:53 tcr: thank you 13:06:00 leo: Which part of "because the first argument is a (UNSIGNED-BYTE 32), not a FIXNUM" confuses you? 13:06:17 blandest [n=blandest@85.204.33.242] has joined #lisp 13:06:22 CyberBlue [n=yong@60.26.103.13] has joined #lisp 13:07:16 that is fine, I understand that, but what to do to improve? 13:07:29 all are uint-32 except t1 13:07:32 Well, do you need (unsigned-byte 32)? 13:07:33 which is 64 13:08:06 If so, then you can't use fixnum, and that's just the way it is. 13:08:24 -!- stassats [n=stassats@wikipedia/stassats] has quit [Read error: 110 (Connection timed out)] 13:09:03 Q is uint-32 random number array 13:09:38 Well, then you'll just have to accept that you can't do fixnum arithmetic on it ... 13:10:19 On its elements, even. 13:10:32 sepult [n=user@xdsl-87-78-75-85.netcologne.de] has joined #lisp 13:11:40 if I get rid of that t1 assignment, the program runs 15 times faster 13:14:19 From (info "(sbcl)Modular arithmetic"), it seems only uint-32 will have Modular arithmetic optimisation 13:17:39 -!- quek [n=read_eva@router1.gpy1.ms246.net] has left #lisp 13:19:11 -!- simplechat [n=simplech@unaffiliated/simplechat] has quit [Remote closed the connection] 13:19:21 jajcloz [n=jaj@pool-98-118-48-50.bstnma.fios.verizon.net] has joined #lisp 13:22:39 -!- mcspiff [n=user@wifi-c102.EE.Dal.Ca] has quit [Read error: 104 (Connection reset by peer)] 13:24:08 -!- wedgeV_ [n=wedgeV@85.31.0.85] has quit [] 13:24:51 bombshelter13_ [n=bombshel@toronto-gw.adsl.erx01.mtlcnds.ext.distributel.net] has joined #lisp 13:28:07 -!- vy [n=user@nbvyazici.cs.bilkent.edu.tr] has quit [Remote closed the connection] 13:30:43 wedgeV [n=wedgeV@85.31.0.85] has joined #lisp 13:31:20 is the sourceforge cvs the official repo? 13:33:28 c|mell [n=cmell@p6037-ipngn101marunouchi.tokyo.ocn.ne.jp] has joined #lisp 13:33:42 merimus: for sbcl, yeah. many people (including some sbcl developers themselves) find it easier to use the git repo, though: http://git.boinkor.net/gitweb/sbcl.git 13:36:30 -!- sepult [n=user@xdsl-87-78-75-85.netcologne.de] has quit ["ERC Version 5.3 (IRC client for Emacs)"] 13:37:32 sepult [n=user@xdsl-87-78-75-85.netcologne.de] has joined #lisp 13:42:09 -!- c|mell [n=cmell@p6037-ipngn101marunouchi.tokyo.ocn.ne.jp] has quit [Read error: 104 (Connection reset by peer)] 13:47:22 -!- mega1 [n=mega@pool-01811.externet.hu] has quit [Read error: 110 (Connection timed out)] 13:50:25 Ringo48 [n=Ringo48@97-122-180-214.hlrn.qwest.net] has joined #lisp 13:53:26 is there a factorial function around? 13:54:17 -!- Adamant_ [n=Adamant@unaffiliated/adamant] has quit [] 13:54:22 jewel [n=jewel@dsl-242-129-65.telkomadsl.co.za] has joined #lisp 13:55:16 is that even useful ? 13:56:20 willb [n=wibenton@h69-129-204-3.mdsnwi.broadband.dynamic.tds.net] has joined #lisp 13:58:22 schoppenhauer [n=senjak@unaffiliated/schoppenhauer] has joined #lisp 13:58:43 which? 13:59:49 the factorial function ? 14:00:08 to some applications, it is useful. 14:00:36 attila_lendvai [n=ati@catv-89-132-189-132.catv.broadband.hu] has joined #lisp 14:01:51 leo2007: #+clisp (ext:! 10) 14:01:54 clop [n=jared@99-23-192-153.lightspeed.austtx.sbcglobal.net] has joined #lisp 14:02:02 alexandria:factorial 14:02:33 leo2007: otherwise, I think fact is the function most often as an example everywhere, you should have no problem coming with the sources of a good version of it. 14:02:41 i thought functions like that were only used in books, but I guess I was wrong since it's already implemented in clisp :) 14:02:55 matimago: a good factorial for large numbers is a bit tricky 14:03:02 Indeed. 14:03:19 But IIRC good hints were given on cll a few years ago. 14:03:49 matimago: alexandria's one should not suck too badly, and is portable 14:04:19 ok. 14:05:02 Nshag [i=user@Mix-Orleans-106-4-142.w193-248.abo.wanadoo.fr] has joined #lisp 14:05:08 it doesn't do the prime stuff, but it's smart enough to avoid extraneous bignum arithmetic 14:06:03 thank you all, I'm mearely curious about it tho. Emacs calc can do that already. 14:07:39 seejay [n=seejay@unaffiliated/seejay] has joined #lisp 14:08:11 phf [n=phf@host.icnfull.com] has joined #lisp 14:09:31 -!- amnesiac [n=amnesiac@p3m/member/Amnesiac] has quit ["This computer has gone to sleep"] 14:11:15 nikodemus: Is there a good way to see whether a lambda came from a binding operator? 14:11:31 brickhazel [n=brickhaz@63.144.132.78] has joined #lisp 14:11:43 -!- schoppenhauer [n=senjak@unaffiliated/schoppenhauer] has left #lisp 14:11:48 ygi [i=foobar@gateway.blegkilde.dk] has joined #lisp 14:12:45 c|mell [n=cmell@p31046-ipngn101marunouchi.tokyo.ocn.ne.jp] has joined #lisp 14:12:55 ejs [n=eugen@252-237-124-91.pool.ukrtel.net] has joined #lisp 14:14:17 -!- sepult [n=user@xdsl-87-78-75-85.netcologne.de] has quit [Read error: 104 (Connection reset by peer)] 14:15:07 tcr: do you mean from defun or defmacro? 14:15:13 -!- brickhazel [n=brickhaz@63.144.132.78] has quit [Client Quit] 14:15:15 Jabberwockey [n=jens@port-9135.pppoe.wtnet.de] has joined #lisp 14:15:19 or also flet / labels? 14:16:11 On thing that is specified for these lambda is that they contain a block named as the function. If function-lambda-expression returns it, you can see it there. 14:16:16 s/On/One/ 14:16:17 I mean LET, LET*, multiple-value-bind 14:16:52 sepult [n=user@xdsl-87-78-131-34.netcologne.de] has joined #lisp 14:16:58 though I take that request back, I think I found a better way anyway 14:17:13 ok, because then I'd be even more puzzled. 14:18:39 SBCL translates LET to ((lambda ...) ...), and LET* to ((lambda (&aux ...) ...) 14:19:02 well, essentially anyway 14:19:41 -!- cracki [n=cracki@41-144.eduroam.RWTH-Aachen.DE] has quit ["If technology is distinguishable from magic, it is insufficiently advanced."] 14:19:54 And you archive getting a reference to such a lambda? 14:20:07 KingNatoG5_ [n=patrik@84-217-7-184.tn.glocalnet.net] has joined #lisp 14:20:20 I would have thought they'd be optimized out right away. 14:20:59 sure when hacking at the first-pass of the compiler :) 14:21:08 s/at/on/ 14:21:20 ok. 14:21:27 schoppenhauer [n=senjak@unaffiliated/schoppenhauer] has joined #lisp 14:23:39 postamar [n=postamar@x-132-204-254-17.xtpr.umontreal.ca] has joined #lisp 14:24:00 amnesiac [n=amnesiac@p3m/member/Amnesiac] has joined #lisp 14:25:05 mmm... it seems that with sbcl's generational garbage collector, it's incapable of purifying the memory space at all before saving the lisp machine, so that every exported core or executable contains lots of garbage? 14:25:51 ejs1 [n=eugen@nat.ironport.com] has joined #lisp 14:26:02 roidrage [n=roidrage@p54BEB25E.dip0.t-ipconnect.de] has joined #lisp 14:26:16 -!- Davse_Bamse [n=davse@130.226.210.2] has quit ["leaving"] 14:26:21 how did you come to that conclusion? 14:26:23 sytse: that doesn't sound right. 14:26:57 purify is a no-op on arches with the generational garbage collector, that much is true 14:28:00 teilzeitstudent [n=teilzeit@p5B17C49B.dip.t-dialin.net] has joined #lisp 14:28:15 sytse: perhaps purify doesn't mean what you think it means? 14:28:18 and I tried uninterning some symbols, doesn't decrease the resulting core size (of sb-ext:save-lisp-and-die), and running some forms that aren't supposed to leave anything behind (like, disassemble) before saving lisp, and that actually increases the core size 14:28:52 luis: oh indeed, purify only marks non-garbage as definitely non-garbage 14:29:00 there you go. 14:29:16 -!- KingNatoG5 [n=patrik@84-217-14-183.tn.glocalnet.net] has quit [Read error: 110 (Connection timed out)] 14:29:23 but there's apparently no way to move objects around when saving lisp? 14:29:43 stepnem_ [i=stepnem@173-19-7-99.client.mchsi.com] has joined #lisp 14:29:54 sytse: move them where? 14:30:09 -!- ejs [n=eugen@252-237-124-91.pool.ukrtel.net] has quit [Read error: 104 (Connection reset by peer)] 14:30:42 luis: compact the memory space 14:31:22 so that when saving the lisp machine, it saves a contiguous (except for padding for efficiency reasons) blob of useful data, with no garbage inbetween 14:32:21 it apparently doesn't, or else uninterning functions/macros would have made the core smaller, and running lisp code that doesn't save anything into an accessible location wouldn't make it bigger 14:32:39 sbcl's gc is also conservative, which prevents it from moving things around. 14:32:39 or am I missing something here? 14:33:10 i can tell you it does; maybe you don't have as complete a picture of the heap as you think. 14:33:12 kuwabara: not quite true. It's only conservative wrt stack references pretty much. 14:33:44 and not conservative at all when saving a core. 14:33:52 so it can move a function around and then adjust pointers in code? 14:33:55 kuwabara: and it's only conservative on x86 14:33:59 luis: oh, so when it save-lisp-and-die, it could assume it's precise, and move things around... 14:34:09 oh btw, I should've mentioned: I'm using amd64 14:35:09 sytse: yes it moves code around. 14:35:12 -!- ejs1 [n=eugen@nat.ironport.com] has quit [Read error: 60 (Operation timed out)] 14:35:38 after using lots of logands in http://paste.lisp.org/display/82281#3, the speed is still less than optimal 14:35:47 luis: then I'm surprised that it can't clean up after disassemble, for example.. 14:35:51 any suggestion on further improvement 14:35:52 just inserting a 14:35:53 (disassemble #'fun :stream *error-output*) 14:36:03 before save-lisp-and-die 14:36:10 makes the core bigger 14:36:12 sytse: clean what up? 14:36:21 *, *, ***, etc and caches. 14:36:42 well, caches for example ;-) 14:37:33 and btw, I tried loading a compiled file with functions and macros in it: uninterning the functions and macros before save-lisp-and-die can't make the core smaller, only bigger 14:38:13 what does uninterning a function mean? 14:38:15 as in, (let ((pkg (find-package 'pkg))) (unintern (find-symbol "FUN" pkg) pkg)) 14:38:24 ejs1 [n=eugen@91.124.237.252] has joined #lisp 14:38:48 sytse: the function is probably still callable from other code, for instance. 14:38:53 sure, use fmakunbound 14:38:56 tcr: it means that you remove the symbol pointing to it in a package, thus making it inaccessible if it isn't referenced in some other accessible function 14:39:11 sytse: it's probably referenced in the global info db 14:39:14 luis: even if it weren't, you'd have to fmakunbound. The info db uses strong references. 14:40:02 sytse: it's the reason why DELETE-PACKAGE will probably do much less than people may suspect 14:40:04 how much bigger it became? enough for you to care? 14:40:06 damn. 14:40:08 I feel so stupid 14:40:20 tcr: global info db.. :( 14:40:41 how do you remove something from there? (anyway, fmakunbound works indeed) 14:43:00 -!- ruepel0r [n=rue@203-97-49-162.dsl.clear.net.nz] has quit [Read error: 110 (Connection timed out)] 14:43:48 elliotstern [n=chatzill@cpe-67-240-162-85.rochester.res.rr.com] has joined #lisp 14:44:01 sytse: ah, what pkhuong and tcr said is a better explanation. Though, if you had a reference to #'fun elsewhere that would prevent collection as well. 14:44:41 cinayakoshka [n=bar@adaptive.cs.unm.edu] has joined #lisp 14:45:00 if delete-packagage desn't deleta a package, but keeps poinnters, SBCL sucks in that area 14:45:14 (weak pointers would have been fine) 14:45:30 No, you don't understand what delete-package does 14:45:35 -!