00:03:45 fvw [n=sdfpme@59.39.243.199] has joined #lisp 00:04:07 -!- Yuuhi [i=benni@p5483F7F4.dip.t-dialin.net] has quit ["ERC Version 5.3 (IRC client for Emacs)"] 00:04:50 -!- amnesiac [n=amnesiac@p3m/member/Amnesiac] has quit ["Leaving"] 00:06:37 -!- bughunter2 [n=bughunte@unaffiliated/bughunter2] has left #lisp 00:09:41 -!- Fufie [n=innocent@86.80-203-225.nextgentel.com] has quit [Read error: 110 (Connection timed out)] 00:10:09 -!- Jabberwock [n=jens@port-13309.pppoe.wtnet.de] has quit [Remote closed the connection] 00:18:28 sellout [n=greg@c-24-128-50-176.hsd1.ma.comcast.net] has joined #lisp 00:19:47 lixujia [n=lixujia@wuxi.asianux.net] has joined #lisp 00:20:07 teilzeitstudent_ [n=teilzeit@p5B17EC3B.dip.t-dialin.net] has joined #lisp 00:20:51 -!- schoppenhauer [n=christop@unaffiliated/schoppenhauer] has quit [Remote closed the connection] 00:22:24 Guest25711 [i=Anthony@c-69-181-117-45.hsd1.ca.comcast.net] has joined #lisp 00:22:27 t a conference and other people are having trouble connecting, please mention it to staff 00:22:42 http://www.codewithus.net << CODING COMMUNITY 00:22:44 -!- Guest25711 [i=Anthony@c-69-181-117-45.hsd1.ca.comcast.net] has left #lisp 00:23:37 O.o 00:23:40 -!- fvw [n=sdfpme@59.39.243.199] has quit [Remote closed the connection] 00:29:05 -!- rpg [n=rpg@216.243.156.16.real-time.com] has quit [Remote closed the connection] 00:29:22 rpg [n=rpg@216.243.156.16.real-time.com] has joined #lisp 00:29:42 -!- rpg [n=rpg@216.243.156.16.real-time.com] has quit [Remote closed the connection] 00:30:43 rpg [n=rpg@216.243.156.16.real-time.com] has joined #lisp 00:35:40 dnm_ [n=dnm@c-68-49-47-248.hsd1.va.comcast.net] has joined #lisp 00:36:23 -!- teilzeitstudent [n=teilzeit@p5B17EB93.dip.t-dialin.net] has quit [Read error: 101 (Network is unreachable)] 00:36:47 oudeis [n=oudeis@bzq-79-176-53-236.red.bezeqint.net] has joined #lisp 00:38:26 wgl [n=wgl@216.145.227.9] has joined #lisp 00:40:31 -!- asksol [n=ask@247.243.251.212.customer.cdi.no] has quit [Success] 00:41:47 beach` [n=user@ABordeaux-158-1-125-158.w90-60.abo.wanadoo.fr] has joined #lisp 00:42:24 -!- kmels [n=kmels@190.148.177.22] has quit [Read error: 104 (Connection reset by peer)] 00:43:54 kmels [n=kmels@190.148.177.22] has joined #lisp 00:44:25 QinGW [n=wangqing@203.86.81.2] has joined #lisp 00:45:24 kmels_ [n=kmels@190.148.177.22] has joined #lisp 00:45:52 Anac [n=user7994@187.10.22.185] has joined #lisp 00:46:05 -!- Anac is now known as dalton 00:46:07 -!- kmels [n=kmels@190.148.177.22] has quit [Read error: 104 (Connection reset by peer)] 00:48:26 -!- blackened` [n=blackene@ip-89-102-28-224.karneval.cz] has quit [] 00:50:10 mattrepl [n=mattrepl@c-76-104-2-182.hsd1.va.comcast.net] has joined #lisp 00:54:01 bgs100 [n=ian@unaffiliated/bgs100] has joined #lisp 00:57:41 -!- beach [n=user@ABordeaux-158-1-29-202.w90-55.abo.wanadoo.fr] has quit [Read error: 110 (Connection timed out)] 01:04:15 -!- athos [n=philipp@92.250.250.68] has quit [Remote closed the connection] 01:04:29 -!- yahooooo [n=yahooooo@c-76-104-183-185.hsd1.wa.comcast.net] has quit [Read error: 60 (Operation timed out)] 01:06:16 -!- weirdo [i=foo123@c139-243.icpnet.pl] has quit [] 01:08:03 drafael [n=tapio@ip-118-90-143-243.xdsl.xnet.co.nz] has joined #lisp 01:16:30 yahooooo [n=yahooooo@c-76-104-183-185.hsd1.wa.comcast.net] has joined #lisp 01:16:35 Hello, all 01:17:11 nyef: got an idea re why EQL doesn't get constant-folded in eric marsden's test case? 01:17:13 TDT [n=TDT@126.91.248.216.dyn.southslope.net] has joined #lisp 01:18:19 Pepe__ [n=ppjet@78.114.139.251] has joined #lisp 01:18:34 -!- bdowning [n=bdowning@mnementh.lavos.net] has quit [Read error: 60 (Operation timed out)] 01:18:54 -!- dialtone [n=dialtone@unaffiliated/dialtone] has quit [Read error: 110 (Connection timed out)] 01:20:02 bdowning [n=bdowning@mnementh.lavos.net] has joined #lisp 01:21:33 -!- Adlai [n=user@93-173-254-22.bb.netvision.net.il] has quit [Remote closed the connection] 01:22:06 Adlai [n=user@93-173-254-22.bb.netvision.net.il] has joined #lisp 01:28:59 pjb: amazing, i'm inserting sql now *phew* 01:29:28 -!- Pepe_ [n=ppjet@78.116.2.231] has quit [Read error: 110 (Connection timed out)] 01:29:51 fvw [n=sdfpme@59.39.243.199] has joined #lisp 01:30:52 Demosthenex pasted "Working regexp to sql" at http://paste.lisp.org/display/83672 01:31:03 as a sample. 01:31:16 what i don't get, is the part at the end where i had to use an eval 01:31:57 -!- carlocci [n=nes@93.37.222.75] has quit ["eventually IE will rot and die"] 01:38:40 Demosthenex annotated #83672 "Fixed a compile error" at http://paste.lisp.org/display/83672#1 01:42:17 hey Demosthenex, the problem is that insert-object* is a function. Maybe you want (defmacro insert-object (type args) (insert-object* type args)), no quasi quotes 01:44:03 nyef: nevermind, casts update the type correctly but don't pretend the lvar is a constant. 01:45:01 jchicas [n=jchicas@137smtph3wg.tigo.com.sv] has joined #lisp 01:45:13 -!- daniel [i=daniel@unaffiliated/daniel] has quit [Read error: 110 (Connection timed out)] 01:45:43 -!- kmels_ is now known as kmels 01:48:14 gko [n=gko@211.21.137.140] has joined #lisp 01:48:36 -!- Odin- [n=sbkhh@s121-302.gardur.hi.is] has quit [] 01:48:43 -!- mattrepl [n=mattrepl@c-76-104-2-182.hsd1.va.comcast.net] has quit [Read error: 110 (Connection timed out)] 01:49:06 -!- jchicas [n=jchicas@137smtph3wg.tigo.com.sv] has quit [Client Quit] 01:51:23 -!- ia [n=ia@89.169.189.230] has quit ["...may the Source be with you..."] 01:51:35 -!- cracki [n=cracki@sglty.kawo2.RWTH-Aachen.DE] has quit ["If technology is distinguishable from magic, it is insufficiently advanced."] 01:51:55 -!- dalton is now known as ausente 01:52:17 -!- ASau [n=user@193.138.70.52] has quit [Remote closed the connection] 01:52:29 ASau [n=user@193.138.70.52] has joined #lisp 01:54:08 ilitirit: thats likely, but someone convinced me earlier it'd be easier to wrap a function in a macro... and if i unquote it, it fails. 01:55:18 yes it is often easier to have a macro call a normal function because it is sometimes convenient to work mostly with normal functions 01:57:09 second Demosthenex you should change your strange (append `(make-instance ',type) args )) to `(make-instance ',type ,@args) which is more idiomatic 01:58:32 cracki [n=cracki@sglty.kawo2.RWTH-Aachen.DE] has joined #lisp 01:58:42 to stop it failing maybe don't quote ',type but just put ,type? 02:01:28 -!- ajhager [n=ajhager@c-98-223-251-77.hsd1.in.comcast.net] has quit ["A way a lone a last a loved a long the..."] 02:01:54 hey, thats what i was missing :P 02:06:06 -!- Pepe__ is now known as Pepe_ 02:06:24 ia [n=ia@89.169.189.230] has joined #lisp 02:08:19 mattrepl [n=mattrepl@c-76-104-2-182.hsd1.va.comcast.net] has joined #lisp 02:11:29 -!- ignas [n=ignas@ctv-79-132-160-221.vinita.lt] has quit ["Ex-Chat"] 02:14:13 wakeup_vanuber [n=wakeup@koln-4db43df6.pool.einsundeins.de] has joined #lisp 02:15:44 -!- wakeup [n=wakeup@koln-5d817342.pool.einsundeins.de] has quit [Read error: 60 (Operation timed out)] 02:17:02 -!- dnm_ [n=dnm@c-68-49-47-248.hsd1.va.comcast.net] has quit [Read error: 110 (Connection timed out)] 02:17:41 -!- saikat_ [n=saikat@adsl-71-139-210-181.dsl.snfc21.pacbell.net] has quit [] 02:18:54 dys``` [n=andreas@krlh-5f736ea8.pool.einsundeins.de] has joined #lisp 02:21:32 why did Common Lisp choose an "ugly" convention (compared to, say, scheme) of predicates ending in P, and destructive stuff having N (iirc), instead of the arguably prettier ? and ! of scheme? limitations? 02:22:28 Tordek: reader macros :) 02:22:36 hmm? 02:22:43 probably not the actual reason. 02:22:59 but a good post-hoc one, I say. 02:23:16 how are reader macros a reason? 02:24:28 matimago: I have never heard of anyone keeping a TV loud to be able to have private discussions in Vietnam. 02:26:39 Tordek: Say you have a reader macro like ?var ... as used in Norvig's PAIP. 02:27:11 Suddenly END? is END followed by some var with a zero-length name. 02:28:36 Tordek: my guess is: compatibility with existing code. 02:29:01 Xach: damn you! 02:30:22 -!- mattrepl [n=mattrepl@c-76-104-2-182.hsd1.va.comcast.net] has quit [] 02:30:54 don't forget people using character sets that don't include ! or ? 02:31:57 -!- dys`` [n=andreas@krlh-5f72d918.pool.einsundeins.de] has quit [Connection timed out] 02:34:23 -!- mrsolo [n=mrsolo@adsl-68-126-177-41.dsl.pltn13.pacbell.net] has quit ["This computer has gone to sleep"] 02:35:25 looks like i will make it to boston lisp july after all 02:35:28 foom: or - 02:35:50 It's not actually as insane as you might think, there's a standard ISO 6-bit character set that only includes capital letters, 0-9 ()*+,-./:;$%&'[] and british pound. 02:36:36 and another one that includes ()*+,-./:;<=>'[] and pound instead of those symbols. :) 02:37:01 No backquote or #; that makes for a hard lisping life. 02:38:02 I wouldn't put it past someone to have made a custom character set that replaced [] with `# 02:38:32 simonhaines [n=simon@dsl-210-15-193-181-static.TAS.netspace.net.au] has joined #lisp 02:40:20 here ya go: http://www-formal.stanford.edu/jmc/recursive/node4.html 02:40:37 six 6-bit characters make up a 36-bit word, perfect 02:42:30 foom: Yeah, SIXBIT might have been the cause, would need to get my hands on working maclisp to check it again, though 02:43:58 interesting how SIXBIT had left impression in many languages (like Pascal or Lisp) 02:44:08 and C! 02:44:28 foom: C didn't use SIXBIT because the machine was more 8bit oriented 02:44:51 But the C standard does have a limit on the maximum name length you're allowed to assume will be treated as unique by the linker. 02:45:09 which I believe happens to be 6 02:46:04 And I don't have a quote offhand, but I wouldn't be surprised if early dialects don't require case sensitive linkage 02:46:41 foom: 6 8bit-padded characters fit in 3 16bit words 02:47:25 yeah but 3 is a strange number, who'd ever use that? :) 02:47:42 foom: it's not that strange for word-aligned data 02:48:04 also, it fits in single word on later LispMs from Symbolics :D 02:49:11 as case sensitivity, I think original C was case-sensitive, some dialects might not be, but that was before standards 02:50:02 wlr [n=walt@c-65-96-92-150.hsd1.ma.comcast.net] has joined #lisp 02:50:12 here's a thing: http://h71000.www7.hp.com/commercial/c/docs/6180p001.html 02:50:22 that's even talking about relatively *modern* systems 02:50:28 An identifier without external linkage has at most 32,767 significant characters. An identifier with external linkage has 1023 significant characters on Tru64 UNIX systems and 31 significant characters for OpenVMS platforms. ( Section 2.8 describes linkage in more detail.) Case is not significant in external identifiers on OpenVMS systems. 02:52:48 foom: external identifiers on VMS are governed by VMS ABI, which is shared among all languages 02:52:57 bhyde1 [n=Adium@c-66-30-202-56.hsd1.ma.comcast.net] has joined #lisp 02:53:31 at least that's what I remember from my little forays into VMSland 02:54:21 *p_l* still wonders who was the 'genius' who created di/trigraphs 02:56:23 -!- simonhaines [n=simon@dsl-210-15-193-181-static.TAS.netspace.net.au] has quit ["Leaving"] 02:57:35 bighouse [n=bighouse@modemcable239.238-202-24.mc.videotron.ca] has joined #lisp 02:58:35 here ya go, I don't have the standard itself, but this seems authoritative enough. :) http://www.lysator.liu.se/c/c-faq/c-5.html#5-14 02:58:38 hawkbill [n=pradyus@nat/yahoo/x-24bab7109e3eba1d] has joined #lisp 03:00:46 -!- bhyde [n=Adium@c-66-30-202-56.hsd1.ma.comcast.net] has quit [Read error: 113 (No route to host)] 03:04:38 Xach: ah! you're *that* xach! 03:08:03 Xach: and you play protoss? ;] 03:08:58 -!- nvoorhies [n=nvoorhie@208.80.66.30] has quit [] 03:16:47 yoonkn [n=yoonkn@210.108.170.70] has joined #lisp 03:22:36 jlf` [n=user@netblock-208-127-247-82.dslextreme.com] has joined #lisp 03:33:43 -!- rpg [n=rpg@216.243.156.16.real-time.com] has quit [] 03:36:19 davazp [n=user@56.Red-79-153-148.dynamicIP.rima-tde.net] has joined #lisp 03:39:43 TDT` [n=user@vs1202.rosehosting.com] has joined #lisp 03:43:51 saikat_ [n=saikat@c-98-210-13-214.hsd1.ca.comcast.net] has joined #lisp 03:45:21 -!- fvw [n=sdfpme@59.39.243.199] has quit [Remote closed the connection] 03:45:47 -!- hawkbill [n=pradyus@nat/yahoo/x-24bab7109e3eba1d] has quit ["Leaving."] 03:49:12 kei__ [n=gueorgui@gmt.su] has joined #lisp 03:51:17 acieroid` [n=quentin@ks23738.kimsufi.com] has joined #lisp 03:52:38 hawkbill [n=pradyus@nat/yahoo/x-0a7b96dd5b748134] has joined #lisp 03:52:38 -!- kei [n=gueorgui@gmt.su] has quit [Read error: 104 (Connection reset by peer)] 03:55:03 -!- willb [n=wibenton@h69-129-204-3.mdsnwi.broadband.dynamic.tds.net] has quit [Read error: 110 (Connection timed out)] 03:57:25 willb [n=wibenton@h69-129-204-3.mdsnwi.broadband.dynamic.tds.net] has joined #lisp 03:57:52 Ah Symbolics...such a shame. Will we ever see another such radical break from convention? 03:58:18 Hopefully not. 03:59:19 Why not? 04:01:40 -!- acieroid [n=quentin@ks23738.kimsufi.com] has quit [Read error: 110 (Connection timed out)] 04:03:17 -!- TDT` [n=user@vs1202.rosehosting.com] has quit [Remote closed the connection] 04:03:28 -!- TDT [n=TDT@126.91.248.216.dyn.southslope.net] has quit ["Lost terminal"] 04:04:54 In what manner "break from convention"? 04:06:33 the machines itself weren't so revolutionary, maybe except for hw assisted GC 04:11:58 dnm_ [n=dnm@c-68-49-47-248.hsd1.va.comcast.net] has joined #lisp 04:12:10 -!- pstickne [n=pstickne@c-24-21-76-57.hsd1.wa.comcast.net] has quit [Client Quit] 04:13:49 -!- parolang [n=user@keholmes.oregonrd-wifi-1261.amplex.net] has quit [Remote closed the connection] 04:14:35 -!- davazp [n=user@56.Red-79-153-148.dynamicIP.rima-tde.net] has quit [Remote closed the connection] 04:15:12 TDT [n=user@vs1202.rosehosting.com] has joined #lisp 04:17:17 parolang [n=user@keholmes.oregonrd-wifi-1261.amplex.net] has joined #lisp 04:17:46 -!- bohanlon [n=bohanlon@TUBERIUM.CLUB.CC.CMU.EDU] has quit [Read error: 104 (Connection reset by peer)] 04:17:49 bohanlon [n=bohanlon@TUBERIUM.CLUB.CC.CMU.EDU] has joined #lisp 04:17:57 davazp [n=user@56.Red-79-153-148.dynamicIP.rima-tde.net] has joined #lisp 04:18:09 -!- mikezor [n=mikael@213.112.224.79] has quit [Remote closed the connection] 04:18:20 mikezor [n=mikael@c-4fe070d5.04-404-7570701.cust.bredbandsbolaget.se] has joined #lisp 04:18:37 are there any lisp libraries similar to OpenMP? 04:19:01 it seems like it'd be possible to achieve something similar using macros 04:19:53 openmp... i don't think so, but writing an equivalent using bordeaux-threads shouldn't be very hard 04:20:18 you shouldn't even need macros 04:21:52 or even anything-threads 04:22:06 haha 04:22:17 hefner: just pointing b-t so that it would be portable ;-) 04:22:41 why bother when there's only one and two halves lisps worth using it with? :) 04:23:26 lexa_ [n=lexa_@seonet.ru] has joined #lisp 04:23:44 ? 04:23:54 -!- lexa_ is now known as Guest47226 04:24:54 SBCL is one, CCL is at least one half, and I guess you can divide the other half evenly between the commercial lisps 04:25:03 lol 04:27:57 hmm... SCL is surprisingly cheap, for a commercial lisp 04:28:05 really? 04:28:32 fvw [n=sdfpme@59.39.243.199] has joined #lisp 04:29:00 hefner: ~$330 USD per system, with tax 04:29:41 Ringo48, alex fukunaga made some bindings for open mp i think 04:29:43 separate support contract @~$1600 USD/year 04:30:50 That's 237 Euro. 04:31:11 About as much as the full version of MS-Vista. 04:31:12 thanks 04:31:16 scieneer cl is quite closely based on cmucl -- douglas crocker (the author) maintains a patched set of most open source lisp projects to work on it 04:31:40 ilitirit: CMUCL compatibility is mentioned on their website 04:31:49 Is CMUCL so bad that it doesn't count? 04:32:23 btw, support&maintenance contract covers all system license you bought (system licenses don't have time limit) 04:32:34 pbusser3: CMUCL is a wonderful lisp implementation, but lacking native threads, no, it doesn't count at all for this discussion. 04:32:39 I'm not really concerned about an actual binding, just some way of automatically breaking loops up into parallel sections 04:32:54 hefner: Ah ok. That makes sense. 04:34:00 -!- rread [n=rread@nat/sun/x-109a1d4a55e462c5] has quit [] 04:34:21 instead of using OpenMP's "#pragma parrallel for ...", something that'd let me change loop or dotimes into par-loop or par-dotimes 04:35:01 you can do something a little similar with ccl pand and por http://www.cs.utexas.edu/~ragerdl/parallelism/ 04:35:28 thanks 04:35:46 oh, just write it yourself. not much to it, unless you want to overengineer. 04:35:49 i think ccl is the only lisp with a serious threading implementation, though sbcl might have improved enough to not crash so much now 04:36:00 sbcl threads on linux are very stable. 04:36:48 i last used them about three years ago and while generally they are stable, creating say 1k threads/second caused horrible crashes 04:37:05 ilitirit: I think you might need to talk with your kernel :D 04:37:11 that's a feature, to stop you from doing insane things. 04:37:26 about inferior lisp giving up the ghost? 04:37:47 ilitirit: at that speed, I'd expect linux killing itself in half a minute 04:37:53 mega1 has been fixing many of the sort of issues that caused these shenanigans 04:37:58 linux was fine with it 04:38:12 ilitirit: max-threads on 64bit linux is 32621 04:38:43 yes, to be clear i created the threads and they only lived for a short time so i only had a few thousand at any time 04:39:04 still, three years seems to be a long time for SBCL 04:39:27 on a 32-bit linux, I think you fill your address space somewhere approaching the 1,000 mark, with the default stack size 04:39:57 ok, i'm stuck on how to pass a plist of keyword/value pairs to make-instance for slots (initargs?) 04:40:14 hefner: Only on a really old box. 04:40:24 pbusser3: oh? 04:41:04 pbusser3: I think you might hit memory management limits even on newer box 04:41:13 p_l, yes and mega1 recently fixed many bugs that are relevant(?) around 1.0.26, e.g "bug fix: fix deadlocks related to starting threads" 04:41:19 hefner: Maybe not. I was only thinking about the kernel stack, but you also need a stack in user space. :-( 04:41:22 pbusser3: you can't access 36bit space from userspace, after all 04:41:41 hell, you can't really access it from kernelspace either 04:42:32 What is the stack space of a POIX thread? 1M? For 1000 threads that would be 1G. 04:42:44 pbusser3: system-dependant 04:43:01 s/POIX/POSIX/ 04:43:30 also, nothing stops your stack from growing, also, stack could be separate between threads, right? 04:43:49 at least, linux gives no guarantees that threads address spaces are 1:1 04:43:55 p_l: The stacks between threads must be separated, otherwise you get a big mess. 04:44:38 p_l: And the stack size needed depends on the level of recursion if we're talking about a C program. 04:45:10 pbusser3: but they could be placed in the same address space, just have different pointers (though it wouldn't be sane, IMHO) 04:46:23 p_l: I don't know. 04:46:38 p_l: I doubt POSIX demands a certain implementation though. 04:48:08 so what's this month's Boston Lisp Meeting topic? 04:48:30 -!- bgs100 [n=ian@unaffiliated/bgs100] has quit ["Leaving"] 04:49:15 pbusser3: if I understood linux manpage on clone(2) correctly, you can have two separate processes sharing virtual memory like threads and have threads that don't share virtual memory at all 04:49:50 Demosthenex pasted "Working regexp to SQL" at http://paste.lisp.org/display/83674 04:50:18 right, so that works. *phew* what i didn't get was why i could get the last step to work in eval, but trying to make a macro died 04:51:53 lnostdal [i=lnostdal@149-187-178.oke2-bras5.adsl.tele2.no] has joined #lisp 04:52:35 bgs100 [n=ian@unaffiliated/bgs100] has joined #lisp 04:52:58 -!- existentialmonk [n=carcdr@64-252-32-194.adsl.snet.net] has quit [Remote closed the connection] 04:55:15 -!- Axioplase_ is now known as Axioplase 04:56:33 -!- Axioplase [n=Axioplas@fortigate.kb.ecei.tohoku.ac.jp] has quit ["brb"] 04:56:39 Axioplase [n=Axioplas@fortigate.kb.ecei.tohoku.ac.jp] has joined #lisp 05:03:57 dnm__ [n=dnm@c-68-49-47-248.hsd1.va.comcast.net] has joined #lisp 05:04:50 evening 05:07:13 -!- Quadrescence [n=quad@unaffiliated/quadrescence] has quit [Remote closed the connection] 05:08:32 aja_ [n=aja@S01060018f3ab066e.ed.shawcable.net] has joined #lisp 05:12:08 -!- dnm_ [n=dnm@c-68-49-47-248.hsd1.va.comcast.net] has quit [Connection timed out] 05:14:36 -!- ASau [n=user@193.138.70.52] has quit ["off"] 05:24:35 splittist [n=dmurray@63-55.5-85.cust.bluewin.ch] has joined #lisp 05:24:39 morning 05:27:46 fusss [i=73802a57@gateway/web/freenode/x-9a353d231db74804] has joined #lisp 05:28:35 just got my first pay check as "Lisp and Web Developer" :-) wooo \o/ 05:28:48 -!- beach` is now known as beach 05:28:56 though my employers call it "the lisp" :-D 05:29:09 Good morning. 05:29:19 hey beach :-) 05:29:20 fusss: I love it - The Lisp! 05:29:44 they googled it and came back with much respect 05:30:10 chocolait [n=user@dsl-239-115.melsa.net.id] has joined #lisp 05:30:43 hahaha 05:30:47 Reav__ [n=Reaver@212.88.117.162] has joined #lisp 05:31:16 Jacejkn [n=chatzill@198-242-58-66.gci.net] has joined #lisp 05:31:42 the project is marginally healthcare related, so i show them lispjobs.com with a bunch of healthcare industry people using CL. then showed the Roger Corman's resume (using another man's resume to pad my own, :meta-class lisp-resume ;-) 05:32:05 -!