- ygi [i=foobar@gateway.blegkilde.dk] has quit [Read error: 60 (Operation timed out)] 14:46:04 luis: I don't. 14:47:01 sytse: then that's why fmakunbound works for you. 14:47:25 *sytse* reads compiler/globaldb.lisp 14:47:26 tcr: explain. Forget the standard and start with the SBCL spesifics. Please. 14:49:14 It's all about the standard 14:49:32 It precisely specifies what delete-package is supposed to do 14:50:05 Odin-MAC [n=sbkhh@adsl-2-92.du.snerpa.is] has joined #lisp 14:52:12 Jonathan-Smith [n=jon@ip24-250-13-137.ri.ri.cox.net] has joined #lisp 14:52:43 -!- adityo_ [n=adityo@122.169.38.55] has quit ["leaving"] 14:52:57 I concur with tcr 14:53:28 -!- cinayakoshka [n=bar@adaptive.cs.unm.edu] has quit ["Leaving"] 14:54:08 stuart71 [n=soverton@75-145-221-229-Memphis.hfc.comcastbusiness.net] has joined #lisp 14:54:16 -!- drafael [n=tapio@ip-118-90-139-108.xdsl.xnet.co.nz] has quit [Read error: 110 (Connection timed out)] 14:55:35 I am using clx under sbcl. Can anyone tell me how to (or if I can) use clue and clio? 14:55:49 -!- Odin- [n=sbkhh@adsl-2-92.du.snerpa.is] has quit [Read error: 110 (Connection timed out)] 14:57:47 ygi [i=foobar@gateway.blegkilde.dk] has joined #lisp 14:58:01 yes, you can use them, though they are a bit dated 14:58:19 jthing: have you tried using them? 14:59:47 I'm open to suggestions on alternatives, too. I really like mcclim, but it does take ~20sec to load on my machine. 15:00:10 mattrepl_ [n=mattrepl@pm2-users.caci.com] has joined #lisp 15:00:10 what is the most efficient way of getting the higher 32 bit of a 64-bit integer? 15:00:46 -!- mattrepl [n=mattrepl@hq-users.caci.com] has quit [Read error: 104 (Connection reset by peer)] 15:00:55 mattrepl [n=mattrepl@hq-users.caci.com] has joined #lisp 15:01:29 -!- CyberBlue [n=yong@60.26.103.13] has quit ["Leaving"] 15:01:48 -!- ASau [n=user@host169-231-msk.microtest.ru] has quit ["off"] 15:03:03 -!- teilzeitstudent [n=teilzeit@p5B17C49B.dip.t-dialin.net] has quit ["Leaving"] 15:03:19 -!- Jonathan-Smith [n=jon@ip24-250-13-137.ri.ri.cox.net] has quit ["Leaving."] 15:03:27 -!- ejs1 [n=eugen@91.124.237.252] has quit ["This computer has gone to sleep"] 15:03:40 I started writing my own widget toolkit, and have had some success, but I hate to reinvent the wheel, especially when my wheel is probably not nearly as well thought-out as some other rolling around out there that I don't know about :) 15:05:13 ejs0 [n=eugen@252-237-124-91.pool.ukrtel.net] has joined #lisp 15:05:24 On the upside, I now have a much greater respect for the design of the x window system. 15:05:36 Jonathan-Smith [n=jon@ip24-250-13-137.ri.ri.cox.net] has joined #lisp 15:06:04 leo2007: (ldb (byte 32 32) my64) 15:06:40 ikki [n=ikki@201.155.75.146] has joined #lisp 15:07:15 nyef [n=nyef@65-125-125-98.dia.static.qwest.net] has joined #lisp 15:07:19 -!- mattrepl_ [n=mattrepl@pm2-users.caci.com] has quit [Read error: 60 (Operation timed out)] 15:08:11 -!- stassats` [n=stassats@wikipedia/stassats] has quit [Read error: 60 (Operation timed out)] 15:11:30 btw, is there a way to ask sbcl about the details of specific objects, like, for example, their machine-dependent size in bytes? 15:11:51 stuart71: why not work with CLIM? :D 15:12:35 or with gtk 15:12:41 or cells-gtk 15:12:43 or ....... 15:12:46 kuwabara: that should be about the same speed as (ash my64 -32), right? 15:12:55 sytse: it has been requested, see the comment in sb-introspect:allocation-information 15:13:17 Adamant [n=Adamant@c-76-29-188-60.hsd1.ga.comcast.net] has joined #lisp 15:13:29 or clg. or commonqt ;-) 15:13:31 p_l: I love CLIM, i just have a couple of issues with the load time and I can't figure how to do any focus change without clicking. 15:14:12 ...and the number of dependencies 15:14:50 leo2007: ldb may be faster on 32 bits arch 15:15:15 leo2007: no, (ash my64 -32) is *MUCH* more efficient on sbcl 15:15:31 at least on amd64 (where I can test it) 15:15:45 ace4016 [i=ace4016@cpe-76-168-248-118.socal.res.rr.com] has joined #lisp 15:16:17 sytse & p_l: do you guys do much gui development in common-lisp? If so, what do you use? 15:16:35 *sytse* doesn't at all 15:16:39 I just started learning lisp :P 15:16:57 been doing lisp for about 10 to 15 days now 15:18:10 Really, it does seem as if mcclim is the most active, am i right? 15:18:14 just wrote a very nice getopt parser though, which should get most argp functionality (and already has a compile-time abbreviation finder using a binary tree to find common bases.. which might be a bit overkill for anything below 500 options, but ah well) 15:18:15 uh that's good progress, sytse 15:18:21 thx :) 15:18:45 hey sytse have you seen wclas http://cl-www.msi.co.jp/projects/wclas/index.html 15:18:47 *sytse* is always fast when learning new languages, because I immediately dive into the specs 15:19:03 -!- mvilleneuve [n=mvillene@ABordeaux-253-1-5-135.w82-125.abo.wanadoo.fr] has quit [Read error: 113 (No route to host)] 15:19:13 I'd read through PCL first, then go to the hyperspec 15:19:18 *stuart71* wishes he could learn that quickly 15:19:29 PCL is much better at teaching the basics of CLOS, and also LOOP 15:19:32 yes, PCL I read about 70% of 15:19:44 that's the first thing I did, it's a great book 15:19:59 tcr: and of macros 15:20:05 paste some code if you want critique 15:20:24 will do later 15:21:02 sytse: There's a way to ask for the size in octets of a given object in SBCL, but it's rarely what you want, as it ignores any objects that would be considered "components" of the object asked about. 15:21:49 Any given CONS, for example, is always two words, but if you're wondering about a -list-, then you need to do the iteration yourself. And arrays, classes, structures, and so on get complicated fast. 15:21:51 sohail [n=Sohail@unaffiliated/sohail] has joined #lisp 15:22:04 nyef: of course, same thing in any other language I guess 15:22:16 Right. 15:22:26 sizeof(some_cplusplus_class) is only informative if you know what you're doing too 15:22:46 nyef: Would be cool if you could hack it up (see sb-introspect:allocation-information.) I'd add it to the slime inspector then 15:22:48 but, how do you do it? It isn't documented I think? 15:23:06 -!- blandest [n=blandest@85.204.33.242] has quit ["Leaving."] 15:23:24 I'd cheerfully call GC internals to determine the size, but there may be an easier way. 15:23:37 (And to start finding that easier way, I'd look at ROOM...) 15:24:00 -!- ia [n=ia@89.169.189.230] has quit [Connection timed out] 15:25:13 ia [n=ia@89.169.189.230] has joined #lisp 15:25:31 only thing I see is sb!c::sb-allocated-size.. 15:25:39 but that's not it 15:26:06 -!- Fufie [n=poff@Gatekeeper.vizrt.com] has quit ["Leaving"] 15:26:41 -!- stepnem [n=stepnem@topol.nat.praha12.net] has quit [Read error: 113 (No route to host)] 15:28:29 nyef: nice quote that I just can't pass: "<1> our company moved our MRP system from a reliable as/400 based system to a java platform this weekend. <2> i bet that turned out swell for you <1> the server went down a half hour ago" 15:28:48 Nice. 15:29:05 someone who knows iolib? src/syscalls/ffi-wrappers-unix.grovel-tmp.lisp (which is autogenerated by cffi-grovel) is a file which contains ... -- so the reader chokes on it 15:29:40 it contains three dots 15:29:56 yeah tcr cffi-wrappers is not safe outside with-standard-io-syntax 15:29:57 Heh. Someone set print-level? 15:30:14 we had the same problem at work 15:30:14 nyef: Good catch! 15:30:37 Yeah I do because otherwise I'd have to wait minutes for some of the debugging output come across 15:31:03 c|mell: will you bug fe[nl]ix about it? 15:31:27 -!- ygi [i=foobar@gateway.blegkilde.dk] has quit [Read error: 110 (Connection timed out)] 15:31:29 mega1 [n=mega@3e44af28.adsl.enternet.hu] has joined #lisp 15:31:58 cenan [n=vircuser@88.240.31.213] has joined #lisp 15:32:18 mega1: Are you the person to ask about race conditions involving asynchronous signals and GC invarients? 15:32:32 I don't know what you mean. 15:32:36 -!- cenan [n=vircuser@88.240.31.213] has quit [Client Quit] 15:32:45 in fact, I'm only passing by :-) 15:32:55 Heh. 15:33:02 tell me that you have a mutant tree 15:33:02 hjpark [n=user@221.138.197.236] has joined #lisp 15:33:08 or a mutant OS. 15:33:27 mutant CPU 15:33:31 Sorry, no can do. 15:33:37 athos [n=philipp@92.250.250.68] has joined #lisp 15:34:01 On the upside, the question is more of a "why the hell -isn't- this a problem" thing than a "this is a problem, how can it be fixed" thing. 15:34:08 sytse: thanks for that 15:34:21 milanj [n=milan@79.101.231.108] has joined #lisp 15:34:42 nyef: that's totally unexpected. Please ask. 15:34:57 -!- schoppenhauer [n=senjak@unaffiliated/schoppenhauer] has left #lisp 15:35:09 Essentially, on cheneygc, there's a 1-2 instruction window (depending on the presence of branch-delay slots) wherein the program counter does not point within the range specified by reg_CODE, and a worse window on return. 15:35:27 do you know why some of the functions by http://www.cl-user.net/asp/libs/num-recipes-olum 15:35:30 This strikes me as a race condition waiting to happen, and I'm unable to find out why it wouldn't be. 15:35:31 are defined using defvar 15:36:04 cenan [n=vircuser@88.240.31.213] has joined #lisp 15:36:11 -!- alinp [n=alinp@86.122.9.2] has quit [No route to host] 15:36:42 nyef: where is that window? 15:36:44 If an interrupt happens in one of these windows, and the handler invokes a GC, and the two code segments move relative to each other, doesn't this derange up to three (depending on platform) pointers due to the way scavenge_interrupt_context works? 15:36:57 -!- elliotstern [n=chatzill@cpe-67-240-162-85.rochester.res.rr.com] has quit [Read error: 113 (No route to host)] 15:37:22 -!- aerique [i=euqirea@xs2.xs4all.nl] has quit ["..."] 15:37:38 On call, it's the branch delay slot for the call instruction (optional) and before executing the first instruction of the XEP (which sets reg_CODE correctly). 15:38:08 On return, the window varies by backend, return convention, how many values expected, etc. 15:38:29 -!- stuart71 [n=soverton@75-145-221-229-Memphis.hfc.comcastbusiness.net] has left #lisp 15:39:17 -!- jlf [n=user@unaffiliated/jlf] has quit [Read error: 104 (Connection reset by peer)] 15:40:30 nyef: doesn't scavenge_interrupt_context in cheneygc.c fix the registers up? 15:40:38 (An answer of "asynchronous interrupts don't happen often enough and don't trigger GC often enough for this to have been noticed yet" would be believable, if not exactly what I'm looking for.) 15:40:47 jlf [n=user@unaffiliated/jlf] has joined #lisp 15:40:54 AFAICT, not if reg_CODE is wrong. 15:41:47 dean_ [i=4b107cb6@gateway/web/freenode/x-284627a82af23b21] has joined #lisp 15:42:20 (+ 2 2) 15:44:41 4 15:45:05 nyef: looks broken on a quick glance 15:45:14 Oh, good. I'm not going insane, then. 15:45:26 krumholt [n=krumholt@port-92-193-115-209.dynamic.qsc.de] has joined #lisp 15:45:26 but I can't believe it is. 15:45:37 mega1: nice work with stack scrub scrubbing 15:45:43 I actually can believe it is. 15:45:46 it would have been triggerred many times. 