- bgs100 [n=ian@unaffiliated/bgs100] has quit ["Leaving"] 05:33:41 fusss: I need to use that meta-class next time :P 05:34:24 -!- chocolait [n=user@dsl-239-115.melsa.net.id] has left #lisp 05:34:48 strange how my first two web apps are "ad network" and "social network". I should have started with the twitter clone or url shortener. 05:36:16 speaking of which, i have two reference-counting GCs in my stack; hunchentoot sessions and clsql views. inevitable clsql hacking awaits me down the road ;-( 05:41:00 -!- bighouse [n=bighouse@modemcable239.238-202-24.mc.videotron.ca] has quit ["Leaving."] 05:42:48 -!- davazp [n=user@56.Red-79-153-148.dynamicIP.rima-tde.net] has quit ["ERC Version 5.2 (IRC client for Emacs)"] 05:43:04 tcr [n=tcr@host145.natpool.mwn.de] has joined #lisp 05:47:24 -!- Jacejkn [n=chatzill@198-242-58-66.gci.net] has left #lisp 05:48:37 i am not sure if this is idiomatic lisp, but i am using filter functions with &key arguments to strip &allow-other-keys from standard key arguments. 05:48:39 let me explain 05:50:00 make-instance &allows-other-keys, and i am wrapping defclass with a macro that allows other keywords arguments for a slot specification 05:50:42 (defclass foo () ((name :type string :display-info '(:input :text :length 20)))) 05:50:58 sorry, make that defview, not defclass 05:51:28 -!- Taggnostr [n=x@86.50.68.11] has quit [No route to host] 05:51:36 (defmacro defview (class supers slots) `(defclass ,class ,supers ,(filter-slots slots))) 05:53:06 dialtone [n=dialtone@c-69-181-124-148.hsd1.ca.comcast.net] has joined #lisp 05:53:40 then inside fliter-slots i have a labels function that looks like (remove-display-args (&optional &key display-info &allow-other-keys t) (remove display-info slots)) then when it returns, the caller function, filter-slots removes all non standard keywords, such :display-info 05:54:46 -!- fvw [n=sdfpme@59.39.243.199] has quit [Remote closed the connection] 05:54:58 for some reason i think i need some kind of destructuring-bind for values, and not just *structure* 05:59:31 -!- 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.""] 06:02:13 -!- TuxPurple [n=TuxPurpl@unaffiliated/tuxpurple] has quit [Client Quit] 06:02:33 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 110 (Connection timed out)] 06:04:24 -!- fusss [i=73802a57@gateway/web/freenode/x-9a353d231db74804] has quit [Ping timeout: 180 seconds] 06:12:32 lispm [n=joswig@e177153051.adsl.alicedsl.de] has joined #lisp 06:13:41 Taggnostr [n=x@dyn57-11.yok.fi] has joined #lisp 06:15:19 abeaumont [n=abeaumon@85.48.202.13] has joined #lisp 06:18:42 -!- parolang [n=user@keholmes.oregonrd-wifi-1261.amplex.net] has quit [Remote closed the connection] 06:20:04 g'day 06:20:22 schme: g'day 06:21:24 (: 06:29:51 -!- kmels [n=kmels@190.148.177.22] has quit [] 06:32:01 -!- Taggnostr [n=x@dyn57-11.yok.fi] has quit [Read error: 113 (No route to host)] 06:37:11 -!- slyrus [n=slyrus@adsl-68-121-172-169.dsl.pltn13.pacbell.net] has quit [Remote closed the connection] 06:39:54 slyrus [n=slyrus@adsl-68-121-172-169.dsl.pltn13.pacbell.net] has joined #lisp 06:40:56 mrSpec [n=NoOne@88.208.105.1] has joined #lisp 06:41:05 -!- fe[nl]ix [n=algidus@88-149-210-67.dynamic.ngi.it] has quit ["Valete!"] 06:41:10 Adlai` [n=user@c-69-181-142-8.hsd1.ca.comcast.net] has joined #lisp 06:42:38 hello schme 06:43:06 Taggnostr [n=x@dyn57-11.yok.fi] has joined #lisp 06:45:21 -!- tmh [n=thomas@pdpc/supporter/sustaining/tmh] has left #lisp 06:47:04 ASau [n=user@host58-231-msk.microtest.ru] has joined #lisp 06:47:21 -!- tcr [n=tcr@host145.natpool.mwn.de] has quit ["Leaving."] 06:48:50 and hello beach ! 06:49:22 -!- Adlai` [n=user@c-69-181-142-8.hsd1.ca.comcast.net] has quit [Remote closed the connection] 06:52:36 Davse_Bamse [n=davse@4306ds4-soeb.0.fullrate.dk] has joined #lisp 06:52:55 tomoyuki28jp [n=tomoyuki@EM114-51-145-92.pool.e-mobile.ne.jp] has joined #lisp 06:53:03 hello 06:53:10 hello tomoyuki28jp 06:53:16 schme: :) 06:53:46 -!- gko is now known as keca 06:54:05 -!- keca [n=gko@211.21.137.140] has quit ["rcirc on GNU Emacs 22.0.990.1"] 06:55:04 keca [n=Keca@211.21.137.140] has joined #lisp 06:55:17 -!- trebor_home [n=user@dslb-084-059-005-168.pools.arcor-ip.net] has quit [Read error: 110 (Connection timed out)] 06:56:05 The differences between O/R mapper and persistent object database like Elephant are class and slot names are correspondding to table and column names in O/R mapper whereas persistent object database does not, correct? Is there any other differences? 06:57:42 tomoyuki28jp: ORM "maps" data in tables into objects, so there is no real 1:1 relationship, while object database will store objects directly and probably interface languages' object system intimately 06:57:54 daniel [i=daniel@unaffiliated/daniel] has joined #lisp 06:59:32 hello 06:59:33 p_l: I see, thanks for the info. 06:59:43 hello mrSpec 07:00:21 hello mrSpec 07:00:41 papermachine [n=papermac@74.83.204.11] has joined #lisp 07:00:48 -!- papermachine [n=papermac@74.83.204.11] has quit [Remote closed the connection] 07:01:01 merus [n=merus@74.83.204.11] has joined #lisp 07:02:23 legumbre [n=user@r190-135-3-234.dialup.adsl.anteldata.net.uy] has joined #lisp 07:03:11 -!- tomoyuki28jp [n=tomoyuki@EM114-51-145-92.pool.e-mobile.ne.jp] has quit [Remote closed the connection] 07:10:31 -!- sohail [n=Sohail@unaffiliated/sohail] has quit [Read error: 113 (No route to host)] 07:12:27 mrsolo [n=mrsolo@adsl-68-126-177-41.dsl.pltn13.pacbell.net] has joined #lisp 07:13:11 aerique [i=euqirea@xs2.xs4all.nl] has joined #lisp 07:15:50 -!- legumbre_ [n=user@r190-135-75-52.dialup.adsl.anteldata.net.uy] has quit [Read error: 110 (Connection timed out)] 07:24:37 jdz [n=jdz@85.254.193.149] has joined #lisp 07:25:26 -!- merus [n=merus@74.83.204.11] has left #lisp 07:26:28 -!- borism [n=boris@195-50-199-19-dsl.krw.estpak.ee] has quit [Read error: 131 (Connection reset by peer)] 07:29:42 jdz_ [n=jdz@85.254.211.133] has joined #lisp 07:33:05 alinp [n=alinp@86.122.9.2] has joined #lisp 07:33:46 ziggurat [n=quassel@pool-96-226-117-126.dllstx.fios.verizon.net] has joined #lisp 07:34:42 riecon [n=nathan@dsl2-189.gsinet.net] has joined #lisp 07:41:48 -!- ziggurat [n=quassel@pool-96-226-117-126.dllstx.fios.verizon.net] has quit [Remote closed the connection] 07:42:12 ziggurat [n=quassel@pool-96-226-117-126.dllstx.fios.verizon.net] has joined #lisp 07:45:37 lujz [n=lujz@cpe-92-37-0-24.dynamic.amis.net] has joined #lisp 07:47:16 -!- jdz [n=jdz@85.254.193.149] has quit [Read error: 110 (Connection timed out)] 07:47:51 blandest [n=user@softhouse.is.ew.ro] has joined #lisp 07:48:41 bobf_ [n=bob@host86-148-145-225.range86-148.btcentralplus.com] has joined #lisp 07:51:02 -!- ziggurat [n=quassel@pool-96-226-117-126.dllstx.fios.verizon.net] has quit [Remote closed the connection] 07:51:29 nostoi [n=nostoi@80.31.172.237] has joined #lisp 07:53:01 -!- bobf [n=bob@unaffiliated/bob-f/x-6028553] has quit [Read error: 54 (Connection reset by peer)] 07:53:54 -!- nostoi [n=nostoi@80.31.172.237] has quit [Client Quit] 07:59:32 addled [n=adlirc@209.20.68.236] has joined #lisp 07:59:55 -!- addled [n=adlirc@209.20.68.236] has quit [Remote closed the connection] 08:00:09 -!- Gertm [n=user@81.83.0.66] has quit [Read error: 110 (Connection timed out)] 08:00:42 hkBst [n=hkBst@gentoo/developer/hkbst] has joined #lisp 08:01:00 -!- mrsolo [n=mrsolo@adsl-68-126-177-41.dsl.pltn13.pacbell.net] has quit [Read error: 104 (Connection reset by peer)] 08:06:50 mrsolo [n=mrsolo@adsl-68-126-177-41.dsl.pltn13.pacbell.net] has joined #lisp 08:07:08 -!- ace4016 [i=ace4016@cpe-76-168-248-118.socal.res.rr.com] has quit ["night"] 08:08:28 pstickne [n=pstickne@c-24-21-76-57.hsd1.wa.comcast.net] has joined #lisp 08:09:04 -!- Harag [n=phil@wbs-196-2-98-168.wbs.co.za] has left #lisp 08:10:26 manuel_ [n=manuel@91.89.172.94] has joined #lisp 08:15:09 -!- alexbobp [n=alex@ppp-70-253-66-27.dsl.austtx.swbell.net] has quit [Read error: 110 (Connection timed out)] 08:19:00 ejs1 [n=eugen@nat.ironport.com] has joined #lisp 08:19:40 -!- lispm [n=joswig@e177153051.adsl.alicedsl.de] has quit [Read error: 110 (Connection timed out)] 08:23:34 morning 08:23:43 hello bulibuta 08:24:03 hey beach 08:24:53 hmmm, has anybody worked with identi.ca? 08:25:28 dmiles [n=dmiles@c-71-197-210-170.hsd1.wa.comcast.net] has joined #lisp 08:27:01 -!- dmiles_afk [n=dmiles@c-71-197-210-170.hsd1.wa.comcast.net] has quit [Read error: 54 (Connection reset by peer)] 08:29:04 Adlai` [n=user@c-69-181-142-8.hsd1.ca.comcast.net] has joined #lisp 08:29:43 moghar [n=user@unaffiliated/moghar] has joined #lisp 08:30:13 looks like it has multiple APIs, including a HTTP ("REST") one .. so, http://weitz.de/drakma/ 08:31:43 asksol [n=ask@249.241.251.212.customer.cdi.no] has joined #lisp 08:32:19 athos [n=philipp@92.250.250.68] has joined #lisp 08:33:18 -!- 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.""] 08:33:29 spilman [n=spilman@ARennes-552-1-36-239.w92-135.abo.wanadoo.fr] has joined #lisp 08:34:45 jamief [n=jamie@158.223.53.104] has joined #lisp 08:45:12 prg [n=prg@ns.alusht.net] has joined #lisp 08:48:23 fvw [n=sdfpme@59.39.243.199] has joined #lisp 08:49:42 good morning 08:49:50 hello serichsen 08:50:18 -!- drafael [n=tapio@ip-118-90-143-243.xdsl.xnet.co.nz] has quit ["Leaving."] 08:50:57 hi beach :) 08:53:46 -!- riecon [n=nathan@dsl2-189.gsinet.net] has quit [Remote closed the connection] 08:56:30 Yuuhi [i=benni@p5483EB87.dip.t-dialin.net] has joined #lisp 09:04:17 mega1 [n=mega@pool-00839.externet.hu] has joined #lisp 09:19:10 shiny. Allegro 9.0 will have SMP, and there's already some documentation explaining the new macros and behaviour that will change. 09:19:50 commercial lisp leaps boldly into the 90s 09:20:00 I saw that, and wondered why it didn't show up on Planet Lisp. Maybe they didn't post it to the feed I'm reading. 09:20:41 *p_l* thinks that Weblocks definitely could use a bigger example 09:21:53 Xach: I'll show up on Planet Prime Lisp, as befits the parallel worlds of lisp 09:22:07 s/I'll/It'll/ 09:24:56 p_l: Do you think so? 09:25:14 *pbusser3* is creating Debian packages for Weblocks. 09:27:13 pbusser3: definitely. I lost a lot of time playing around instead of writing my demo app, now I'm writing in Ruby, but I'd really like an example that showed more than the few simple examples that are placed in weblocks-dev 09:27:32 or maybe I just don't get something, trying to use weblocks like I'd use a GUI toolkit :D 09:28:33 ruediger [n=ruediger@91-115-20-76.adsl.highway.telekom.at] has joined #lisp 09:28:40 -!