15:45:57 Not really, look at the requirements for triggering it. 15:46:16 nikodemus: let's wait for the regressions ;-) 15:46:16 1.) An asynchronous interrupt has to hit in a vulnerable window. 15:46:37 2.) The interrupt has to cause a GC. 15:47:10 3.) The GC has to relocate the two code-components relative to each other. 15:47:33 isn't 3.) practically every time? 15:47:58 a fixed point analysis is in order for cheneygc 15:48:16 It's 2 that I think is less likely. 15:48:34 but it's easy to test. 15:49:05 *nyef* doesn't have a cheneygc machine. 15:49:10 hehe 15:49:16 nyef: gcc farm 15:49:29 yes, join the thundering herd 15:49:54 it's where nikodemus tests mips builds 15:50:05 *nyef* initially read that and thought "shouldn't that be 'thundering hurd'?" 15:50:09 eh, um 15:50:33 I'm -so- looking forward to getting myself a MIPS system sorted out. 15:50:48 Unfortunately, it looks like it'll run Linux 2.4. 15:50:59 -!- cenan [n=vircuser@88.240.31.213] has quit ["User pushed the X - because it's Xtra, baby"] 15:52:06 ((lambda (x) (funcall x x)) (lambda (x) (funcall x x))) 15:52:13 one mips in the gcc compile farm has linux 2.6.24 15:52:16 for syste 15:52:29 -!- asksol_ [n=ask@pat-tdc.opera.com] has quit [Client Quit] 15:52:54 asksol [n=ask@213.236.208.247] has joined #lisp 15:54:36 fe[nl]ix [n=algidus@89.202.147.18] has joined #lisp 15:54:41 Anyway, I still think it's an unlikely scenario, simply because of how rarely I see asynchronous interrupts used. 15:55:05 tcr: you summoned me ? 15:55:22 -!- asksol [n=ask@213.236.208.247] has quit [Read error: 60 (Operation timed out)] 15:55:26 nyef: sb-sprof? 15:55:51 Hrm. You may have a point there. 15:56:01 Can it trigger a GC, though? 15:56:12 -!- dean_ [i=4b107cb6@gateway/web/freenode/x-284627a82af23b21] has left #lisp 15:56:12 fe[nl]ix: I got an error trying to compile iolib with *print-length* set to 4 15:56:24 fe[nl]ix: the groveller needs some with-standard-io-syntax somewhere 15:56:29 I guess it's optimized to cons little, but in all likelihood it can. 15:56:49 interesting 15:57:29 Hrm. It's doing its work in a without-gcing. 15:58:12 yes, but that doesn't mean anything for the code before and after it 15:58:28 There is no code after it, and the code before it doesn't cons. 15:58:40 when the without-gcing form ends any deferred gc will run 15:58:54 That's what I was looking at next. 15:59:04 tcr: I'll fix it tonight, I'm going home now 15:59:11 -!- fe[nl]ix [n=algidus@89.202.147.18] has quit [Client Quit] 15:59:47 There's also the question of if it conses or not, regardless of the without-gcing. 16:00:06 RECORD does 16:00:15 but it's hard to estimate the probability of this happening in practice in any case 16:00:18 -!- kpreid [n=kpreid@cpe-67-249-58-190.twcny.res.rr.com] has quit [] 16:00:41 my bet it's much easier to make it trip over with a well constructed test case 16:00:49 kpreid [n=kpreid@cpe-67-249-58-190.twcny.res.rr.com] has joined #lisp 16:00:52 Yes, that seems more likely. 16:01:21 mcspiff [n=user@DC7CB.WPA.Dal.Ca] has joined #lisp 16:01:45 hmm, RECORD may not cons in the normal case 16:02:04 Doesn't look like it does, especially given that it's inlined. 16:02:13 only when the sample vector is full. 16:03:08 So, that makes it rarer. 16:03:11 rread [n=rread@192.18.41.196] has joined #lisp 16:03:22 -!- ikki [n=ikki@201.155.75.146] has quit [Read error: 60 (Operation timed out)] 16:03:34 Which brings us back to the "no sufficiently common use-case for it to have been noticed as a problem". 16:03:35 infinitely so 16:03:54 timers, if anyone uses them ... 16:04:05 Can't be infinitely so unless the sample vector never fills up. 16:04:17 i have a file where two functions call each other... how do i make the compiler not warn me that the second is undefined when the first is processed? 16:04:28 clop: Compile the file instead of just loading it. 16:04:45 elliotstern [n=chatzill@cpe-67-240-162-85.rochester.res.rr.com] has joined #lisp 16:04:46 I've never seen that message about it having filled up. 16:05:22 although I sprofiled a bit back in my day. 16:05:37 But I'm not the patient sort. 16:06:15 -!- Phoodus [i=foo@ip68-231-38-131.ph.ph.cox.net] has quit [Read error: 104 (Connection reset by peer)] 16:07:35 -!- sepult [n=user@xdsl-87-78-131-34.netcologne.de] has quit ["ERC Version 5.3 (IRC client for Emacs)"] 16:08:33 -!- srcerer_ is now known as srcerer 16:09:14 I know how to close the smaller of the windows if it's open, but not the larger. 16:09:38 And there's no point in doing so if it's compensated for elsewhere. 16:09:51 Blkt [n=wasdf@dynamic-adsl-94-37-248-35.clienti.tiscali.it] has joined #lisp 16:10:35 tseug [n=tseug3@cpe-70-112-21-137.austin.res.rr.com] has joined #lisp 16:10:36 how can you close the smaller? 16:11:04 stassats` [n=stassats@wikipedia/stassats] has joined #lisp 16:11:07 -!- tseug [n=tseug3@cpe-70-112-21-137.austin.res.rr.com] has quit [Client Quit] 16:11:22 There are two cases. First, the instruction pointer indicates an instruction that writes to reg_CODE. 16:11:51 Second, we're in a branch-delay slot and the -next- instruction pointer indicates an instruction that writes to reg_CODE. 16:13:12 If reg_CODE is the wrong thing to use here, then reg_LIP is the right thing (due to the calling convention). 16:13:25 schoppenhauer [n=senjak@unaffiliated/schoppenhauer] has joined #lisp 16:13:38 Where's the code that emits the note of a VOP ? 16:13:42 Closing the larger window will likely involve modifying the code sequences for recieving unknown-values. 16:14:08 tcr: "the note of a VOP"? 16:14:36 signed word to integer coercion etc 16:14:40 Ah. 16:15:09 the former conditions are necessary, but are they sufficient? 16:16:08 Hmm? 16:16:56 -!- ignas [n=ignas@78-60-73-85.static.zebra.lt] has quit [Read error: 110 (Connection timed out)] 16:17:14 tcr: There's a maybe-emit-coerce-efficiency-note in compiler/represent, that might be it. 16:17:24 yeah just found it 16:17:25 thank you 16:17:38 -!- kiuma [n=kiuma@proxy.emea.fedex.com] has quit ["Bye bye ppl"] 16:17:55 mrsolo__ [n=mrsolo@adsl-68-126-189-5.dsl.pltn13.pacbell.net] has joined #lisp 16:18:03 Or... Hrm. There's an actual :NOTE parameter to define-vop? 16:18:58 gigamonkey [n=user@adsl-99-17-205-4.dsl.pltn13.sbcglobal.net] has joined #lisp 16:18:59 nyef: may reg_CODE be written by code that does not follow this convention? That is, is there a different kind of race with entering/exiting foreign land? 16:19:11 mega1: That I haven't examined yet. 16:20:23 <_3b> mega1: if you mean the "Profiler sample vector full ... doubling the size" message from sb-sprof, i get that regularly 16:20:50 strange, what are you doing differently from me? 16:21:18 <_3b> posisbly testing large apps/libs instead of specific functions? 16:22:00 <_3b> for example profiling compilation of some of my pathological code that sbcl dislikes :) 16:22:01 a uint-32 number times a constant 18782, how to do this so that no bignum or uint-64 is not involved 16:22:22 in sbcl if that matters 16:22:27 <_3b> i think i've also seen it profiling a running cl-opengl app 16:23:20 leo2007: do you know the result fits inside an uint32? 16:23:43 the results can go beyond uint32 16:24:06 then uint64 or a bignum _has_ to be involved, pretty much 16:24:21 -!- envi^home [n=envi@220.121.234.156] has quit ["Leaving"] 16:24:31 but I only need quotient and remainder of (truncate result 2^32) 16:25:08 Nicodemus: unfortunately if that's the case, it slows down drastically 16:25:56 -!- hjpark [n=user@221.138.197.236] has quit [Read error: 110 (Connection timed out)] 16:26:05 tcr: Also note-rejected-templates in compiler/ltn. 16:26:37 dv_ [n=dv@85-127-129-32.dynamic.xdsl-line.inode.at] has joined #lisp 16:27:18 nikodemus: the function I am trying to port is only 10-lines of C http://school.anhb.uwa.edu.au/personalpages/kwessen/shared/Marsaglia03.html 16:27:20 -!- Odin-MAC [n=sbkhh@adsl-2-92.du.snerpa.is] has quit [] 16:27:27 the key part is only 4 lines 16:28:02 unfortunately, the version have ported so far is like > 20 times slower 16:28:50 leo2007: the last one from that page? 16:28:58 yes 16:29:01 sorry 16:29:32 -!- shmho [n=user@58.142.15.103] has quit [Remote closed the connection] 16:30:24 -!- nenorbot [n=nix@bzq-219-145-16.static.bezeqint.net] has quit [] 16:32:14 -!- hyperboreean [n=none@unaffiliated/hyperboreean] has quit ["leaving"] 16:33:23 -!- milanj [n=milan@79.101.231.108] has quit [Read error: 60 (Operation timed out)] 16:34:31 ikki [n=ikki@201.155.75.146] has joined #lisp 16:34:53 rullie [n=rullie@129.97.143.15] has joined #lisp 16:35:24 -!- rullie [n=rullie@129.97.143.15] has quit [Client Quit] 16:35:43 -!- mrSpec is now known as spec[away] 16:36:48 -!- Athas [n=athas@0x50a157d6.alb2nxx15.dynamic.dsl.tele.dk] has quit [Remote closed the connection] 16:37:31 what is the right way to compile a lisp file and have it loaded? when i just use compile-file it doesn't seem to load it 16:37:50 (load (compile-file file)) 16:38:30 -!- seejay [n=seejay@unaffiliated/seejay] has quit ["Leaving"] 16:38:38 Easy, ain't it? 16:38:39 milanj [n=milan@79.101.231.108] has joined #lisp 16:39:03 hrmn, it seems to complain about my import statement, saying that my import is being done twice or something 16:39:10 maybe ic an't put imports into a file to compile 16:39:32 Well, it's better to use defpackage anyways. 16:39:57 Indeed, import, use-package, shadow, etc, are better reserved to interactive use. 16:40:30 ah, i wrote this code some time ago, i think i hit a lisp where defpackage wasn't working right... probably stupid gcl or something... i shoudl probably use defpackage instead 16:40:40 and things like setfs, are those re-executed? 16:41:28 Normally not. But they could be put in eval-when forms... 16:41:51 ejs [n=eugen@252-237-124-91.pool.ukrtel.net] has joined #lisp 16:42:00 hrmn. is there a command to just compile all functions that have been loaded? 16:42:05 -!- frozsyn [n=FrozSyn@wafer.futurs.inria.fr] has quit ["Leaving"] 16:42:17 -!- oudeis [n=oudeis@bzq-79-182-115-160.red.bezeqint.net] has quit ["This computer has gone to sleep"] 16:42:20 (load (compile-file file)) ; is the best. 16:42:43 Otherwise you cannot distinguish the function that "have just been loaded" from the others. The best you can do is: 16:43:00 (do-symbol (s *package*) (when (fboundp s) (compile s))) 16:43:05 ok 16:43:17 things like (declaim (optimize ...)) are those file scope then? i want to just do that globally 16:43:36 Then declaim outside of the file. 16:43:46 I have some global declaims in my rc files. 16:44:00 -!- mkfort [i=pqq9TWb3@68-189-164-209.dhcp.spbg.sc.charter.com] has quit [Read error: 54 (Connection reset by peer)] 16:44:10 REPLeffect [n=REPLeffe@69.54.115.254] has joined #lisp 16:44:45 ok cool 16:44:53 how about declaim inlines? 16:45:33 seejay [n=seejay@unaffiliated/seejay] has joined #lisp 16:46:21 -!- mcspiff [n=user@DC7CB.WPA.Dal.Ca] has quit [Read error: 110 (Connection timed out)] 16:46:53 leo2007: the unsigned long long makes that a bitch to write efficiently in lisp on a 32 bit platform 16:47:08 nikodemus: sb-bignum ;) 16:47:20 -!- elliotstern [n=chatzill@cpe-67-240-162-85.rochester.res.rr.com] has quit [Read error: 60 (Operation timed out)] 16:47:20 ha 16:47:23 nikodemus: yes, that's my impression too, 70% of time is wasted there 16:47:39 -!- ejs [n=eugen@252-237-124-91.pool.ukrtel.net] has quit ["This computer has gone to sleep"] 16:48:36 also, sb-regpair 16:48:51 oh, true! 16:49:20 clop: well, clhs declaim says it's implementation dependant whether the declaration persists outside of the file, after compiled. But is should be taken into account once loaded. So you'd better compile-file and load before compiling the next file. 16:49:34 milanj- [n=milan@79.101.231.108] has joined #lisp 16:49:41 (loop (load (compile-file f))) is better than (loop (compile-file f)) (loop (load f)) 16:50:37 -!- amnesiac [n=amnesiac@p3m/member/Amnesiac] has quit ["This computer has gone to sleep"] 16:51:04 elliotstern [n=chatzill@cpe-67-240-162-85.rochester.res.rr.com] has joined #lisp 16:51:23 -!- wedgeV [n=wedgeV@85.31.0.85] has quit [] 16:52:18 Xof: do you remember if there was a reason why sb-regpair was better off as a would-be contrib than merged into the mainline? 16:52:39 -!- phf [n=phf@host.icnfull.com] has quit ["Leaving..."] 16:52:48 a long time ago, I was nervous because I think it only worked if the registers were renumbered, and no-one understood why 16:52:52 (not as a contrib, that is) 16:53:26 i know how to solve that! we'll put pkhuong on the case! 16:54:35 hm, i don't see the renumbering 16:55:02 wait, what? 16:55:26 you mean merged and not-a-contrib? 16:55:36 yes 16:56:04 jsoft [n=Administ@unaffiliated/jsoft] has joined #lisp 16:56:08 does it pessimize something, or... what? 16:56:08 can't remember, except for natural conservatism 16:56:19 Elench [n=jarv@dsl-217-155-101-22.zen.co.uk] has joined #lisp 16:56:22 I honestly can't remember 16:56:50 it's possible it gives the compiler more chances to get representation selection and register allocation wrong 16:57:20 hm 16:57:22 I've never used sb-bignum or sb-regpair, any doc on how to use it 16:57:49 sb-regpair you have to fetch from http://repo.or.cz/w/sbcl/lichteblau.git 16:58:11 it is not part of sbcl? 16:58:15 nope 16:58:54 sb-bignum is sbcl's bignum package -- internal and has no real docs except in comments 16:59:53 to do the that 64-bit thing, which is better to use, sb-regpair or sb-bignum? 17:00:00 if speed is the main concern 17:00:41 <_3b> leo2007: a 64bit lisp would probably be the best choice 17:00:41 i would guess sb-regpair -- i don't think pkhuong was entirely serious in suggesting sb-bignum 17:00:42 if speed is the main concern, use mersenne twister. 17:00:56 or write a VOP :) 17:01:15 leo2007: what is the goal here? 17:01:27 nikodemus: well, sb-bignum:%multiply-and-add would be nearly perfect. 17:01:36 pkhuong: oh, true! 17:01:52 -!- milanj [n=milan@79.101.231.108] has quit [Read error: 110 (Connection timed out)] 17:01:57 i wonder if we could add some cleverness to the compiler about that 17:02:11 -!- dalton [n=user7994@189-19-121-109.dsl.telesp.net.br] has quit ["eject"] 17:02:43 dalton [n=user7994@189-19-121-109.dsl.telesp.net.br] has joined #lisp 17:03:14 Does anyone know of a good RSS reader(or lib) for CL? 17:03:42 -!- Fade [i=fade@outrider.deepsky.com] has quit [Read error: 60 (Operation timed out)] 17:03:45 Fade [i=fade@outrider.deepsky.com] has joined #lisp 17:04:51 Elench: cl-rss? 17:05:27 nikodemus: to get a new rng with different algorithm than mt 17:05:32 Elench: Doesn't seem to have much of a webpage, though. 17:06:03 antoszka: quite possibly, i'd only got as far as searching cliki for "RSS" 17:06:18 Mhm. 17:06:36 If not i'll (try to)write one 17:06:40 Elench: It's available prepackaged for Gentoo and Debian from what I'm seeing. Dunno what OS you're using. 17:06:58 (k)ubuntu at the moment, so the Debian version might work 17:07:05 Thanks for the hint :-) 17:09:06 leo2007: and speed comparable to the sbcl's built-in mt? 17:09:10 amnesiac [n=amnesiac@p3m/member/Amnesiac] has joined #lisp 17:10:41 oudeis [n=oudeis@bzq-84-109-137-43.red.bezeqint.net] has joined #lisp 17:11:15 rdd [n=user@c83-250-157-93.bredband.comhem.se] has joined #lisp 17:13:07 a-s` [n=user@92.81.45.243] has joined #lisp 17:14:13 -!- elliotstern [n=chatzill@cpe-67-240-162-85.rochester.res.rr.com] has quit [Read error: 113 (No route to host)] 17:14:29 mcspiff [n=user@hlfxns0163w-142068210006.pppoe-dynamic.ns.aliant.net] has joined #lisp 17:15:26 Hun [n=hun@pd956be5d.dip0.t-ipconnect.de] has joined #lisp 17:15:27 -!- roidrage [n=roidrage@p54BEB25E.dip0.t-ipconnect.de] has quit [] 17:15:33 -!- oudeis [n=oudeis@bzq-84-109-137-43.red.bezeqint.net] has quit [Client Quit] 17:16:34 milanj [n=milan@79.101.231.108] has joined #lisp 17:18:21 anyone run a desktop completely in mcclim? 17:18:44 merimus: In what sense "a desktop"? 17:19:07 hmm... only using clim apps? 17:19:32 No CLIM window manager, for starters. 17:19:33 -!- milanj- [n=milan@79.101.231.108] has quit [Read error: 60 (Operation timed out)] 17:19:35 merimus: I doubt you really can... 17:19:45 nikodemus: close is fine 17:20:12 nikodemus: i did one using lcg and it seems to come really close 17:20:13 can it run directly off of a framebuffer? 17:21:23 it can, just write a framebuffer back-end 17:22:51 stuart71 [n=soverton@75-145-221-229-Memphis.hfc.comcastbusiness.net] has joined #lisp 17:24:18 Is calling someone from England a "Brit" in any way offensive? 17:24:50 It probably depends on context and intonation. 17:24:58 Probably about as offensive as calling an American "Yank" 17:25:16 On the other hand, britanic and yankee... 17:26:25 So, I'm looking at the cheneygc stuff with respect to call-out / alien-funcall and... It looks like it's trivial to persuade the GC to scavenge a foreign register context as if it were a lisp register context. 17:26:26 offense is in the ear of the listener 17:26:43 *Elench* probably wouldn't be offended by "Brit" 17:26:47 -!- a-s [n=user@92.81.130.69] has quit [Connection timed out] 17:26:54 Elench: are you one? 17:26:54 *merimus* is certainly not offended by yank 17:27:13 gigamonkey: Yup 17:27:28 pom! 17:28:00 The pom thing never made sense to me 17:28:22 I think it is short for pompous 17:28:46 Sbidicuda [n=antani@host11-26-dynamic.59-82-r.retail.telecomitalia.it] has joined #lisp 17:29:03 ejs [n=eugen@252-237-124-91.pool.ukrtel.net] has joined #lisp 17:29:10 Prisoner of Her Majesty 17:29:18 -!- Sbidicuda [n=antani@host11-26-dynamic.59-82-r.retail.telecomitalia.it] has quit [Client Quit] 17:29:45 Is what i've heard from australian people 17:29:52 "The Oxford English Dictionary (OED) strongly supports the theory that pommy originated as a contraction of "pomegranate".[3] The OED also suggests that the reason for this is that pomegranate is extinct Australian rhyming slang for immigrant;" [Wikipedia] 17:30:13 *Elench* sides with the OED 17:30:39 According to Wikipedia the Prisoner of Her Majesty is a false etymology. 17:31:03 Is there a way to find out whether a node came in via inlining? 17:31:15 Well anyway, i don't feel offended at that one either 17:33:12 gigamonkey: heh, webster's simply says '[origin unknown]' 17:33:22 dialtone [n=dialtone@adsl-99-136-101-166.dsl.pltn13.sbcglobal.net] has joined #lisp 17:33:35 -!- ejs [n=eugen@252-237-124-91.pool.ukrtel.net] has quit [Client Quit] 17:33:46 milanj- [n=milan@79.101.231.108] has joined #lisp 17:34:36 Mazingaro [n=tetsuja@host11-26-dynamic.59-82-r.retail.telecomitalia.it] has joined #lisp 17:34:38 but that's a pretty old webster's dictionary (third new international, unabridged) 17:34:46 salex [n=user@216.80.147.206] has joined #lisp 17:36:50 'afternoon 17:37:04 -!- dv_ [n=dv@85-127-129-32.dynamic.xdsl-line.inode.at] has quit ["Verlassend"] 17:37:05 TDT [n=dthole@dhcpw80ff97da.dynamic.uiowa.edu] has joined #lisp 17:38:59 Taking some opinions yesterday, spending my lunch hour creating a macro for define-condition, so I can loop through it and create a condition for each git command, here's what I have: http://paste.lisp.org/display/82289 - I'm getting an error running this, which hasn't really been the most helpful. 17:39:17 Would anyone have an idea, maybe an obvious one of what I'm not seeing? 17:40:05 -!- dalton [n=user7994@189-19-121-109.dsl.telesp.net.br] has quit ["Nettalk6 - www.ntalk.de"] 17:40:29 Are the TeX sources for CLtL2 available on the net? My current PDF rendering is terrible and I'd like to regenerate it. 17:41:05 TDT: macros do not give you eval-at-runtime 17:41:25 within the macro expander cmd is bound to the symbol k 17:41:58 -!- Mazingaro [n=tetsuja@host11-26-dynamic.59-82-r.retail.telecomitalia.it] has quit [Remote closed the connection] 17:42:53 -!- milanj [n=milan@79.101.231.108] has quit [Read error: 110 (Connection timed out)] 17:43:21 Jason2gs [n=Jason2gs@71.238.211.166] has joined #lisp 17:43:27 -!- Jason2gs [n=Jason2gs@71.238.211.166] has left #lisp 17:43:34 TDT: http://paste.lisp.org/display/82289#1 17:43:48 Also, you should have pasted for channel #lisp, not None 17:43:59 lisppaste: url? 17:43:59 To use the lisppaste bot, visit http://paste.lisp.org/new/lisp and enter your paste. 17:45:50 Mazingaro [n=tetsuja@host11-26-dynamic.59-82-r.retail.telecomitalia.it] has joined #lisp 17:46:35 Athas [n=athas@0x50a157d6.alb2nxx15.dynamic.dsl.tele.dk] has joined #lisp 17:46:54 sreeram [n=sreeram@122.174.68.30] has joined #lisp 17:48:06 antoszka: try ftp://ftp.cs.cmu.edu/user/ai/lang/lisp/doc/cltl/ 17:48:27 kpreid: weird, it's not showing that annotation 17:48:41 nm there it is, thanks 17:49:04 salex: thx, all the links on the web I found were broken 17:52:38 Odin- [n=sbkhh@adsl-2-92.du.snerpa.is] has joined #lisp 17:53:32 dysinger [n=tim@71.20.231.3] has joined #lisp 17:56:12 nikodemus: it is extremely slow to clone that sbcl fork, can I just download the regpair package? 17:56:28 kpreid: Thanks, I'm a bit confused though why the original idea didn't work. 17:57:00 leo2007: i don't know 17:57:03 TDT: what would you would you expect (loop for k in '(a b c) do (setf k 10)) to do? 17:59:07 rread_ [n=rread@nat/sun/x-8152bcdf0923881e] has joined #lisp 17:59:44 To return (10 10 10) 17:59:51 at least that's my guess, I didn't run it 17:59:52 antoszka: ftp ftw! 17:59:52 hjpark [n=user@221.138.197.236] has joined #lisp 17:59:58 jfrancis_ [n=jfrancis@66.194.68.209] has joined #lisp 18:00:34 :) 18:02:22 is it possible that define macro that make sequence of list like "(a) (b) (c)" 18:02:53 hjpark: "(progn (a) (b) (c))"? 18:03:26 If macro expanded, progn is exist on that code? 18:03:49 TDT: if that were the case, you could never assign to any variable whose value is a symbol 18:03:58 without progn, (a) (b) (c) 18:04:01 leo2007: You might try starting from another SBCL git tree, adding nikodemus' repository as a remote and then pulling the remote branch you want. 18:04:31 nyef: he want's lichteblau's stuff, not mine 18:04:37 Oh, sorry. 18:04:38 TDT: another perspective: in your nonworking code, the macro is *completely irrelevant*: it would fail to work in exactly the same way if you wrote define-condition inside the loop body 18:04:38 Either way. 18:04:40 slyrus_ [n=slyrus@dsl092-019-253.sfo1.dsl.speakeasy.net] has joined #lisp 18:04:42 It's a viable approach. 18:04:49 (defmacro mymacro (a b c) `((,a) (,b) (,c))) 18:04:55 sure 18:04:57 there's an option to only grab the tree 18:05:06 Xach [n=xach@unnamed.xach.com] has joined #lisp 18:05:11 Hello Xach. 18:05:17 hello nyef 18:06:29 hjpark: it is not possible in general to make a macro that "expands into multiple forms", such that e.g. (list (mymacro)) has more than one element. Usually progn *is* what you want though. 