- Drakeson` [n=user@69-196-169-53.dsl.teksavvy.com] has quit [Connection timed out] 09:29:56 densem [i=5648d9da@gateway/web/freenode/x-31c1a1ed2d4eec52] has joined #lisp 09:30:13 'hi 09:30:17 http://sw.nostdal.ath.cx/text-input getting more enterpricy :) .. (source code link at the bottom) .. i'm trying to come up with good ways to handle cases where there are multiple connections or channels of dataflow between a model and a view/widget instance .. not sure i'm happy with the "return a list of connections" thing 09:30:23 hello densem 09:30:38 hi beach 09:30:56 ready to welcome a noob ? :p 09:31:24 densem: we always are. 09:31:28 densem: that's entirely up to you 09:31:33 lnostdal: 404 09:32:07 stassats, oh .. hm .. i changed ip recently i think 09:33:03 -!- densem [i=5648d9da@gateway/web/freenode/x-31c1a1ed2d4eec52] has quit [Client Quit] 09:33:17 densem [i=5648d9da@gateway/web/freenode/x-325eacb5f2325226] has joined #lisp 09:33:24 densem: What brings you to #lisp? 09:33:38 lnostdal: I managed to end on some voip thingy that asked for username/password but still gave access if you clicked "cancel" enough times... 09:33:53 there .. not sure how fast that'll propagate, stassats .. dyndns does tend to be quick though 09:33:59 hi back, sorry .. the ajax irc client crashed 09:34:07 lnostdal: http://sw.nostdal.org/text-input works, though 09:34:18 p_l, yeah, sorry .. that's someone else's IP entirely .. i run dns on localhost so i did not notice :P 09:34:34 ahok, stassats 09:34:36 i wanted to do some xml filtering as a lisp first try 09:34:54 i used ip from your irc connection 09:35:12 fetching some webpages, pruning the dom .. some sort of html summary 09:35:28 lnostdal: hahaha. BTW, do you really use extended charset for your code or is it some nice display? :D 09:36:03 p_l, i do .. heh :P .. e.g., http://paste.lisp.org/display/82845#6 09:36:51  unicode even has a snowman-character .. how cool is that 09:37:11 i smell APL 09:37:24 hehe, yeah stassats 09:37:43 Odin- [n=sbkhh@s121-302.gardur.hi.is] has joined #lisp 09:38:26 lnostdal: how do you enter lambda? some special mapping? 09:38:45 p_l, shift-Caps Lock 09:38:55 awesome 09:39:23 hmm... I'll have to add mappings 09:39:39 -!- kpreid [n=kpreid@cpe-67-249-58-190.twcny.res.rr.com] has quit [] 09:39:46 (and maybe make myself a new set of keylabels) 09:39:54 M-x set-input-method RET greek RET ! 09:40:29 i got ( at my left pinky .. and ) at my right pinky .. so i can go; ())()()()(())() .. really easy to reach :) 09:41:01 already doing the dvorak thing, so why not .. might as well go all out nuts 09:41:18 *p_l* suddenly gets an urge to release into wild world code depending on kanjis as symbols 09:41:25 haha 09:41:47 lnostdal: doesn't that hurt your pinkies? the ctrl button hurts my pinkies enough as it is 09:42:19 maybe i should map the j and f keys to ) and ( instead 09:42:48 or having two tabs 09:42:48 lnostdal: I am a dvorak user too ;-) 09:42:54 gonzojive: well, control/meta/super/hyper should be under your thumbs, but modern keyboards don't follow that... 09:43:46 gonzojive, it's fine actually .. from a "starting position" i move my left one about 1cm to the left to reach ( .. and the right one 1cm to the right to reach ) .. beats the standard norwegian layout for sure 09:44:03 fe[nl]ix [n=algidus@89.202.147.22] has joined #lisp 09:45:20 hello 09:45:24 hi 09:45:26 hey 09:45:37 hello fe[nl]ix 09:45:42 felix the dutch fenix ? 09:45:47 lnostdal: does sw gracefully degrades in case JS isn't there? :-) 09:45:49 hi beach 09:45:55 p_l, nope 09:45:58 densem: nope 09:46:05 >< famn 09:46:08 d* 09:46:57 fe[nl]ix: How are exams going? 09:47:15 not bad 09:47:25 i wonder how the java guys do mvc .. i can't seem to grok the high-level overviews of some of the java frameworks 09:48:26 java is like programming with two by fours. it always looks the same: classes, classes, classes and classes 09:49:13 heh, yeah -- "Execution in the Kingdom of Nouns" 09:51:41 i don't know if you can achieve anything but conventions with frameworks in java 09:51:48 you end up writing so much code anyway 09:53:43 Athas [n=athas@0x50a157d6.alb2nxx15.dynamic.dsl.tele.dk] has joined #lisp 09:54:26 -!- Odin- [n=sbkhh@s121-302.gardur.hi.is] has quit [] 09:54:30 serichse1 [n=harleqin@xdsl-81-173-189-89.netcologne.de] has joined #lisp 09:55:57 silenius [n=jl@yian-ho03.nir.cronon.net] has joined #lisp 09:56:17 -!- jdz_ is now known as jdz 09:56:45 sometimes I wonder whether my most enduring (and most cited) work in the Lisp community was documenting and providing a place to download for SPLIT-SEQUENCE 09:56:58 if only comp.lang.lisp were indexed for citations 09:58:05 HET2 [i=diman@128.131.95.22] has joined #lisp 09:59:44 -!- HET2 [i=diman@128.131.95.22] has quit [Client Quit] 10:06:37 -!- QinGW [n=wangqing@203.86.81.2] has quit [Read error: 104 (Connection reset by peer)] 10:06:39 -!- serichsen [n=harleqin@xdsl-81-173-152-231.netcologne.de] has quit [Read error: 110 (Connection timed out)] 10:07:51 -!- hyperboreean [n=none@unaffiliated/hyperboreean] has quit ["leaving"] 10:08:56 -!- manuel_ [n=manuel@91.89.172.94] has quit [] 10:09:52 Xof: most enduring, perhaps. But you're not so old that it need be your most cited... 10:12:28 fvw_ [n=sdfpme@59.39.243.199] has joined #lisp 10:12:46 schoppenhauer [n=christop@unaffiliated/schoppenhauer] has joined #lisp 10:13:34 stassats` [n=stassats@wikipedia/stassats] has joined #lisp 10:17:01 -!- stassats [n=stassats@wikipedia/stassats] has quit [Read error: 104 (Connection reset by peer)] 10:17:38 -!- dmiles [n=dmiles@c-71-197-210-170.hsd1.wa.comcast.net] has quit [Read error: 54 (Connection reset by peer)] 10:21:07 -!- lixujia [n=lixujia@wuxi.asianux.net] has quit [""] 10:22:28 jewel [n=jewel@dsl-247-203-169.telkomadsl.co.za] has joined #lisp 10:23:01 -!- Axioplase is now known as Axioplase_ 10:23:45 hyperboreean [n=none@unaffiliated/hyperboreean] has joined #lisp 10:25:05 "I argue that nouns are good for sorting objects into categories and associating responses with the categories. Verbs are better for describing processes and relationships. People using verb based languages are very likely to be a population that assumes juxtapositional thinking is available to them, ..." 10:25:10 http://the-programmers-stone.com/about/neuroscience-questions/ 10:25:40 Start reading the whole text from http://the-programmers-stone.com/about/ 10:27:19 -!- hkBst [n=hkBst@gentoo/developer/hkbst] has quit [Read error: 104 (Connection reset by peer)] 10:28:24 hmmm.... but we *do* use separate stores for "actions" and "memories" 10:29:02 oO 10:30:27 -!- fvw [n=sdfpme@59.39.243.199] has quit [Read error: 110 (Connection timed out)] 10:38:26 ThomasI [n=thomas@unaffiliated/thomasi] has joined #lisp 10:39:41 interesting read 10:42:00 holly_ [n=holly@ronaldann.demon.co.uk] has joined #lisp 10:48:16 nvoorhies [n=nvoorhie@adsl-75-36-204-63.dsl.pltn13.sbcglobal.net] has joined #lisp 10:50:01 -!- keca [n=Keca@211.21.137.140] has quit [Read error: 60 (Operation timed out)] 10:54:15 -!- athos [n=philipp@92.250.250.68] has quit ["leaving"] 10:55:47 wbraun [n=chatzill@vpn0577.extern.uni-tuebingen.de] has joined #lisp 10:57:38 -!- hawkbill [n=pradyus@nat/yahoo/x-0a7b96dd5b748134] has quit ["Leaving."] 10:58:29 -!- dys``` is now known as dys 10:58:48 -!- Holcxjo [n=holly@ronaldann.demon.co.uk] has quit [Read error: 110 (Connection timed out)] 10:58:58 -!- saikat_ [n=saikat@c-98-210-13-214.hsd1.ca.comcast.net] has quit [] 11:01:38 -!- ruediger [n=ruediger@91-115-20-76.adsl.highway.telekom.at] has quit ["Leaving"] 11:05:38 -!- schoppenhauer [n=christop@unaffiliated/schoppenhauer] has quit [Remote closed the connection] 11:07:05 -!- kiuma [n=kiuma@85-18-55-37.ip.fastwebnet.it] has quit ["Bye bye ppl"] 11:07:36 Hello, I have a problem with asdf and/or the package system. I have made a package which is a pattern matching tool. I load it with the asdf tool. When I use the exported functions of the package, sometimes it doesn't work with examples, but these examples work if i am in the pattern-matching package. 11:08:00 manuel_ [n=manuel@213.144.1.98] has joined #lisp 11:08:48 -!- ThomasI [n=thomas@unaffiliated/thomasi] has quit ["Bye Bye!"] 11:09:51 -!- fvw_ is now known as fvw 11:12:02 "doesn't work" is a bit vague :) 11:12:02 -!- oudeis [n=oudeis@bzq-79-176-53-236.red.bezeqint.net] has quit ["This computer has gone to sleep"] 11:13:24 oudeis [n=oudeis@bzq-79-176-53-236.red.bezeqint.net] has joined #lisp 11:13:26 p_l: Aren't there any freely available web applications which use Weblocks? 11:13:41 it returns a wrong value. Instead using a valid pattern and return a variable binding, it returns nil. In the patternmatch-package the same example returns valid bindings.. 11:14:16 that is probably because you are misusing the software 11:14:38 probably you think that ? or _ is a wildcard, whereas instead it's pmp:? or pmp:_ or something 11:14:42 borism [n=boris@195-50-206-70-dsl.krw.estpak.ee] has joined #lisp 11:15:22 Xof's remote debugging at it's best 11:15:29 s/it's/its 11:15:52 -!- oudeis [n=oudeis@bzq-79-176-53-236.red.bezeqint.net] has quit [Client Quit] 11:16:03 -!- fvw is now known as fvw_ 11:16:30 oh, hey, I'm even debugging his own code, not even anything public 11:16:43 exactly 11:16:43 I wonder if I'm right 11:18:15 -!- fvw_ is now known as fvw 11:18:21 mmhh i don't think this is the problem... a variable is a ?x (? at the first char in the symbol name), with normal variables it works fine. I have pattern variables of the form (?* ?x) there it doesn't work 11:18:29 -!- TR2N [n=email@89.180.155.255] has left #lisp 11:19:03 then maybe you are doing something like (intern (frob arg)) and comparing for equality with EQ 11:20:05 (without an explicit package argument for the interning) 11:20:07 trying with multiple views of the same shared model; http://sw.nostdal.ath.cx/text-input .. psychedelic x) 11:20:35 -!- fe[nl]ix [n=algidus@89.202.147.22] has quit ["Valete!"] 11:20:56 ( http://sw.nostdal.org/text-input if the .ath.cx one doesn't work ... :/ .. curse dns) 11:21:05 Xof: I don't get what you mean ;) 11:21:07 wbraun: in any case, tracing appropriate functions (functions which should do one thing but do do another) usually helps 11:21:20 ManateeLazyCat [n=user@119.141.230.249] has joined #lisp 11:21:29 wbraun: you are describing a run-time dependency on *package*; interning a symbol at runtime without an explicit package is one way to get that run-time dependency 11:21:46 matimago: thanks for the link. it's very interesting 11:21:53 pbusser3: there isn't anything more complicated, i.e. no "completed" apps with available source 11:22:03 ecraven [n=nex@140.78.42.103] has joined #lisp 11:22:15 pbusser3: at least none that google uncovered so far (I know there are deployed apps, but I'm looking for source) 11:25:33 -!