18:06:29 hey xach ! 18:06:30 kpreid: I think the concept of symbols vs variables is still confusing me. 18:06:40 TDT: variables are code; symbols are data 18:06:43 yo ska, you controversy-hound! 18:06:47 heh 18:06:58 i'm thinking out loud and sketching a lecture 18:07:03 hopefully not too boring 18:07:03 TDT: forms, function calls, special operators, etc are code concepts 18:07:09 kpreid: got it, thanks. 18:07:10 TDT: symbols and conses are data concepts 18:07:26 TDT: certain data is interpreted as code 18:07:29 aka [i=aka@guest221.ibr.cs.tu-bs.de] has joined #lisp 18:07:52 (that and i'm professionally obligated to decry abuse of interesting mathematical objects, natch) 18:07:56 how've you been? 18:09:04 kpreid: Ah, hmm - this is definitely something I need to work on more understanding, the concepts that is. Thanks for helping on that - I'll research these concepts more to get a better grasp on it, and will work with macros more to hopefully help this some. 18:09:25 salex: lately, pretty sad about erik naggum 18:09:37 yeah, i just heard that :( 18:10:02 Xach!! 18:10:03 i was thinking of writing something about it, actually 18:10:14 i was very surprised (though i'm not sure why) 18:10:16 it shouldn't be surprising, but this has been an opportunity for everyone whose feelings he hurt to complain, and for those who have never seen his style before to leap to judgement 18:10:24 -!- merimus [n=wroth@nat-49.laurelnetworks.com] has quit [] 18:10:27 -!- rdd [n=user@c83-250-157-93.bredband.comhem.se] has quit [Read error: 104 (Connection reset by peer)] 18:10:34 -!- mcspiff [n=user@hlfxns0163w-142068210006.pppoe-dynamic.ns.aliant.net] has quit [Read error: 60 (Operation timed out)] 18:10:47 i'm glad on balance it's more fond reminiscing than that, though 18:10:50 yo slyrus_ 18:10:53 nikodemus: sb-regair can not be compiled any more 18:11:04 Xach: true. 18:11:21 sb-bignum would work, for a quick hack. 18:11:54 Xach: hi. i know exactly how you feel -- reddit comments seldom irk me ask much as the naggum ones have done 18:12:20 do you know how old he was? 18:12:31 pkhuong: I don't how to start with sb-bignum 18:12:32 nikodemus: 44 18:12:34 nikodemus: 43 or 44 18:12:48 leo2007: with sb-bignum:%multiply-and-add. 18:13:06 oh, that's young 18:13:23 Clearly, therefore, he was good. 18:14:00 yeah, young enough that it was a shock. 18:14:04 quite 18:14:51 -!- rread [n=rread@192.18.41.196] has quit [Read error: 110 (Connection timed out)] 18:15:06 does someone have sbcl pre-1.0.16.10 on x86 or x86-64 handy? 18:15:45 Oldest I have is 1.0.19. 18:16:03 oh, well -- i'll just build a vintage one :) 18:16:04 saikat_ [n=saikat@76.102.86.244] has joined #lisp 18:16:14 1.0.11 here. 18:16:20 Those memories :) 18:16:35 milanj [n=milan@79.101.231.108] has joined #lisp 18:16:46 pkhuong: how to use it in this case http://paste.lisp.org/display/82281#3 18:17:01 oh, cool. can you paste the output of (disassemble (compile nil `(lambda (x &optional y) (quux x y))))? 18:18:07 nyef pasted "for nikodemus" at http://paste.lisp.org/display/82291 18:19:25 thanks 18:19:59 i'm trying to figure out if that inline error trap has been there always, or if it's a regression 18:20:06 leo2007: (sb-bignum:%multiply-and-add a (aref Q i) c) will return the high and low halves of the operations, as double return values. 18:20:34 (as opposed to a jump to the trap in elsewhere segment) 18:20:47 Okay, let me see if I can get you another data poin then. 18:21:47 nyef annotated #82291 "Another data point" at http://paste.lisp.org/display/82291#1 18:22:03 pkhuong: thank you. let me plug it in 18:22:11 cracki [n=cracki@sglty.kawo2.RWTH-Aachen.DE] has joined #lisp 18:22:14 fe[nl]ix [n=algidus@88-149-210-231.dynamic.ngi.it] has joined #lisp 18:22:24 I'm going to go with "it's been there a while". 18:22:48 ah, and cmucl does the same 18:23:42 can't be good for performance is what i'm thinking -- bloat _and_ invalid instruction in the middle of the stream... 18:24:37 -!- milanj- [n=milan@79.101.231.108] has quit [Read error: 110 (Connection timed out)] 18:26:04 So, my current impression is that, on cheneygc platforms, SBCL handling of asynchronous signals, or any signals in alien code, is unsafe at any speed. 18:26:11 kpreid: For concepts like what you were talking about, related to the data vs code concepts, what do you recommend? 18:26:33 (Or, more accurately, establishing a lisp handler for the signals is unsafe.) 18:26:43 I don't have any documents to recommend; I learned these things by long experience 18:27:13 (I don't mean trial and error on lisp; I mean understanding how these things work in general) 18:27:34 the essence is knowing that most things work on *one* level of indirection, not two 18:27:46 and being able to count how many you want 18:28:21 nyef: urk 18:28:40 pkhuong: I'm getting this The variable SB-BIGNUM:%MULTIPLY-AND-ADD is unbound. 18:28:55 leo2007: it's a function, not a variable. 18:29:03 (multiple-value-setq (c x) (sb-bignum:%multiply-and-add a (aref Q i) c)) 18:29:06 salex: Amazing what a desk-check will do, huh? 18:29:14 That is how I use it 18:30:11 ThomasI [n=thomas@unaffiliated/thomasi] has joined #lisp 18:30:41 nikodemus: probably not... Might be a good idea to have a dedicated error segment (instead of sharing elsewhere between allocation trampolines and error sequences) and to insert padding between the regular constant vector and code. 18:31:18 bobf_ [n=bob@host81-154-14-74.range81-154.btcentralplus.com] has joined #lisp 18:31:43 nyef: indeed. 18:31:44 kpreid: k k, I'll just keep going the way I am and hopefully I'll master the concepts sometime. Thanks again for your help before. 18:31:52 i'd also like to teach the code-generator to reuse already emitted traps when possible 18:33:24 So, you know how all of the cheneygc port use LIP to hold the code object during funcall and the LRA during return and then do an (inst compute-code-from-FOO ...) in the XEP and values-reception sequences? 18:33:30 -!- fe[nl]ix [n=algidus@88-149-210-231.dynamic.ngi.it] has quit [Read error: 60 (Operation timed out)] 18:33:38 I'm getting the distinct feeling that ARM doesn't need to use LIP there. 18:33:45 *nikodemus* most certainly doesn't know :) 18:34:48 Reason being that the program counter is actually a GPR, thus can be used as the target of an ALU instruction... or the source of one. 18:34:50 Distressingly few responses re complexes. I certainly don't feel qualified to make a decision on the issues I raised :\ 18:35:01 -!- stuart71 [n=soverton@75-145-221-229-Memphis.hfc.comcastbusiness.net] has left #lisp 18:35:08 pkhuong: you mean on -devel? 18:35:10 i haven't looked 18:35:14 pkhuong: have you pinged rtoy? 18:35:18 i meant to, after you mentioned them 18:35:22 That said, it might need a multiple-instruction sequence, given the limits on immediate constants. 18:35:22 he may also have insights 18:35:34 but i've really been buried at them moment, and will be for a couple weeks 18:35:56 rtoym: ping ;) 18:35:57 (notwithstanding blathering on my blog during coffee break, which didn't need much thought) 18:35:59 But, already having established that interrupts are gc-unsafe anyway, that shouldn't be a problem. 18:36:21 pkhuong: this is re: sse complex patch, yes? 18:36:43 -!- TDT [n=dthole@dhcpw80ff97da.dynamic.uiowa.edu] has quit ["low battery"] 18:36:50 (You know, at least for the lisp side of this, we could apply a pclsring approach.) 18:37:02 salex: well, tangent issues. 18:37:38 right 18:37:39 pkhuong: I still cannot figure out why it complains The variable SB-BIGNUM:%MULTIPLY-AND-ADD is unbound. 18:38:11 which is the best introdutory common lisp book ? 18:38:16 teilzeitstudent [n=teilzeit@p5B17C49B.dip.t-dialin.net] has joined #lisp 18:38:29 from clhs, (multiple-value-setq (c x) (sb-bignum:%multiply-and-add a (aref Q i) c)) is correct use, right? 18:38:58 george__: depends a bit on your background and what you want to learn 18:39:00 leo2007: I'm afraid my mind reading skills aren't very good today. 18:39:28 -!- kidd1 [n=kidd@32.Red-79-150-112.dynamicIP.rima-tde.net] has quit [Read error: 104 (Connection reset by peer)] 18:39:29 fe[nl]ix [n=algidus@88-149-208-148.dynamic.ngi.it] has joined #lisp 18:39:32 leo2007: provide more context, like the form surrounding that one 18:39:45 salex: if i have learned scheme already ... 18:40:19 sorry, http://paste.lisp.org/display/82281#4 18:40:22 leo2007: "The variable is unbound" when you were expecting a function, or vice versa, usually means that an enclosing special form or macro is being misused or is not actually defined (e.g. wrong package, lack of import, not defined yet). 18:40:52 I find it 18:41:01 a stupid typo outside that function 18:41:03 ljosa [n=ljosa@gm3b6-dd5.broad.mit.edu] has joined #lisp 18:42:10 pkhuong: the speed has improve from 0.65 sec to 0.09 for 1 million generations. Thank you so much. 18:42:44 ZabaQ [n=johnc@host86-177-151-212.range86-177.btcentralplus.com] has joined #lisp 18:42:58 -!- benny [n=benny@i577A2A13.versanet.de] has quit [Success] 18:43:54 kidd1 [n=kidd@111.Red-193-152-161.dynamicIP.rima-tde.net] has joined #lisp 18:45:06 -!- gemelen [n=shelta@shpd-92-101-154-182.vologda.ru] has quit ["I wish the toaster to be happy, too."] 18:46:31 kpreid: thank you 18:47:42 -!- bobf [n=bob@unaffiliated/bob-f/x-6028553] has quit [Read error: 110 (Connection timed out)] 18:49:52 hkBst [n=hkBst@gentoo/developer/hkbst] has joined #lisp 18:51:41 nikodemus pasted "reusing traps" at http://paste.lisp.org/display/82292 18:53:19 REM has no diference with respect to MOD as regards execution efficiency in SBCL right? 18:53:36 nikodemus: Wouldn't that confuse the debugger's notion of control flow? And what about traps with use-value restarts, should one of those patches ever get committed? 18:53:38 -!- teilzeitstudent [n=teilzeit@p5B17C49B.dip.t-dialin.net] has quit ["Leaving"] 18:53:57 *jthing* seems to remember something along thse lines 18:54:03 do we have traps with working use-value restarts? 18:54:50 Not yet, but I've produced a couple patches to add them for x86oids. 18:54:56 Hmm "(probe-file "ccl-tools;asdf-install;") => NIL". Something is very wrong. 18:55:13 jthing: REM should be faster, since MOD is REM+some 18:55:33 antgreen [n=green@nat/redhat/x-dd7b125fbf213f58] has joined #lisp 18:55:34 (in the general case, that is) 18:55:57 nikodemus: but the compiler optimizes it, I think 18:56:15 jthing: test, and then you'll know 18:56:16 Jacob_H [n=jacob@92.5.116.107] has joined #lisp 18:56:38 stuart71 [n=soverton@75-145-221-229-Memphis.hfc.comcastbusiness.net] has joined #lisp 18:56:49 nyef: i think it should be possible to make this play nice with use-value 18:57:05 Okay. 18:57:29 less sure about the debugger, since i don't remember what it knows (or thinks it knows) about control flow 18:57:30 And I probably should spend the time required to bring the use-value patches up to HEAD and throw them in my public tree. 18:57:30 -!- jsoft [n=Administ@unaffiliated/jsoft] has quit [Remote closed the connection] 18:57:34 ZabaQ: try (probe-file "ccl:tools;asdf-install;") instead 18:58:10 anyhow, that (logand #xFFFFFFFF z) should be replaced by (rem z #.