- huangjs [n=user@watchdog.msi.co.jp] has quit [Remote closed the connection] 11:26:28 -!- BrianRice [n=water@c-71-59-210-115.hsd1.or.comcast.net] has quit [Read error: 104 (Connection reset by peer)] 11:26:43 BrianRice [n=water@c-71-59-210-115.hsd1.or.comcast.net] has joined #lisp 11:28:36 -!- bobf_ [n=bob@host86-148-145-225.range86-148.btcentralplus.com] has quit ["leaving"] 11:29:18 kiuma [n=kiuma@85-18-55-37.ip.fastwebnet.it] has joined #lisp 11:30:23 -!- mrsolo [n=mrsolo@adsl-68-126-177-41.dsl.pltn13.pacbell.net] has quit [Read error: 60 (Operation timed out)] 11:31:03 mstevens [n=mstevens@zazen.etla.org] has joined #lisp 11:34:21 TuxPurple [n=TuxPurpl@unaffiliated/tuxpurple] has joined #lisp 11:34:59 ... being out of the loop sucks. I now have to read through japanese version of Merb book cause I couldn't find my skills back :/ 11:36:20 So, re: noun/verb, notice that lisp is verb based (verb [noun...]), and it may explain beyond the parentheses why it doesn't appeal to most programmers: they're too stressed to be able to use verbs, they only can grunt nouns. 11:37:34 wbraun pasted "pattern" at http://paste.lisp.org/display/83686 11:37:44 Xof: I exported a predicate function which i found out doesn't work right, by tracing it. 11:38:04 Xof: But I don't get the point, why it fails... 11:41:22 -!- Adlai` [n=user@c-69-181-142-8.hsd1.ca.comcast.net] has quit [Read error: 60 (Operation timed out)] 11:41:44 whoops .. ~R does not print insanely huge numbers and floats .. ohwell 11:43:12 ejs2 [n=eugen@nat.ironport.com] has joined #lisp 11:45:59 Edico [n=Edico@unaffiliated/edico] has joined #lisp 11:47:21 -!- spilman [n=spilman@ARennes-552-1-36-239.w92-135.abo.wanadoo.fr] has quit ["Quitte"] 11:52:23 -!- ejs1 [n=eugen@nat.ironport.com] has quit [Read error: 110 (Connection timed out)] 11:52:33 Yes, if 10^66-1 is not insanely huge. 11:53:12 fe[nl]ix [n=algidus@89.202.147.22] has joined #lisp 11:53:47 -!- younder [n=jthing@165.244.251.212.customer.cdi.no] has quit [Read error: 54 (Connection reset by peer)] 11:54:07 With European number names it could go much higher... 11:54:29 dwave [n=ask@pat-tdc.opera.com] has joined #lisp 11:55:37 -!- dwave [n=ask@pat-tdc.opera.com] has quit [SendQ exceeded] 11:57:42 -!- mathrick [n=mathrick@users177.kollegienet.dk] has quit [Read error: 110 (Connection timed out)] 11:57:46 tcr [n=tcr@host145.natpool.mwn.de] has joined #lisp 11:58:29 -!- dnm__ [n=dnm@c-68-49-47-248.hsd1.va.comcast.net] has quit [Client Quit] 11:59:15 -!- asksol [n=ask@249.241.251.212.customer.cdi.no] has quit [Read error: 110 (Connection timed out)] 11:59:40 asksol [n=ask@pat-tdc.opera.com] has joined #lisp 12:00:23 -!- hyperboreean [n=none@unaffiliated/hyperboreean] has quit [Remote closed the connection] 12:05:03 timor [n=martin@w4087.dip.tu-dresden.de] has joined #lisp 12:05:16 -!- timor [n=martin@w4087.dip.tu-dresden.de] has left #lisp 12:05:37 Quadrescence [n=quad@24.118.241.200] has joined #lisp 12:06:52 brb 12:06:55 -!- densem [i=5648d9da@gateway/web/freenode/x-325eacb5f2325226] has left #lisp 12:07:03 qwer [n=fasteez2@218.217.72-86.rev.gaoland.net] has joined #lisp 12:07:12 *qwer* back 12:08:41 -!- qwer is now known as densem 12:08:49 schoppenhauer [n=senjak@unaffiliated/schoppenhauer] has joined #lisp 12:10:49 -!- wbraun [n=chatzill@vpn0577.extern.uni-tuebingen.de] has quit ["ChatZilla 0.9.85 [Firefox 3.5/20090624012820]"] 12:13:04 envi^home [n=envi@220.121.234.156] has joined #lisp 12:14:03 lacedaemon [n=algidus@89.202.147.18] has joined #lisp 12:14:15 -!- fe[nl]ix [n=algidus@89.202.147.22] has quit [Nick collision from services.] 12:14:23 -!- lacedaemon is now known as fe[nl]ix 12:14:42 -!- abeaumont [n=abeaumon@85.48.202.13] has quit [Read error: 110 (Connection timed out)] 12:17:03 -!- Jasko2 [n=tjasko@c-98-235-105-148.hsd1.pa.comcast.net] has quit ["Leaving"] 12:17:45 -!- sellout [n=greg@c-24-128-50-176.hsd1.ma.comcast.net] has quit [] 12:18:42 Jasko [n=tjasko@c-98-235-105-148.hsd1.pa.comcast.net] has joined #lisp 12:23:12 -!- schoppenhauer [n=senjak@unaffiliated/schoppenhauer] has left #lisp 12:23:55 -!- Athas [n=athas@0x50a157d6.alb2nxx15.dynamic.dsl.tele.dk] has quit [Read error: 110 (Connection timed out)] 12:29:11 Spyderco [n=nash@194.45.110.65] has joined #lisp 12:30:27 shmho [n=user@58.142.15.103] has joined #lisp 12:30:42 -!- ManateeLazyCat [n=user@119.141.230.249] has quit [Success] 12:31:29 ruediger [n=ruediger@91-115-20-76.adsl.highway.telekom.at] has joined #lisp 12:32:48 -!- bhyde1 [n=Adium@c-66-30-202-56.hsd1.ma.comcast.net] has quit ["Leaving."] 12:33:07 -!- kei__ is now known as kei 12:34:23 hyperboreean [n=none@unaffiliated/hyperboreean] has joined #lisp 12:36:01 ejs1 [n=eugen@77.222.151.102] has joined #lisp 12:36:15 -!- ejs2 [n=eugen@nat.ironport.com] has quit [Read error: 60 (Operation timed out)] 12:37:01 ryepup [n=ryepup@one.firewall.gnv.acceleration.net] has joined #lisp 12:37:16 -!- yoonkn [n=yoonkn@210.108.170.70] has quit [Read error: 60 (Operation timed out)] 12:38:05 TR2N [i=email@89.180.149.201] has joined #lisp 12:38:30 mathrick [n=mathrick@wireless.its.sdu.dk] has joined #lisp 12:40:13 -!- lujz [n=lujz@cpe-92-37-0-24.dynamic.amis.net] has quit [Remote closed the connection] 12:40:15 schoene [n=ms@cpe-76-188-180-22.neo.res.rr.com] has joined #lisp 12:45:16 -!- Phoodus [i=foo@ip68-231-38-131.ph.ph.cox.net] has quit [Read error: 104 (Connection reset by peer)] 12:49:29 yoonkn [n=yoonkn@210.108.170.70] has joined #lisp 12:51:37 -!- arbscht [n=arbscht@unaffiliated/arbscht] has quit [Read error: 110 (Connection timed out)] 12:52:02 ... writing another CL webframework... must... resist... urges... 12:52:32 -!- fe[nl]ix [n=algidus@89.202.147.18] has quit ["Valete!"] 12:52:57 alexbobp [n=alex@ppp-70-253-81-11.dsl.austtx.swbell.net] has joined #lisp 12:53:08 p_l: something wrong with hunchentoot? 12:53:11 ManateeLazyCat [n=user@119.141.230.249] has joined #lisp 12:53:36 bah, I had to move some logic in main game loop from CL to C because there were too many FFI calls being made :( went from 13.x seconds to 0.2 12:53:37 hunchentoot is not exactly a framework 12:54:10 Something wrong with UCW? 12:54:13 :-) 12:54:47 matimago: not hunchentoot. It's just that I started wondering if I could write a framework oriented around "widgets" without weblocks defview etc. and without CPS (Cause I don't get it... yet) 12:55:22 aerique: what were the FFI calls for? 12:55:26 OpenGL or something? 12:55:34 also, 13 seconds for what? :) 12:55:50 sellout [n=greg@63.107.91.105] has joined #lisp 12:55:59 p_l: probably. Assuming there are some non-local dataflow around widgets, you could get insipration from setf/get-setf-expansion/define-setf-expander. 12:56:14 p_l: AFAIK, ucw isn't as continuation-centric as it used to be 12:56:58 rsynnott: I blame my lack of experience and lisp making it seem so damn easy to just write another thing :) 12:56:58 rsynnott: or CPSed at all. 12:57:08 -!- ASau [n=user@host58-231-msk.microtest.ru] has quit ["ff"] 12:57:56 rsynnott: yeah OpenGL (or well Ogre sitting on top of it), basically setting vertex on normals an awful lot of times: http://local.wasp.uwa.edu.au/~pbourke/geometry/polygonise/ 12:58:05 -!- ejs1 [n=eugen@77.222.151.102] has quit [Read error: 60 (Operation timed out)] 12:58:18 aerique: ah 12:58:24 rsynnott: "vertex positions and normals" I meant 12:58:52 yep, I can see how that wouldn't be the most practical through FFI :) 12:58:57 shame FFI isn't faster ;) 12:59:03 minion: Logs? 12:59:04 Logs: #lisp logs are available at http://ccl.clozure.com/irc-logs/lisp/ (since 2008-09) and http://tunes.org/~nef/logs/lisp/ (since 2000) 12:59:15 rsynnott: most of the code is still in CL but I still feel dirty :) 13:00:50 *Xach* remembers that excuse next time he writes a CL library for which a C library is available 13:01:12 See, this is what you get for not having your GL client library entirely in CL. :-P 13:01:29 -!- ManateeLazyCat [n=user@119.141.230.249] has quit [Read error: 104 (Connection reset by peer)] 13:01:29 I wish there was one :( 13:01:30 aren't you supposed to batch openGL requests if you want decent performance anyway? 13:01:37 Umm... 13:01:39 ejs1 [n=eugen@nat.ironport.com] has joined #lisp 13:01:49 *nyef* points to the GLX in "portable" CLX. 13:02:36 Doesn't do direct render, and some of the stuff just doesn't work, but it's there. 13:02:44 pkhuong: I'm talking through a 3D engine and I definitely hope it does the sensible thing. 13:03:28 nyef: Yeah, I actually meant more than just OpenGL being available in CL. 13:03:41 ah, then writing some wrapper code is probably best. 13:03:57 you have to drop sending GL commands through GLX anyway if you want high performance. And unless you've got one of those weird cards, you can't really rewrite OpenGL in Lisp... 13:04:29 This 'all the world in lisp' is a kind of harmful fetishism. 13:06:41 blackened` [n=blackene@ip-89-102-28-224.karneval.cz] has joined #lisp 13:06:42 Zhivago: for me it would be to cut down on dependencies 13:06:49 Zhivago: then why python programmers have implemented a md5 module when there's a perfectly working one implemented in CL? 13:07:35 matimago: is it implemented in Python? 13:07:50 Yes, AFAIK. 13:07:54 matimago: You appear to have suffered a failure of reasoning. 13:08:17 What good would it do them to have access to an md5 module written in CL? 13:08:19 athos [n=philipp@92.250.250.68] has joined #lisp 13:08:20 Zhivago: you're critizing the NHI in Lisp, but you are not critizing it when it's done in the other PL. 13:08:44 What good would it do to us to use the md5 module writtenin python or in C or whatever when we can have one in CL? 13:08:59 matimago: it isn't 13:08:59 Really, all this has been done in FORTRAN before, why are we not using that? :-P 13:09:14 yango [n=yango@unaffiliated/yango] has joined #lisp 13:09:19 nyef: you laugh. 13:09:21 -!- Spyderco [n=nash@194.45.110.65] has quit [Read error: 104 (Connection reset by peer)] 13:09:26 matimago: md5, sha etc. that are available in Python are all implemented in C 13:09:39 matimago: CL has decent C bindings. 13:09:41 matimago: possibly linked to another library, even 13:09:54 There's nothing wrong with wasting your time re-implementing them again if it's for fun. 13:09:58 yup, libssl 13:10:04 -!- nvoorhies [n=nvoorhie@adsl-75-36-204-63.dsl.pltn13.sbcglobal.net] has quit [Read error: 104 (Connection reset by peer)] 13:10:05 But too many people are doing it because they feel dirty. 13:10:07 nvoorhies [n=nvoorhie@75.36.204.63] has joined #lisp 13:10:21 Which is why it's a kind of dangerous fetishism. 13:10:24 *rsynnott* awaits SSL in CL :) 13:10:37 just don't hold your breath 13:10:44 (assuming it hasn't been done already; I think most stuff just uses the BSD C lib) 13:10:57 Zhivago: this is long term thinking. We still hope one day to have computer systems with zero C code on them. 13:11:45 Zhivago: it can be pragmatic as well as fun 13:11:46 Isn't that what MenuetOS and the other all-ASM OSes were for? 13:11:49 is there something like numpy for CL? 13:11:55 And beside, there are a lot of CFFI libraries... 13:11:59 Zhivago: depending, of course, on the individual project 13:12:03 Zhivago: you spend a great deal of your life trying to achieve some kind of purity. why not do it in programming? 13:12:22 Zhivago: it's just not optimal to have to use a C++ library through C wrappers, but still I'm doing it since it would take me multiple man-years (if ever) to port it to CL 13:12:30 it's not good for the budget, it probably is good for the soul :) 13:12:38 I thought some people spent a good part of their life trying to -lose- purity? 13:12:39 guaqua: Well, it depends on if you want to get things done or not. 13:12:41 -!- ia [n=ia@89.169.189.230] has quit [Read error: 110 (Connection timed out)] 13:12:51 no it doesn't 13:12:51 (incf guaqua) 13:13:07 There are higher purposes than getting things done. 13:13:24 Zhivago: don't forget that in the end, we're all dead. 13:13:28 for various values of "higher" 13:13:30 mega1: what would it do? There are bindings to lapack and blas, and user friendly interfaces around them. I don't think we need an extension to be able to add vectors of floats with a decent speed. 13:13:49 Zhivago: eg. what good did it do to Nasa engineers to get things done going to the Moon, now, 40 years later, it's all lost. 13:13:57 ia [n=ia@89.169.189.230] has joined #lisp 13:14:27 seeing the big picture, not having to reuse the possibly maintenance-wise costly implementation might save you money and time in the long run 13:14:28 pkhuong: I'm thinking efficient slicing, tensors. 13:14:34 Zhivago: Lispers on the contrary tried for a higher goal, conclusion Lisp is still here. 13:14:54 matimago: So is cobol. 13:15:28 :-) 13:15:48 i think one of the reasons lisp is still here is that millions of dollars have been spent on it 13:15:52 mega1: some of the interfaces on top of lapack/blas offer slicing, iirc. Not sure re tensor operations. 13:16:02 now the money is going elsewhere 13:16:18 jdz: to java, yes 13:16:38 or did go to java for about 10 years straight. i think there's something else up next 13:16:48 ye, that's one of the reasons i'm spending my time with clojure these days 13:16:56 jdz: ISTM that the amount of money spent on Lisp is insignificant, compared to either C, C++, C#, Java, JVM or .Net. 13:17:23 jdz: remember that the whole DARPA financing of LISP has been returned on investment in only ONE software project of the US MIL. 13:17:26 matimago: you must differentiate between money being spent on the technology and advertising 13:18:06 So, Lisp = one software project. vs. uncountable number of libraries and applications developed in any other PL. 13:18:16 academia has spent considerably more time and money on lisp than the languages stated by matimago 13:18:24 pkhuong: spartns looks good, except I don't believe it can come close to blas. 13:18:31 I think that we forget that lisp is a historically filthy family of languages born of horrible pragmatism. ) 13:18:42 (in the dense case, of course) 13:18:44 guaqua: searchers may have _used_ lisp, but I'm not sure they spend a lot of money on it. 13:18:56 guaqua: or else that would have translated into lisp programming jobs. 13:19:09 matimago: researchers eat, too, don't they? 13:19:23 matimago: oh, nvm. misread. 13:19:31 matimago: academia doesn't provide a whole lot of software engineering jobs 13:19:33 mega1: I think matlisp or one of tamas papp's project have slicing at least (and call out to BLAS and LAPACK). 13:19:35 matimago: they also got some NASA stuff out of it 13:19:54 fisxoj [n=fisxoj@149.43.252.13] has joined #lisp 13:20:54 matimago: the returned investment was on *all* of AI work, not only Lisp 13:21:40 mega1: btw, do you have any feeling on making constant-lvar-p/lvar-constant-value slightly smarter and thus avoid problems with VOPs being passed all constant arguments? 13:21:56 zophy [n=sy@host-242-6-111-24-static.midco.net] has joined #lisp 13:22:44 pkhuong: I'll be back shortly 13:23:14 -!- yango [n=yango@unaffiliated/yango] has quit [Remote closed the connection] 13:23:29 yango [n=yango@unaffiliated/yango] has joined #lisp 13:23:45 p_l: even worse then. 13:24:20 -!- borism [n=boris@195-50-206-70-dsl.krw.estpak.ee] has quit [Read error: 145 (Connection timed out)] 13:24:29 -!- Pegazus [n=gawgaw@host250.190-224-109.telecom.net.ar] has left #lisp 13:24:55 hkBst [n=hkBst@gentoo/developer/hkbst] has joined #lisp 13:25:20 matimago: considering *what* kind of project it was, I'd say it was completely apriopriate 13:26:56 -!- stassats` [n=stassats@wikipedia/stassats] has quit [Read error: 110 (Connection timed out)] 13:26:57 segv [n=mb@ita4fw1.itasoftware.com] has joined #lisp 13:28:55 -!- fvw [n=sdfpme@59.39.243.199] has quit [Remote closed the connection] 13:31:16 p_l: of course, that's the problem with military research. :-) 13:32:07 when it comes to it, the vast, vast majority of military research has never produced any practical return at all 13:32:15 bombshelter13_ [n=bombshel@toronto-gw.adsl.erx01.mtlcnds.ext.distributel.net] has joined #lisp 13:32:27 death and destruction! 13:32:38 so a new life can take it all over 13:32:40 It's kept researchers off the streets where they might harm others. 13:32:41 jdz: probably the bulk of it hasn't even produced that 13:33:09 there's been an enormous amount of spending on things like ICBMs and missile shields, but of course they've never been used 13:33:18 researchers usually don't harm anything. they just produce stuff that others use to do harm. 13:34:11 rsynnott: "It's not the using, it's the having." 13:36:47 Oh, joy. I manage to get a shell prompt on my ps2, and the keyboard workingish, but C-c doesn't seem to do anything... 13:38:21 well, possibly :) 13:38:35 rsynnott: iirc DARPAs mission was actually to fund stuff that might not even have actual practical impact, but which might uncover something important/useful or just prove dead ends in research 13:39:00 (and very early ICBMs did contribute towards orbital launcher tech) 13:39:41 as for ICBMS, the true inventors were in Soviet Russia, as they had *need* unlike USA (which had convenient bomber bases) 13:40:21 (there was also considerable blurring of budgets; the Soviet Proton rocket was misleadingly referred to as an ICBM during development, though it was clearly never meant to be one; far too big) 13:40:38 -!- yahooooo [n=yahooooo@c-76-104-183-185.hsd1.wa.comcast.net] has quit [Success] 13:41:05 rsynnott: actually, early soviet ICBMs were much bigger due to much heavier payload than US designs 13:41:22 younder [n=jthing@165.244.251.212.customer.cdi.no] has joined #lisp 13:41:38 proton's payload was 20 tonnes into _orbit_ though, so presumably much higher in an ICBM role 13:41:55 (also, by the time it was under development, liquid-fuel icbms were largely obsolete) 13:42:02 USA afaik waited till getting proven thermonuclear warheads, while USSR started building ICBMs with fission nukes 13:42:27 moghar` [n=user@157.185.jawnet.pl] has joined #lisp 13:42:37 and purely fission-based nuke is much, much heavier than fusion one 13:42:37 -!- moghar [n=user@unaffiliated/moghar] has quit [Remote closed the connection] 13:44:17 also, USSR had at one point ICBM system based on orbital bombardment 13:44:56 (which IMHO beats one of their early space stations having point-defense gun) 13:44:57 So, did a fission based nuke weigh 50 tonnes? 13:45:12 first Soviet Tellar Ulam device was in 1955, before even the ICBM that was the precursor to the vostok/soyuz launchers was tested 13:46:15 fiveop [n=fiveop@g229237200.adsl.alicedsl.de] has joined #lisp 13:46:27 Zhivago: one could have been built which would weigh that much, certainly 13:46:43 rsynnott: rocket research was started earlier - I recall reading once that Stalin pushed it forward a lot 13:46:45 I guess the americans were pretty stupid. 13:47:01 The russians still have the better rockets, after all. 13:47:09 They might have just been confused. 13:47:29 Zhivago: the Americans had a major advantage in terms of siting for bombers and regional ballistic missles, so not as much pressure to develop rockets 13:47:38 (or big rockets, anyway) 13:47:48 also, very arguable who has the best rockets these days :) 13:48:07 The argument here appears to be that one side things that the americans deliberately misclassified a rocket as an icbm for budget purposes 13:48:35 The other side seems to think that they actually thought that a rocket to put a 20 tonne payload into orbit was an icbm. 13:49:19 so when do the lisp missile-control panel come into this story? 13:49:30 By the way fission nukes are much lighter than fusion nukes. And each fusin nuke has a fission nuke inside.. 13:49:39 pem [n=pem@159.226.35.246] has joined #lisp 13:50:09 oh the ever-educational #lisp 13:50:11 <3 13:50:17 younder: fission ignition devices used by modern nukes are much lighter than what was used in old fission-only designs 13:50:21 ok, nuclear weapons are way off topic. 13:50:26 i built a fission device for my dog 13:50:40 and it breaks the golden rule: no religion or politics in the programming channel. 13:50:50 -!- willb [n=wibenton@h69-129-204-3.mdsnwi.broadband.dynamic.tds.net] has quit [Read error: 110 (Connection timed out)] 13:50:54 are dogs off topic ? 13:51:11 zophy: Leash laws? 13:51:23 rockets, missiles, and all kinds of robots is always on-topic in #lisp 13:51:33 especially the human-killing robots 13:51:42 Wasn't the topic "stupid americans"? 13:51:56 ye, the onse killed by robots 13:52:06 they used lisp to calculate trajectories for icbm's 13:52:11 Yeah, the only bomb I've made today uses alien (sb-alien) tecknology 13:52:22 heh 13:52:30 zophy: did they?! 13:52:34 Imagine the joy of having a REPL on a warhead! 13:52:39 i'm sure of it 13:53:07 i actually watched "war games" recently 13:53:13 like last month 13:53:25 deepfire: while today warheads might not give you good REPL experience, that soviet orbital bombardment system might... 13:53:32 good movie, they used lisp in it 13:53:36 p_l: it never went live, though 13:53:37 I was at the smithsonian air and space museum a week ago, but it's still of topic. ;) 13:53:41 Some sap pointers got awry and crashed my lisp system.. 