(expt 2 32)) so the compier can see that it doesn't overflow and has to assume potential bignum. As least thatt's what it looks like to me. 18:58:47 joga_ [i=joga@rikki.fi] has joined #lisp 18:58:51 jthing: The compiler knows about what logand does, though, doesn't it? 18:59:05 stassats`: D'oh. I never really did grok the pathname stuff. 18:59:06 -!- milanj [n=milan@79.101.231.108] has quit [Connection timed out] 18:59:23 pathnames are not made to be easily (at all?) grokked 18:59:37 ZabaQ: neither did i 18:59:47 -!- ski [n=slj@c-e113e055.1149-1-64736c10.cust.bredbandsbolaget.se] has quit [Read error: 104 (Connection reset by peer)] 18:59:53 -!- fe[nl]ix [n=algidus@88-149-208-148.dynamic.ngi.it] has quit [Read error: 60 (Operation timed out)] 19:02:14 milanj [n=milan@79.101.231.108] has joined #lisp 19:02:47 salex: I got that impression, too :-) 19:02:53 Fufie [n=innocent@86.80-203-225.nextgentel.com] has joined #lisp 19:03:01 -!- joga [i=joga@unaffiliated/joga] has quit [Nick collision from services.] 19:03:07 -!- joga_ is now known as joga 19:03:36 bakkdoor [n=bakkdoor@xdslfr200.osnanet.de] has joined #lisp 19:03:42 fundamental [n=fundamen@24-148-122-247.ip.mhcable.com] has joined #lisp 19:04:20 benny [n=benny@i577A062A.versanet.de] has joined #lisp 19:04:51 ski [n=slj@c-e113e055.1149-1-64736c10.cust.bredbandsbolaget.se] has joined #lisp 19:04:55 nyef: I don't think it is that smart.. 19:05:06 jthing: Try it. 19:05:18 will do 19:05:52 *nyef* is fairly confident that it -is- that smart, as he remembers discussions about modular arithmetic a while back, and Xof had some sort of paper on rings and SBCL. 19:06:20 *ZabaQ* vaguely remembers that paper. 19:06:50 (SETQ C (LOGAND 4294967295 (1+ C))) 19:06:50 ; 19:06:50 ; note: doing unsigned word to integer coercion (cost 20), for: 19:06:50 ; the second argument of VALUE-CELL-SET 19:07:19 so it seems to be a problem 19:08:33 -!- amnesiac [n=amnesiac@p3m/member/Amnesiac] has quit ["This computer has gone to sleep"] 19:09:26 -!- zophy [n=sy@host-242-6-111-24-static.midco.net] has quit [Read error: 104 (Connection reset by peer)] 19:09:43 jthing: Again, value cells are not stored unboxed. 19:10:19 a version with pkhuong's help has been posted in http://paste.lisp.org/display/82281#5 19:10:40 The speed has come really close to the default mt rng 19:10:45 Davse_Bamse [n=davse@4306ds4-soeb.0.fullrate.dk] has joined #lisp 19:11:04 Is there a reason why you don't like mt? 19:11:46 nyef: no, I like it, but I want three rngs 19:12:09 A promise rng, an engagement rng and a wedding rng? 19:12:11 mt would be my choice but other rngs are for testing purposes 19:12:18 nyef: ;) 19:12:24 what's your third? 19:12:38 three: lcg, mt, cmwc 19:12:40 (* t2 rr) returns floating point.. perhaps (the (unsigned-byte 32) (* t2 rr)) 19:12:44 4. Chosen by fair die roll, guaranteed to be correct. 19:12:59 jthing: Umm... No. That's how you lie to the compiler. 19:13:03 jthing: that's intended for uniform (0,1) 19:13:41 Odin-MAC [n=sbkhh@adsl-2-92.du.snerpa.is] has joined #lisp 19:13:52 82281#5 still has three compiler notes, so there might be some room for improvement 19:13:53 -!- joast [n=rick@76.178.184.231] has quit [Read error: 110 (Connection timed out)] 19:14:24 4 notes 19:15:17 Well, it didn't take me long to crash ccl. 19:15:31 fe[nl]ix [n=algidus@88-149-209-105.dynamic.ngi.it] has joined #lisp 19:15:37 3 in my sbcl 1.0.27 19:15:43 http://paste.lisp.org/display/82281#6 19:15:53 any idea how to get rid of them? 19:16:08 tcr: done 19:16:44 asksol [n=ask@062016247101.customer.alfanett.no] has joined #lisp 19:17:57 joast [n=rick@76.178.184.231] has joined #lisp 19:18:55 leo: where does monte-carlo::I come from? 19:19:11 i is 4095 19:19:16 which is the best common lisp book , for people migrating from scheme ? 19:19:24 -!- saikat_ [n=saikat@76.102.86.244] has quit [] 19:19:30 Jonathan-Smith: in the outer let's 19:19:46 oh i see, sorry didn't scroll up 19:19:52 george__: practical common lisp and paradigms of AI programming are good books regardless of your background 19:20:25 amnesiac [n=amnesiac@p3m/member/Amnesiac] has joined #lisp 19:21:14 mt : cmwc = 0.778 : 0.901 seconds for 10 m random numbers 19:21:34 bye 19:21:42 -!- ZabaQ [n=johnc@host86-177-151-212.range86-177.btcentralplus.com] has quit [Remote closed the connection] 19:22:47 Krystof [n=csr21@84-51-132-95.christ977.adsl.metronet.co.uk] has joined #lisp 19:23:06 davazp [n=user@56.Red-79-153-148.dynamicIP.rima-tde.net] has joined #lisp 19:23:15 -!- xan [n=xan@cs78225040.pp.htv.fi] has quit ["leaving"] 19:23:21 leo2007: you could try punning C as the first or last element of the arrat 19:24:07 -!- seejay [n=seejay@unaffiliated/seejay] has quit ["Leaving"] 19:24:24 nikodemus: would that be roughly the same as creating a zero array for C? 19:24:33 george__: sorry, i actually thought i'd answered you the first time.! xach gave the recommends i would have anyway 19:24:36 seejay [n=seejay@unaffiliated/seejay] has joined #lisp 19:25:21 -!- hjpark [n=user@221.138.197.236] has quit [Remote closed the connection] 19:25:50 leo2007: probably better than that 19:26:07 since you already need to hold on to Q 19:26:38 -!- aka [i=aka@guest221.ibr.cs.tu-bs.de] has quit [] 19:26:48 nikodemus: good idea 19:29:23 -!- Odin- [n=sbkhh@adsl-2-92.du.snerpa.is] has quit [Connection timed out] 19:29:31 -!- legumbre_ [n=user@r190-135-4-18.dialup.adsl.anteldata.net.uy] has quit [Read error: 110 (Connection timed out)] 19:29:47 oudeis [n=oudeis@89-138-239-170.bb.netvision.net.il] has joined #lisp 19:29:54 Jarvellis [n=jarv@dsl-217-155-101-22.zen.co.uk] has joined #lisp 19:30:03 gigamonk` [n=user@adsl-99-184-205-104.dsl.pltn13.sbcglobal.net] has joined #lisp 19:30:19 -!- Elench [n=jarv@dsl-217-155-101-22.zen.co.uk] has quit [Nick collision from services.] 19:30:23 -!- Jarvellis is now known as Elench 19:31:57 illuminati1113 [n=user@pool-71-114-64-62.washdc.dsl-w.verizon.net] has joined #lisp 19:31:58 -!- krumholt [n=krumholt@port-92-193-115-209.dynamic.qsc.de] has quit [Read error: 60 (Operation timed out)] 19:32:15 krumholt [n=krumholt@port-92-193-16-131.dynamic.qsc.de] has joined #lisp 19:32:56 nikodemus: I don't know what to do with m-v-s if c is in Q 19:34:20 leo2007: m-v-b, then setf the variables? 19:34:48 -!- gigamonkey [n=user@adsl-99-17-205-4.dsl.pltn13.sbcglobal.net] has quit [Read error: 104 (Connection reset by peer)] 19:36:59 -!- davazp [n=user@56.Red-79-153-148.dynamicIP.rima-tde.net] has quit [Remote closed the connection] 19:39:02 -!- mrsolo__ [n=mrsolo@adsl-68-126-189-5.dsl.pltn13.pacbell.net] has quit [Read error: 104 (Connection reset by peer)] 19:41:03 replace it with (setf (values ...) ...) 19:41:16 and use (aref Q ...) in place of C 19:41:22 *nikodemus* is off, cheers 19:41:23 -!- nikodemus [n=nikodemu@cs181229041.pp.htv.fi] has quit ["Leaving"] 19:43:22 -!- joast [n=rick@76.178.184.231] has quit [Read error: 110 (Connection timed out)] 19:44:00 jlf` [n=user@alumni.EECS.Berkeley.EDU] has joined #lisp 19:44:43 dstatyvka [i=ejabberd@pepelaz.jabber.od.ua] has joined #lisp 19:45:30 joast [n=rick@76.178.184.231] has joined #lisp 19:46:30 what is a foreign variable and how do i see/change it? 19:48:20 Jasko2 [n=tjasko@75-149-33-105-SFBA.hfc.comcastbusiness.net] has joined #lisp 19:52:19 ... Hunh. It just occurred to me that you can capture a free variable in a closure, but you can't bind one. 19:52:33 *Riastradh* blinks. 19:52:50 clop: that's a variable in non-lisp-managed (mostly C or java) memory 19:53:14 -!- schoppenhauer [n=senjak@unaffiliated/schoppenhauer] has left #lisp 19:53:29 nyef, what would it mean to bind a free variable? If you bound it, it would be a bound variable, not a free variable. 19:53:36 Exactly. 19:53:56 clop: it depends on your implementation how you can access them; it's best to consult the manual, or use cffi if you want to access foreign stuff from C. 19:54:03 Is there anything more to this revelation, nyef? 19:54:31 ok thanks, it was just telling me in the cmucl gc settings that one of them is a foreign variable, i wasn't sure what it meant 19:54:38 No, just the silliness inherent in the thought that Lisp obey some rules with respect to treatment of prisoners^Wvariables. 19:54:50 a-s`` [n=user@92.81.136.29] has joined #lisp 19:54:51 haha 19:54:58 ...ah. 19:55:07 also, they don't even have to tell you their name. 19:55:26 -!- a-s`` [n=user@92.81.136.29] has quit [Remote closed the connection] 19:56:18 legumbre [n=user@r190-135-27-74.dialup.adsl.anteldata.net.uy] has joined #lisp 20:02:03 mrsolo [n=mrsolo@nat/yahoo/x-a0e4be9eb94da9bc] has joined #lisp 20:06:07 -!- mattrepl [n=mattrepl@hq-users.caci.com] has quit [] 20:06:24 -!- Jasko [n=tjasko@c-98-235-105-148.hsd1.pa.comcast.net] has quit [Read error: 110 (Connection timed out)] 20:08:48 -!- a-s` [n=user@92.81.45.243] has quit [Connection timed out] 20:09:27 -!- segv [n=mb@p4FC1E3F4.dip.t-dialin.net] has quit [Read error: 60 (Operation timed out)] 20:10:55 wljyhfgkhgfjh [n=ahhaha@89.232.55.197] has joined #lisp 20:13:16 -!- wljyhfgkhgfjh is now known as makrateli 20:15:37 -!- makrateli [n=ahhaha@89.232.55.197] has quit ["Leaving"] 20:16:18 -!- ljosa [n=ljosa@gm3b6-dd5.broad.mit.edu] has quit [] 20:19:36 -!- Holcxjo [n=holly@ronaldann.demon.co.uk] has quit [Read error: 110 (Connection timed out)] 20:19:54 -!- ikki [n=ikki@201.155.75.146] has quit [Read error: 60 (Operation timed out)] 20:20:14 Holcxjo [n=holly@ronaldann.demon.co.uk] has joined #lisp 20:21:23 segv [n=mb@p4FC1FF7B.dip.t-dialin.net] has joined #lisp 20:21:49 -!- Blkt [n=wasdf@dynamic-adsl-94-37-248-35.clienti.tiscali.it] has quit [Read error: 113 (No route to host)] 20:24:31 -!- REPLeffect [n=REPLeffe@69.54.115.254] has quit [Read error: 110 (Connection timed out)] 20:28:39 -!- Davse_Bamse [n=davse@4306ds4-soeb.0.fullrate.dk] has quit [] 20:29:41 Bleys [n=Bleys@nat/intel/x-512378be1b757387] has joined #lisp 20:30:32 -!- Jonathan-Smith [n=jon@ip24-250-13-137.ri.ri.cox.net] has left #lisp 20:31:09 slash_ [n=Unknown@whgeh0138.cip.uni-regensburg.de] has joined #lisp 20:32:17 -!- bakkdoor [n=bakkdoor@xdslfr200.osnanet.de] has quit [Read error: 113 (No route to host)] 20:35:06 -!- spec[away] is now known as mrSpec 20:36:15 Cuddles [n=asparagu@pool-71-245-7-148.sctnpa.east.verizon.net] has joined #lisp 20:36:37 anyone alive in here? 20:36:45 no 20:37:52 oh, well is there anyone dead in here who knows why emacs with sbcl and slime seems to have decided that the repl only evaluates, prints and loops if i do M-x slime every time, and hope that nothing goes wrong, because errors get skipped over? 20:38:31 -!- ejs0 [n=eugen@252-237-124-91.pool.ukrtel.net] has quit [Read error: 60 (Operation timed out)] 20:38:44 What do you mean? 20:38:57 i mean the repl is unresponsive 20:39:10 -!- Athas [n=athas@0x50a157d6.alb2nxx15.dynamic.dsl.tele.dk] has quit [Remote closed the connection] 20:39:19 which os? 20:39:33 and my .emacs file is exactly the same as it was before i updated sbcl/slime, and my friend who uses a very similar setup had no problem on two different computers...vista. 20:39:49 im so fucking irritated with it that i want to reboot into ubuntu, but i would miss google chrome =( 20:40:03 are there any good, fast, high-memory capacity (i.e., 10+ GB) lisps other than sbcl and ccl? 20:40:17 roidrage [n=roidrage@ip-80-226-17-88.vodafone-net.de] has joined #lisp 20:40:37 Scieneer comes to mind 20:41:24 Cuddles: It becomes unresponsive on error? 