13:53:42 jdz, do you know that CLHS has a direct reference to that film? 13:53:46 :) 13:53:52 deepfire: nope, where? 13:53:55 rsynnott: they did at least one test 13:54:28 rsynnott: upon seeing it, USA apparently crapped in pants and quickly got a treaty banning weaponry in orbit :P 13:54:33 and then it was banned by treaty (the same one that broke the orion project finally :( ) 13:55:19 jdz: Somewhere in the discussion of the condition system. Might have been ERROR. 13:55:28 error not continuing has some strange reference 13:55:30 i'll just grep my local CLHS copy 13:55:42 jdz, http://www.lispworks.com/documentation/HyperSpec/Body/f_error.htm 13:56:32 deepfire, nyef: thanks 13:57:25 bgs100 [n=ian@unaffiliated/bgs100] has joined #lisp 13:57:26 lujz [n=lujz@cpe-92-37-0-24.dynamic.amis.net] has joined #lisp 13:57:46 Ok, see, nuclear weapons are never offtopic -- they are mentioned in CLHS! 13:58:14 ye, let's play global thermonuclear war\ 13:58:59 jdz: Mmm... A very silly card game. 13:59:10 *Fade* imagines a topic bypass whereby any topic was fair game so long as it was phrased as a valid sexp. 13:59:14 nyef: you have it on punchards? :) 13:59:49 ASau [n=user@193.138.70.52] has joined #lisp 14:00:02 anyway, there is also the game called "defcon" which should look very familiar to those seen the movie (and the other way around) 14:00:17 fe[nl]ix [n=algidus@89.202.147.18] has joined #lisp 14:00:39 bhyde [n=Adium@67.95.132.162] has joined #lisp 14:00:43 dv_ [n=dv@83-64-248-68.inzersdorf.xdsl-line.inode.at] has joined #lisp 14:01:16 fvw [n=sdfpme@59.39.243.199] has joined #lisp 14:01:18 (proposition (proposer myself) (proposee jason) (what (play-game (game thermonuclear-war))) (when now)) 14:01:32 (proposition (proposer myself) (proposee jason) (what (play-game (game thermonuclear-war (option global)))) (when now)) 14:03:02 -!- jlf` [n=user@netblock-208-127-247-82.dslextreme.com] has quit [Read error: 110 (Connection timed out)] 14:05:58 syamajala [n=syamajal@140.232.182.22] has joined #lisp 14:08:19 danlei [n=user@217.226.218.153] has joined #lisp 14:08:46 Setting --dynamic-space-size doesn't seem to work on SBCL 1.0.25.debian. (This: http://identi.ca/notice/3809855 always shows the same value no what I enter at startup) 14:09:31 Zhivago: there's also a good reason why to rewrite a library in CL instead of C. It is that of program verification/certification. It would be easier to verify automatically a lisp program than a C program. (But even without a formal automatic verification, I'd feel safer with a CL library, knowing that there are less bugs and foremost, less run-time crash possibility in CL than in C). 14:10:04 matimago: And people wonder why python and ruby are popular languages :) 14:10:28 Zhivago: have you tried to write a CFFI interface to a C library? 14:10:36 -!- fe[nl]ix [n=algidus@89.202.147.18] has quit ["Valete!"] 14:10:38 How many times did you crash your image because of that C library? 14:10:47 Um, not at all? 14:10:48 aerique: that seems quite convoluted. pkhuong suggested just (define-alien-variable dynamic-space-size unsigned-long) 14:11:01 To be honest, I don't really have problems talking to C code. 14:11:08 aerique: oh, sorry! that's the same thing 14:11:17 I guess this is a special trait that python programmers must also have. 14:12:02 It's not the talking to C code the problem. It's that C code is much less amenable to interactive and exploratory use, because when you make a mistake in the calling sequence, C code can easily overide anything in memory rather than just signal a condition. 14:12:37 Perhaps python crashes as much as its C libraries, or perhaps they run their C libraries in a forked child? 14:12:40 And that's why you don't use C code for exploratory use. 14:12:54 rdd [n=user@c83-250-157-93.bredband.comhem.se] has joined #lisp 14:13:01 matimago: Ah, you are completely ignorant of it. 14:13:09 Of python, yes. 14:13:21 I only know about the significant indent. 14:13:31 matimago: No. They call C code directly. They don't have significant problems. 14:13:46 fe[nl]ix [n=algidus@89.202.147.18] has joined #lisp 14:13:50 -!- densem [n=fasteez2@218.217.72-86.rev.gaoland.net] has quit ["reboot"] 14:13:53 matimago: the core is implemented in C, and their FFI is done the other way around - you write it in C calling CPython's API 14:13:54 How do they deal with memory corruption then? 14:13:56 Xach: yes, nevermind the problem seems to be me but I haven't quite figured it out yet 14:14:36 willb [n=wibenton@compsci-wifi-39.cs.wisc.edu] has joined #lisp 14:14:49 In some regards, python has the benefit of having filthy, practical developers rather than people waffling on about code verification techniques that they'll never actually use. 14:14:51 -!- sykopomp [n=sykopomp@unaffiliated/sykopomp] has quit [Connection timed out] 14:15:03 matimago: by carefully writing C FFI. You don't do "exploratory" FFI coding in CPython or Ruby 14:15:33 at least there's PyRex, it made writing FFI much easier 14:15:52 pl: There's also that inline C++ thingy. 14:16:36 gko [n=gko@122-116-15-138.HINET-IP.hinet.net] has joined #lisp 14:17:52 How do they prevent you to pass a pointer to an object of class A to a C function expecting a pointer to an "object" of class B in python? Is their FFI strongly typed? CFFI is not typed. 14:18:15 That'd be a defect of CFFI, clisp FFI is better in this respect. 14:18:33 Zhivago: that's probably PyRex you are talking about :) 14:19:17 mrSpec_ [n=NoOne@88.208.105.1] has joined #lisp 14:19:17 -!- mrSpec [n=NoOne@88.208.105.1] has quit [Read error: 104 (Connection reset by peer)] 14:19:26 CFFI seems to be using SBCLs default dynamic space size instead of the one I'm supplying 14:19:28 -!- mrSpec_ is now known as mrSpec 14:19:40 matimago: you have only C's type system. CPython's FFI is pure C (with possible generation through various tools) 14:19:42 p_l: No. I think this was called CWeave or something like that -- it was actual C++ code inline in python. 14:20:12 yahooooo [n=yahooooo@c-76-104-183-185.hsd1.wa.comcast.net] has joined #lisp 14:20:26 -!- shmho [n=user@58.142.15.103] has quit ["ERC Version 5.3 (IRC client for Emacs)"] 14:20:29 Zhivago: PyRex looks similar - you have Python-like statements/code with bits of direct, inline C. Then you run PyRex on it and you get _modulename.so and modulename.py 14:20:31 You can sorta inline C++ code in ECL as well. Let me find the mailinglist post. 14:20:47 http://www.scipy.org/Weave <- there. 14:20:48 p_l: So the C programmer must check the python object type and signal an error to python. 14:21:51 matimago: kind of. Don't remember details 14:24:02 -!- fe[nl]ix [n=algidus@89.202.147.18] has quit ["Valete!"] 14:24:06 christ, sourceforge.. nevermind 14:24:14 matimago: http://docs.python.org/c-api/arg.html 14:26:29 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 14:27:06 tmh [n=thomas@pdpc/supporter/sustaining/tmh] has joined #lisp 14:27:11 davazp [n=user@56.Red-79-153-148.dynamicIP.rima-tde.net] has joined #lisp 14:27:13 Greetings. 14:27:20 hello tmh 14:28:29 -!- bhyde [n=Adium@67.95.132.162] has quit [Read error: 110 (Connection timed out)] 14:29:52 -!- bdowning [n=bdowning@mnementh.lavos.net] has quit [Read error: 60 (Operation timed out)] 14:30:53 jlf` [n=user@unaffiliated/jlf] has joined #lisp 14:31:08 bdowning [n=bdowning@mnementh.lavos.net] has joined #lisp 14:32:47 densem [n=fasteez2@218.217.72-86.rev.gaoland.net] has joined #lisp 14:32:48 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 14:32:57 -!- ejs1 [n=eugen@nat.ironport.com] has quit [Read error: 110 (Connection timed out)] 14:33:34 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 14:34:40 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 14:36:33 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 14:37:19 ThomasI [n=thomas@unaffiliated/thomasi] has joined #lisp 14:38:14 The FFI (CFFI) just allows lisp to call C libraries directly, right? 14:38:41 Trying to see if I understand what it is doing correctly. 14:38:41 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 14:38:53 -!- fvw [n=sdfpme@59.39.243.199] has quit [Remote closed the connection] 14:38:57 -!- alinp [n=alinp@86.122.9.2] has quit [Read error: 60 (Operation timed out)] 14:39:17 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 14:41:00 ace4016 [i=ace4016@cpe-76-168-248-118.socal.res.rr.com] has joined #lisp 14:43:43 SandGorgon [n=OmNomNom@122.160.41.129] has joined #lisp 14:44:10 TDT: correct. 14:44:22 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 14:44:45 TDT: only that a direct call is rarely possible, given the differences in ABI and data types. 14:45:18 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 14:45:20 ikki [n=ikki@201.155.75.146] has joined #lisp 14:45:39 matimago: Yeah, that makes sense. I wish there were easy ways to call python libraries from within lisp..my life would be so much easier, heh 14:45:52 Pepe___ [n=ppjet@78.113.15.169] has joined #lisp 14:45:52 TDT: eg. a fixnum such as 42 is encoded as 0x000000A8 in Lisp, while it's encoded as 0x0000002A in C. 14:46:05 TDT: well, you COULD write a little C layer to link the two :) 14:46:05 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 14:46:27 Why would a C layer be needed? 14:46:37 TDT: Well we could provide the CPython FFI API to be compatible with all the python FFI libraries... 14:46:42 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 14:46:46 oh, actually, I suppose you could do it directly, yep 14:46:56 Or it could be implemented as CFFI callbacks :-0 14:47:09 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 14:47:14 Now I'm getting a minor headache contemplating alien-metaobject-protocols. 14:47:19 ejs [n=eugen@77.222.151.102] has joined #lisp 14:47:26 matimago: heh, I think a lot of htis stuff is over my head..I need to read some programs that use the cffi I think to understand this a bit, heh. 14:47:30 -!- ejs [n=eugen@77.222.151.102] has quit [Remote closed the connection] 14:47:33 Yes, we'd have to provide and get back python objects instead of C objects :-0 14:47:42 Grr, this Shift is broken?! 14:47:55 here are some parens to cut and paste: (((()))) 14:48:00 Thanks. 14:48:46 ejs [n=eugen@77.222.151.102] has joined #lisp 14:50:13 -!