20:41:27 then and cmucl too? 20:41:37 -!- oudeis [n=oudeis@89-138-239-170.bb.netvision.net.il] has quit ["This computer has gone to sleep"] 20:41:37 like if you evaluate (error "FOO")? 20:41:40 no, it just ignores errors most of the time, it's unresponsive on start 20:41:47 seems like cmucl is only 32-bit still? i can't get it to run with more than 1632 mb of heap :) 20:42:01 oh, right 20:42:01 Cuddles: How can it ignore errors most of the time if it's unresponsive on start? 20:42:04 cl-user> (print 'hi) does nothing until i do M-x slime 20:42:22 and then after M-x slime, i have to do M-x slime again if i type anything new 20:42:48 and! if i do something like (print asdfhowiehrwoer) just out of irritation, the error prompt gets skipped over 20:43:12 "error prompt"? a sldb buffer should pop up 20:43:39 You M-x slime for the first time, then type (print 'hello), and press RET, what happens? 20:43:42 is HELLO printed? 20:43:46 no 20:43:49 do you get another CL-USER> prompt? 20:43:53 yes, and it does nothing 20:43:53 hefner, thanks for the pointer to scieneer, i didn't know about it 20:44:06 Cuddles: Look into the *inferior-lisp* buffer 20:44:07 Cuddles: what's in *inferior-lisp* buffer? 20:44:12 Cuddles: Does it contain HELLO? 20:44:15 nothing. 20:44:21 any particular reason sb-ext:process-close would kill the lisp? 20:44:22 except all the crap from the initial slime load 20:44:26 at all? no errors? 20:44:29 not a damn thing 20:44:35 cl-user> 20:44:51 there are some warnings about swank interfaces that are unimplemented, if you want those 20:44:55 Can you restart emacs, do what I said above, then paste the content of *slime-events*? 20:45:02 sure 20:45:09 paste at paste.lisp.org 20:45:16 lisppaste: url? 20:45:16 To use the lisppaste bot, visit http://paste.lisp.org/new/lisp and enter your paste. 20:45:45 Paste *slime-events*, *slime-repl sbcl*, and *inferior-lisp* please 20:45:56 Phoodus [i=foo@ip68-231-38-131.ph.ph.cox.net] has joined #lisp 20:46:09 |2dman| [i=80aefb99@gateway/web/freenode/x-6b000ed43222299d] has joined #lisp 20:46:11 laynor [n=laynor@93.107.5.19] has joined #lisp 20:46:34 Cuddles pasted "*slime-events*" at http://paste.lisp.org/display/82300 20:46:39 -!- |2dman| [i=80aefb99@gateway/web/freenode/x-6b000ed43222299d] has quit [Client Quit] 20:46:53 i dont have *slime-repl* 20:46:59 and inferior-lisp is..... 20:47:17 Cuddles annotated #82300 "*inferior-lisp*" at http://paste.lisp.org/display/82300#1 20:47:24 -!- mega1 [n=mega@3e44af28.adsl.enternet.hu] has quit [Read error: 110 (Connection timed out)] 20:47:46 -!- Elench is now known as latcauxe 20:47:48 i dont know why i dont have *slime-repl*, which is another point i meant to bring up 20:47:48 place (slime-setup '(slime-fancy)) into your ~/.emacs 20:47:57 phf [n=phf@host.icnfull.com] has joined #lisp 20:48:04 -!- latcauxe is now known as latcauxehu 20:48:14 I knew I should have said this right away in the beginning 20:48:21 danka 20:48:29 Jacob_H_ [n=jacob@92.5.116.107] has joined #lisp 20:48:33 or danke....* 20:49:06 im surprised that in the several...well...hours, that i spent trying to fix this, before coming in here, that i didnt see that ANYWHERE 20:49:47 saikat_ [n=saikat@c-98-210-13-214.hsd1.ca.comcast.net] has joined #lisp 20:49:52 Helmut did recently something with respect to sbcl on windows 20:49:58 People often come here and must be told to update their ~/.emacs; there was a big incompatible change a few months back 20:50:19 yes, and i kinda figured it was something like that, but i couldnt find anything on it on the interwebs 20:50:29 stassats`: That nothing gets returned is because sbcl uses the NIL communication style 20:51:48 tcr: well, simple-repl seemed to work on Linux 20:52:23 on windows 20:52:27 Cuddles: If you want to use sbcl and slime seriously on windows, you should talk to nyef about his sbcl patch, and how to set it up 20:52:27 Otherwise, I'd recommend using Linux, possibly in a virtual machine 20:52:28 I also heard there's a Linux beta version for Google Chrome 20:52:57 Should be under http://www.lisphacker.com/temp/article-drafts/ somewhere. 20:53:02 tcr: there has been, for a while, but it's in no way usable...im on the mailing list, and as soon as they email me, i'll go back to ubuntu full time 20:53:21 stassats`: Ok perhaps you're right, I missed any such change, but I've not been looking closely 20:53:27 -!- phf [n=phf@host.icnfull.com] has quit ["Leaving..."] 20:53:28 -!- sellout [n=greg@guest-fw.dc4.itasoftware.com] has quit [] 20:54:09 -!- tcr [n=tcr@host145.natpool.mwn.de] has quit ["Leaving."] 20:57:22 oudeis [n=oudeis@89-138-239-170.bb.netvision.net.il] has joined #lisp 20:58:17 -!- joast [n=rick@76.178.184.231] has quit [Read error: 110 (Connection timed out)] 20:59:04 hmm, read-char-no-hang is hanging on a process-pty stream 21:01:07 joast [n=rick@76.178.184.231] has joined #lisp 21:03:00 -!- laynor_ [n=laynor@93.107.81.90] has quit [Read error: 110 (Connection timed out)] 21:06:46 tombom [i=tombom@82.26.83.160] has joined #lisp 21:06:49 t3eblinder [n=Blinder@p5B200B2B.dip0.t-ipconnect.de] has joined #lisp 21:07:23 -!- amnesiac [n=amnesiac@p3m/member/Amnesiac] has quit ["This computer has gone to sleep"] 21:07:52 -!- ryepup [n=ryepup@one.firewall.gnv.acceleration.net] has left #lisp 21:10:37 -!- jthing [n=jthing@212.251.244.165] has left #lisp 21:10:42 jthing [n=jthing@212.251.244.165] has joined #lisp 21:10:47 -!- gigamonk` [n=user@adsl-99-184-205-104.dsl.pltn13.sbcglobal.net] has quit [Read error: 110 (Connection timed out)] 21:11:04 is there a 32 bit %multiply in SBCL? 21:11:51 (in Lispworks it is int32*) 21:12:16 jthing: what about (apropos "MULTIPLY") ??? 21:14:25 amnesiac [n=amnesiac@p3m/member/Amnesiac] has joined #lisp 21:15:04 -!- pjb [n=t@85-169-63-25.rev.numericable.fr] has quit [Remote closed the connection] 21:15:05 matley [n=matley@host84-12-static.25-87-b.business.telecomitalia.it] has joined #lisp 21:16:56 ken-p [n=ken-p@84.92.70.37] has joined #lisp 21:17:04 -!- Bleys [n=Bleys@nat/intel/x-512378be1b757387] has left #lisp 21:17:08 Bleys [n=Bleys@nat/intel/x-512378be1b757387] has joined #lisp 21:17:12 -!- Bleys [n=Bleys@nat/intel/x-512378be1b757387] has left #lisp 21:17:47 jthing: do you mean (logand (* a b) #xffffffff)? 21:22:07 -!- tombom [i=tombom@wikipedia/Tombomp] has quit ["Peace and Protection 4.22.2"] 21:22:34 Krystof: functionally, but that does a ALLOC-UNSIGNED-BIGNUM-IN-EDX 21:22:53 I wanted a imul instruction 21:24:14 bakkdoor [n=bakkdoor@xdslfr200.osnanet.de] has joined #lisp 21:24:15 check carefully 21:24:33 are you sure the bignum allocation is not for a function return value or equivalent? 21:25:19 any better result if using fixnums? (declaring all needed variables as fixnum, and possibly returning (the fixnum ... 21:26:39 well, sure, but that's not a 32-bit multiply any more 21:26:54 jthing is wrong anyway; my expression compiles to a single multiply instruction 21:29:25 antoni [n=user@251.pool85-53-28.dynamic.orange.es] has joined #lisp 21:29:32 -!- roidrage [n=roidrage@ip-80-226-17-88.vodafone-net.de] has quit [] 21:29:39 luis: ping 21:29:44 -!- jewel [n=jewel@dsl-242-129-65.telkomadsl.co.za] has quit [Remote closed the connection] 21:30:35 -!- antgreen [n=green@nat/redhat/x-dd7b125fbf213f58] has left #lisp 21:30:56 -!- stassats` [n=stassats@wikipedia/stassats] has quit [Read error: 60 (Operation timed out)] 21:31:04 or declared as (unsigned-byte 32) or such also? SBCL supports unboxed 32-bit operations too I think 21:31:21 davazp [n=user@56.Red-79-153-148.dynamicIP.rima-tde.net] has joined #lisp 21:31:28 fe[nl]ix: pong 21:32:37 krystof: http://paste.lisp.org/+1RI8 21:32:40 optimization speed/safety parameters will also affect resulting code 21:33:44 -!- salex [n=user@216.80.147.206] has quit [Remote closed the connection] 21:35:33 jthing: MUL EAX, EDX is your inline (unsigned-byte 32) multiply. Everything after that is boxing the return value. 21:36:18 luis: could you test my latest cffi commits, i.e. see if iolib compiles on your machine ? 21:36:36 Bleys_ [n=bleysg@nat/intel/x-075405fa5b2bbb40] has joined #lisp 21:38:11 -!- bombshelter13_ [n=bombshel@toronto-gw.adsl.erx01.mtlcnds.ext.distributel.net] has quit [] 21:39:53 -!- latcauxehu is now known as JHVH 21:41:12 -!- stuart71 [n=soverton@75-145-221-229-Memphis.hfc.comcastbusiness.net] has quit ["Ex-Chat"] 21:42:14 -!- Bleys_ [n=bleysg@nat/intel/x-075405fa5b2bbb40] has quit [Client Quit] 21:42:23 jewel [n=jewel@dsl-242-129-65.telkomadsl.co.za] has joined #lisp 21:42:23 fe[nl]ix: will do 21:42:29 Bleys [n=bleysg@nat/intel/x-a105d4d99862d821] has joined #lisp 21:42:35 -!- Bleys [n=bleysg@nat/intel/x-a105d4d99862d821] has quit [Remote closed the connection] 21:43:26 -!- matley [n=matley@host84-12-static.25-87-b.business.telecomitalia.it] has quit [Read error: 110 (Connection timed out)] 21:44:02 sellout [n=greg@c-24-128-50-176.hsd1.ma.comcast.net] has joined #lisp 21:44:44 -!- puchacz [n=puchacz@87-194-5-99.bethere.co.uk] has quit [Remote closed the connection] 21:45:06 ruediger [n=ruediger@62-47-158-105.adsl.highway.telekom.at] has joined #lisp 21:48:17 -!- Jacob_H_ [n=jacob@92.5.116.107] has quit [Client Quit] 21:48:55 -!- amnesiac [n=amnesiac@p3m/member/Amnesiac] has quit ["This computer has gone to sleep"] 21:49:16 -!- dlowe [n=dlowe@c-24-91-154-83.hsd1.ma.comcast.net] has quit ["Leaving."] 21:51:56 fe[nl]ix: undefined function: CFFI-GROVEL::STRIP-WHITESPACE 21:52:17 mmm, nice 21:52:23 -!- borism [n=boris@195-50-201-218-dsl.krw.estpak.ee] has quit [Client Quit] 21:52:27 it's actually possible in sbcl to remove many packages from the core 21:53:09 I just made some source code that removes all of the packages sb-bignum, sb-disassem, sb-eval, sb-fasl, sb-format, sb-gray, sb-sequence, and sb-walker and fmakunbounds all functions in them, and save-lisp-and-die and the resulting core still work! \o/ 21:53:32 fe[nl]ix: but iolib fails to compile because of the "IOLIB.BASE also exports the following symbols:" warning 21:53:43 -!- ia [n=ia@89.169.189.230] has quit [Client Quit] 21:53:43 -!- JHVH [n=jarv@dsl-217-155-101-22.zen.co.uk] has quit [Read error: 104 (Connection reset by peer)] 21:53:46 oh, and sb-debug too 21:54:05 -!- asksol [n=ask@062016247101.customer.alfanett.no] has quit [Read error: 110 (Connection timed out)] 21:54:28 and my slightly argp-like option processor still works too 21:54:57 luis: that's the usual sbcl defpackage warning. you can ignore it 21:55:21 fe[nl]ix: sure but ASDF fails. Other than that, it works. 21:55:25 or you can add them to the shadow list 21:55:38 http://www.gigamonkeys.com/book/programming-in-the-large-packages-and-symbols.html 21:55:59 oh, wait.. also *exports*. Hm, don't know 21:56:38 hmmm... 21:56:49 sorry, missed the context there I guess 21:57:10 -!- krumholt [n=krumholt@port-92-193-16-131.dynamic.qsc.de] has quit [Remote closed the connection] 21:57:22 luis: hmmm, it doesn't fail here. I get into the debugger, but selecting the ACCEPT restart makes the compilation finish ok 21:57:24 pjb [n=t@85-169-63-25.rev.numericable.fr] has joined #lisp 21:57:39 fe[nl]ix: that's what I meant by "fail". 21:57:42 anyway, it's easily fixable with shadow lists I guess 21:57:50 oh :D 21:57:56 -!- antoni [n=user@251.pool85-53-28.dynamic.orange.es] has quit [Remote closed the connection] 21:58:01 sytse: There's a lot you can do in terms of a minimal core, but minimal cores can't do much. I have an example that doesn't include the compiler or evaluator, but that means no CLOS and no REPL. 21:58:13 nyef: ooh, nice :-) 21:58:33 nyef: well, 'can't do much' means they still can do as much as C programs ;-) 21:58:57 Sure. 21:59:04 is that 'example' of yours public? 21:59:25 It's SBCL-os, so yes, but it's not really worth looking at. 22:00:09 -!- joast [n=rick@76.178.184.231] has quit [Connection timed out] 22:00:56 still would be nice (or at least, fun) to be able to let sbcl write native binaries (that is, without core), minimal or not imho 22:01:14 thehcdreamer [n=thehcdre@93.37.245.15] has joined #lisp 22:01:17 -!- jewel [n=jewel@dsl-242-129-65.telkomadsl.co.za] has quit ["Leaving"] 22:01:26 ikki [n=ikki@201.155.75.146] has joined #lisp 22:01:42 but that would require a *lot* of changes to the architecture-specific plumbing I would guess 22:01:43 -!- jthing [n=jthing@212.251.244.165] has left #lisp 22:01:51 fe[nl]ix: 3% failures. 22:01:56 Got that too, but it's assembler-only, no actual Lisp in the output executable. 22:02:04 yes, of course 22:02:13 joast [n=rick@76.178.184.231] has joined #lisp 22:02:33 was it hard to do that? 22:02:49 It was a bloody pain. 22:03:23 luis: but all regarding the pathname library, right ? 22:04:16 luis pasted "iolib-test failures" at http://paste.lisp.org/display/82306 22:04:45 -!- nha [n=prefect@137-64.105-92.cust.bluewin.ch] has quit ["And with the clicking / Came the ticking / Of the page that was no more / Quoth the server, 404"] 22:04:49 -!- yango [n=yango@unaffiliated/yango] has quit [Read error: 60 (Operation timed out)] 22:04:49 fe[nl]ix: ah, it's because I'm not running echo, righ? 22:05:36 that sounds familiar :) 22:06:13 nyef: heh, thought so 22:06:17 did we have this conversation in opposite roles some years ago? :) 22:06:24 nyef: have you got that up online somewhere? 22:06:28 -!- bakkdoor [n=bakkdoor@xdslfr200.osnanet.de] has quit ["Verlassend"] 22:06:57 I'd like to have a look at that some time (though not now, no time and first have to familiarize myself with sbcl internals) 22:07:01 jewel [n=jewel@dsl-242-129-65.telkomadsl.co.za] has joined #lisp 22:07:22 luis: yes, we did :D 22:08:10 luis: mmm.. maybe this is just because you don't have ipv6 enabled in your kernel? 22:08:44 I see a weird ipv6-to-ipv4 localhost address somewhere in that paste 22:09:20 mm, I guess that one test is supposed to check ipv6 functionality 22:09:21 -!- thehcdreamer [n=thehcdre@93.37.245.15] has quit [] 22:09:27 sytse: no I think the tests failing require the echo service which is probably disabled in Ubuntu. 22:09:38 sytse: http://www.lisphacker.com/temp/sorcery.tgz 22:09:47 -!- oudeis [n=oudeis@89-138-239-170.bb.netvision.net.il] has quit ["This computer has gone to sleep"] 22:09:50 And it's time for my afternoon commute. 22:09:53 -!- nyef [n=nyef@65-125-125-98.dia.static.qwest.net] has quit ["Leaving the office."] 22:11:10 hmm.. amazing, how easy it is with some sbcl internals to write binary headers + assembly code.. 22:11:21 mswin-only I see 22:11:52 sytse: nyef is weird that way :) 22:12:29 me too ;-) 22:12:53 nice though, no sbcl patch required at all 22:13:49 amnesiac [n=amnesiac@p3m/member/Amnesiac] has joined #lisp 22:15:14 he forgot to unlock some packages though 22:15:50 oh wait. No, I forgot to use a 32-bit sbcl ;-) 22:17:53 pstickne [n=pstickne@c-24-21-76-57.hsd1.wa.comcast.net] has joined #lisp 22:17:53 jthing [n=jthing@212.251.244.165] has joined #lisp 22:18:49 -!- milanj [n=milan@79.101.231.108] has quit ["Leaving"] 22:19:08 hsuh [n=user@187.36.13.33] has joined #lisp 22:19:16 -!- Jabberwockey [n=jens@port-9135.pppoe.wtnet.de] has quit [Remote closed the connection] 22:19:48 -!- Edward_ [n=Ed@AAubervilliers-154-1-43-234.w90-3.abo.wanadoo.fr] has quit ["L'oignon fait la farce."] 22:20:24 -!- jewel [n=jewel@dsl-242-129-65.telkomadsl.co.za] has quit [Read error: 60 (Operation timed out)] 22:20:33 -!- Jacob_H [n=jacob@92.5.116.107] has quit ["Leaving"] 22:21:41 oudeis [n=oudeis@bzq-79-182-115-160.red.bezeqint.net] has joined #lisp 22:21:55 mcspiff [n=user@hlfxns0163w-142068210006.pppoe-dynamic.ns.aliant.net] has joined #lisp 22:28:04 -!- Krystof [n=csr21@84-51-132-95.christ977.adsl.metronet.co.uk] has quit [Read error: 113 (No route to host)] 22:28:31 hsuh` [n=user@187.36.13.33] has joined #lisp 22:31:06 -!- hsuh [n=user@187.36.13.33] has quit [Read error: 60 (Operation timed out)] 22:37:22 mcox [n=markcox@203-206-252-106.dyn.iinet.net.au] has joined #lisp 22:40:25 ManateeLazyCat [n=user@119.127.189.107] has joined #lisp 22:40:58 -!- davazp [n=user@56.Red-79-153-148.dynamicIP.rima-tde.net] has quit [Remote closed the connection] 22:41:36 -!- hkBst [n=hkBst@gentoo/developer/hkbst] has quit [Read error: 104 (Connection reset by peer)] 22:44:21 davazp [n=user@56.Red-79-153-148.dynamicIP.rima-tde.net] has joined #lisp 22:46:45 dreish [n=dreish@minus.dreish.org] has joined #lisp 22:47:54 loz- [n=loz@164.176.dsl.syd.iprimus.net.au] has joined #lisp 22:49:19 -!- Mazingaro [n=tetsuja@host11-26-dynamic.59-82-r.retail.telecomitalia.it] has quit [Remote closed the connection] 22:50:01 ia [n=ia@89.169.189.230] has joined #lisp 22:52:18 -!- t3eblinder [n=Blinder@p5B200B2B.dip0.t-ipconnect.de] has quit ["Leaving."] 22:53:08 -!- slash_ [n=Unknown@whgeh0138.cip.uni-regensburg.de] has quit [Client Quit] 22:55:54 shanmu [n=chatzill@client-86-27-63-173.brhm.adsl.virgin.net] has joined #lisp 22:57:12 Hi, I am using CCL on Windows XP and am looking for a subdirectory search enabled asdf definition 23:00:12 -!- attila_lendvai [n=ati@catv-89-132-189-132.catv.broadband.hu] has quit [Read error: 113 (No route to host)] 23:00:47 Hi, I am using CCL on Windows XP and am looking for a subdirectory search enabled asdf definition - the one on cliki asdf page does not work... asdf still gives error component not found 23:00:53 can some help please? 23:02:38 shanmu: unfortunately not yet 23:02:50 attila_lendvai [n=ati@catv-89-132-189-132.catv.broadband.hu] has joined #lisp 23:03:29 -!- lukego [n=lukegorr@58.173.217.78] has left #lisp 23:03:36 p_l: Thanks for the info! perl is my friend then! I used it to generate my init file as of now... wanted an elegant solution... 23:04:32 shanmu: You could try to make a small project out of it - get cl-fad, copy relevant parts into asdf.lisp, modify function which searches for .asd files 23:05:10 p_l: yes. I intend to.. but I am just through the second chapter of practical common lisp :D 23:06:57 -!- ia [n=ia@89.169.189.230] has quit [Client Quit] 23:09:25 -!- ManateeLazyCat [n=user@119.127.189.107] has quit [Read error: 54 (Connection reset by peer)] 23:10:33 ia [n=ia@89.169.189.230] has joined #lisp 23:10:42 -!- pstickne [n=pstickne@c-24-21-76-57.hsd1.wa.comcast.net] has quit [Read error: 110 (Connection timed out)] 23:10:56 -!- [Blay] [n=blaay@BSN-143-124-250.dial-up.dsl.siol.net] has quit ["Leaving"] 23:13:26 nvoorhies [n=nvoorhie@adsl-75-36-204-63.dsl.pltn13.sbcglobal.net] has joined #lisp 23:13:51 -!- nvoorhies [n=nvoorhie@adsl-75-36-204-63.dsl.pltn13.sbcglobal.net] has quit [Read error: 104 (Connection reset by peer)] 23:14:12 -!- Yuuhi [n=user@p5483E711.dip.t-dialin.net] has quit ["ERC Version 5.3 (IRC client for Emacs)"] 23:14:30 nvoorhies [n=nvoorhie@adsl-75-36-204-63.dsl.pltn13.sbcglobal.net] has joined #lisp 23:14:55 I just rolled my own. 23:15:45 -!- Lectus [n=Frederic@189.105.10.45] has quit [Connection reset by peer] 23:21:47 -!- ikki [n=ikki@201.155.75.146] has quit [Read error: 113 (No route to host)] 23:22:06 -!- ruediger [n=ruediger@62-47-158-105.adsl.highway.telekom.at] has quit ["Leaving"] 23:22:32 ikki [n=ikki@201.155.75.146] has joined #lisp 23:22:50 -!- shanmu [n=chatzill@client-86-27-63-173.brhm.adsl.virgin.net] has quit ["ChatZilla 0.9.85 [Firefox 3.5/20090616224221]"] 23:24:01 -!- hsuh` [n=user@187.36.13.33] has quit [Read error: 110 (Connection timed out)] 23:24:26 -!- davazp [n=user@56.Red-79-153-148.dynamicIP.rima-tde.net] has quit [Remote closed the connection] 23:25:42 -!- seejay [n=seejay@unaffiliated/seejay] has quit [Read error: 110 (Connection timed out)] 23:27:37 seejay [n=seejay@unaffiliated/seejay] has joined #lisp 23:31:59 Lectus [n=Frederic@189.105.10.45] has joined #lisp 23:32:07 -!- Hun [n=hun@pd956be5d.dip0.t-ipconnect.de] has quit [Remote closed the connection] 23:36:27 chavo_ [n=user@c-66-41-11-10.hsd1.mn.comcast.net] has joined #lisp 23:37:15 Adlai pasted "ALREF 2.0" at http://paste.lisp.org/display/82310 23:40:24 -!- jlf [n=user@unaffiliated/jlf] has quit [Read error: 104 (Connection reset by peer)] 23:40:43 jlf [n=user@unaffiliated/jlf] has joined #lisp 23:46:03 plage [n=user@118.68.38.106] has joined #lisp 23:46:07 Good morning. 23:46:15 good morning 23:46:25 in case you missed the lisppaste 23:46:35 I've made ALREF 2.0 23:46:42 http://paste.lisp.org/display/82310 23:47:31 -!- ken-p [n=ken-p@84.92.70.37] has quit [Connection timed out] 23:49:47 -!- fiveop [n=fiveop@pD9E6BE97.dip.t-dialin.net] has quit ["humhum"] 23:51:22 -!- sellout [n=greg@c-24-128-50-176.hsd1.ma.comcast.net] has quit [] 23:51:46 -!- blackened` [n=blackene@ip-89-102-28-224.karneval.cz] has quit [] 23:52:08 blackened` [n=blackene@ip-89-102-28-224.karneval.cz] has joined #lisp 23:52:39 p_l: do you mean something like this? http://common-lisp.net/~loliveira/darcs/asdf-source-search/ 23:52:57 Adlai: congratulations, I guess. 23:53:23 elliotstern [n=chatzill@cpe-67-240-162-85.rochester.res.rr.com] has joined #lisp 23:53:54 plage, thanks. I think it's a really useful utility, and I haven't seen anything like it yet. 23:56:47 -!- jfrancis_ [n=jfrancis@66.194.68.209] has quit [] 23:56:56 -!- sreeram [n=sreeram@122.174.68.30] has quit [] 23:57:07 what is the best CL introductory book ? 23:57:16 is pratical common lisp good ? 23:57:22 yes 23:57:23 minion: tell george__ about PCL 23:57:23 george__: direct your attention towards PCL: pcl-book: "Practical Common Lisp", an introduction to Common Lisp by Peter Seibel, available at http://www.gigamonkeys.com/book/ and in dead-tree form from Apress (as of 11 April 2005). 23:57:39 minion: tell george__ about gentle 23:57:40 george__: please see gentle: "Common Lisp: A Gentle Introduction to Symbolic Computation" is a smoother introduction to lisp programming. http://www.cs.cmu.edu/~dst/LispBook/ 23:57:56 george__: There ^ is another one that some people like. 23:58:05 minion: tell george__ about successful-lisp 23:58:05 george__: direct your attention towards successful-lisp: Successful Lisp is a book providing an all-around summary of Common Lisp, available online at http://www.psg.com/~dlamkins/sl/ 23:58:13 OK, time to go to work. 23:58:16 -!- plage [n=user@118.68.38.106] has left #lisp 23:58:22 HET2 [i=diman@xover.htu.tuwien.ac.at] has joined #lisp 23:58:23 -!- Soulman [n=kvirc@154.80-202-254.nextgentel.com] has quit [Read error: 110 (Connection timed out)] 23:58:55 -!- nvoorhies [n=nvoorhie@adsl-75-36-204-63.dsl.pltn13.sbcglobal.net] has quit [] 23:59:37 thanks