- Pepe___ [n=ppjet@78.113.15.169] has quit [Read error: 60 (Operation timed out)] 14:50:53 Pepe___ [n=ppjet@78.113.38.65] has joined #lisp 14:51:52 isn't python very eval-friendly? .. lisp <--socket--> python doing eval with stuff received .. *shrug* 14:52:24 at ILC, Duane Rettig and presented on a marriage of python to Allegro CL. 14:52:37 showing interactive use of the Allegro debugger on python source 14:53:03 http://www.international-lisp-conference.org/2009/tutorials#debug 14:53:17 Willem Broekma was the other fellow 14:53:26 Do you know why when I'm trying to use (drakma:http-request ... ) in Slime I got "Lisp connection closed unexpectedly: connection broken by remote peer" ? In SBCL all is working fine :S 14:53:55 mrSpec: perhaps it is returning a character that can't be encoded over your slime connection 14:54:16 hmm what can I do with it? 14:54:28 mrSpec: you should be able to tell slime to use utf-8 14:54:39 mrSpec: i set slime-net-coding-system to utf-8-unix 14:54:53 ok I'll set it, thankx 14:55:01 -!- jdz [n=jdz@85.254.211.133] has quit ["Somebody booted me"] 14:56:21 -!- ejs [n=eugen@77.222.151.102] has quit ["This computer has gone to sleep"] 14:57:02 ejs [n=eugen@77.222.151.102] has joined #lisp 15:00:58 -!- Pepe_ [n=ppjet@78.114.139.251] has quit [Success] 15:01:36 -!- prg [n=prg@ns.alusht.net] has quit ["Lost terminal"] 15:02:01 sohail [n=Sohail@unaffiliated/sohail] has joined #lisp 15:02:16 dlowe [n=dlowe@24.91.154.83] has joined #lisp 15:02:25 artagnon [n=artagnon@unaffiliated/artagnon] has joined #lisp 15:02:40 -!- fisxoj [n=fisxoj@149.43.252.13] has quit [Read error: 113 (No route to host)] 15:02:48 SBCL is neither a compiler or an interpreter, right? And it isn't a virtual machine either? Then what is it? 15:02:56 I'm running into a maximum with the amount of foreign memory that SBCL wants to allocate with make-alien even though I've increased its dynamic size on the command line. 15:03:45 I told slime to use utf and now all is working fine :D 15:04:05 artagnon: make-alien calls malloc(), that is not related to the dynamic space. 15:04:05 I'm not even sure I need to increase its dynamic size on the command line, I realize now. 15:04:12 right :) 15:04:30 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 15:04:35 s/artagnon/aerique 15:04:44 c|mell [n=cmell@p2225-ipngn1203marunouchi.tokyo.ocn.ne.jp] has joined #lisp 15:04:48 aerique: if anything, increasing the dynamic space size will reduce the amount of mallocable memory. 15:04:59 artagnon: Who told you SBCL is neither a compiler nor an interpreter? 15:05:04 artagnon: SBCL is both a compiler and an interpreter, actually. 15:05:06 luis: Am I out of luck and should I find a way around it (allocating smaller segments) 15:05:20 artagnon: It's a program that includes an interpreter, a compiler and a lot of library functions. 15:05:29 beach: well, I guess it's both then. 15:05:47 rgz [n=rogue@unaffiliated/rgz] has joined #lisp 15:06:03 pkhuong: ah, that clears it. 15:06:07 artagnon: You didn't answer my question. 15:06:15 lnostdal: While doable, I suspect it would give us even worse performance than with CFFI :) 15:06:55 beach: I read it somewhere. Common Lisp isn't classified as completely compiled or interpreted anywhere. 15:06:55 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 15:06:59 What are the general opinions on phosphorous? 15:07:16 rgz: quite amusing 15:07:34 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 15:07:35 one of the few intentionally hilarious new Lisp implementations. 15:08:20 alinp [n=alinp@86.122.9.2] has joined #lisp 15:08:27 Does SBCL have a fraction of a VM? 15:10:05 ? 15:10:11 artagnon: Common Lisp is a language. You can compile it, and you can interpret it. 15:10:30 artagnon: Why do you want to know? Your questions are a bit strange! 15:10:37 lisp systems 15:10:53 rarely use intermediate bytecode right ? 15:11:09 luis: ofcourse. I meant what SBCL does. 15:11:14 densem: CLISP and CMUCL do. 15:11:14 ( i guess historically it's been done ) 15:11:22 oh thanks :) 15:11:45 densem: but CMUCL also has a native compiler 15:11:46 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 15:11:51 beach: Well, I was actually comparing Common Lisp to Clojure. Clojure seems to be pretty stubborn about using the libraries for JVM (the Java libraries ie) 15:12:08 artagnon: SBCL compiles everything by default, but it's got an interpreter you can use if you want. 15:13:11 luis: compiles it to what? some intermediate form that it interprets, right? 15:13:18 something like LLVM? 15:13:25 luis: x86 machine code if you're on x86. 15:13:26 etc. 15:13:31 artagnon: it compiles to native code. 15:13:53 -!- rouslan [n=Rouslan@unaffiliated/rouslan] has quit [Read error: 104 (Connection reset by peer)] 15:13:54 artagnon: check out DISASSEMBLE 15:14:07 WHAT!? Where's my ELF executable then? 15:14:11 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 15:14:29 Why would you need an ELF executable? 15:14:32 mattrepl [n=mattrepl@76.104.2.182] has joined #lisp 15:14:48 artagnon: see sb-ext:save-lisp-and-die 15:14:49 artagnon: your CPU doesn't know anything about ELF. 15:14:54 schme: Just saying. If it's already native code, why not make a nice elf-executable? 15:15:06 schme: ofcourse, but my OS does. 15:15:12 artagnon: You can if you want. 15:15:14 artagnon: sb-ext:save-lisp-and-die generates an executable for you. 15:15:21 *artagnon* looks 15:15:21 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 104 (Connection reset by peer)] 15:15:51 artagnon: but it'll of course, include the runtime support, so it's somehwat large. 15:16:08 oh I see 15:16:10 artagnon: really, executable delivery (e.g. as an ELF) and compilation to native code are quite independent. 15:16:12 *artagnon* is surprised 15:16:30 pkhuong: ofcourse, I understand. 15:16:55 I've seen Python implementations that produce an ELF: basically an interpreter packaged with a program in an ELF. 15:17:32 artagnon: think of it more like smalltalk likes to behave than what madness python is up to. 15:17:34 luis: I understand. The runtime is huge, so people rarely *want* to produce ELFs 15:18:00 artagnon: People don't want ELFs because we prefer just to keep stuff running ;) 15:18:08 *artagnon* nods 15:18:25 artagnon: right. A common technique is to strip undeed bits of the runtime when generating the executable. But SBCL doesn't have one yet, though there have been some attempts. 15:18:48 *unneeded 15:18:51 luis: ah, I see. Does that process have a name? 15:18:52 luis: the image compresses very nicely. 15:19:12 luis: We would be able to produce very efficient ELFs like C does then. 15:19:12 artagnon: tree shaking 15:19:42 artagnon: I don't think gcc produces much more efficient ELFs. 15:19:53 I'm not sure how that would work. But oh well :) 15:19:55 artagnon: well, C programs usually depends on libc.so, libm.so, etc. Don't forget about that. 15:20:05 luis: ofcourse. libc 15:20:14 that's the first thing an OS needs :) 15:20:26 well if it's a unixy C-based OS, sure. 15:20:29 ;) 15:20:45 What isn't? 15:20:48 minion: what is movitz? 15:20:49 maybe you need to ask my master, chandler - he knows a lot 15:20:53 hepp. 15:20:59 maybe that's not how it works. 15:21:04 *schme* never understands bots. 15:21:04 minion: tell schme about movitz 15:21:05 schme: please see movitz: Movitz is a Common Lisp implementation that targets the x86 PC architecture "on-the-metal". http://www.cliki.net/movitz 15:21:09 ah :D 15:21:16 compressing the runtime (sbcl) works well by the way, artagnon .. they end up at about 4-5 MB 15:21:18 *artagnon* looks 15:21:31 -!- ejs [n=eugen@77.222.151.102] has quit [Read error: 110 (Connection timed out)] 15:21:47 artagnon: well, there's for example NT, which isn't that C-centric, there's VMS, which has runtime completely unrelated to C... 15:21:49 -!- mattrepl [n=mattrepl@76.104.2.182] has quit [] 15:21:51 lnostdal: I see. But it's the complete runtime. Without the stripping. motivz seems to be nice. 15:21:52 -!- rgz [n=rogue@unaffiliated/rgz] has left #lisp 15:21:59 NT? 15:22:06 minion: tell me about NT 15:22:07 Sorry, I couldn't find anything in the database for ``NT''. 15:22:48 *p_l* *headdesks* 15:23:05 hm, VMS is interesting. 15:24:06 artagnon: NT = Windows "New Technology", current production version - Windows NT6.0 aka Windows Vista/2008, next NT6.1 aka Windows 7/2008R2 15:24:14 Windows NT! L.O.L 15:24:41 It isn't C-centric? 15:25:02 ie. it's core OS components aren't written in C? 15:25:14 artagnon: looking at how often C/C++ apps had to ship with their own C runtimes, it isn't 15:25:24 -!- mrSpec [n=NoOne@88.208.105.1] has quit [Remote closed the connection] 15:25:25 what!? 15:25:40 what sort of a stupid OS doesn't include a C/C++ runtime!? 15:25:50 it might be written in C, but it doesn't mean that it's runtime is based upon the concept that "ANSI C is our API" 15:25:53 oh, don't mind answering that. 15:25:55 *its 15:26:05 artagnon: erm, quite a lot? 15:26:39 p_l: really? can you give me more examples? Don't give me anything that M$ has touched. 15:27:00 mrSpec [n=NoOne@88.208.105.1] has joined #lisp 15:27:16 mattrepl [n=mattrepl@76.104.2.182] has joined #lisp 15:27:34 kolibrios (: 15:27:46 so the Windoze C runtime isn't ANSI C? But some mangled language that they came up with? Like the shit that Turbo C++ compiles? 15:27:48 artagnon: What is a "C/C++ runtime" and why do you think an OS is stupid if it doesn't include one? 15:28:18 artagnon: let's see, VMS, Inferno, classic MacOS, L4 is a stretch as it's hardly a full OS by itself, Genera xD, TOPS-10/20, ITS (ok, I'm getting into strange and stranger ;-D) 15:28:26 beach: The equivalent of libc. Because many many programs are written in C/C++ 15:28:41 artagnon: that doesn't mean the OS is required to carry a runtime 15:28:43 artagnon: The problem with unix is that it is pretty much designed to make people write applications in C. It's madness (: 15:28:58 -!- mstevens [n=mstevens@zazen.etla.org] has quit ["leaving"] 15:29:05 mstevens [n=mstevens@zazen.etla.org] has joined #lisp 15:29:06 schme: Oh, I guess I haven't really looked outside *nix 15:29:12 artagnon: you could say that it's the other way around - ANSI C adopted Unix runtime as C runtime 15:29:43 ANSI C was defined by Unix's C runtime!? Wow! That is something. 15:30:06 artagnon: I'm writing a lil' os at the moment that doesn't have a line of C anywhere near it. (: 15:30:17 p_l: L4::Pistachio is some failed microkernel project afaik. 15:30:30 it was even referenced in one of the old releases of "The C programming language" that other C implementations made many of the functions based on the interfaces in UNIX 15:30:31 schme: Wow, that must be really awesome :) 15:30:37 artagnon: Not really :P 15:30:39 artagnon: except it's not failed 15:30:51 *artagnon* looks 15:30:51 schme : for what hardware ? 15:31:01 artagnon: Also I seem to remember that my old zx spectrum was pretty much based around basic and z80 assembly. No C there (: 15:31:06 densem: ARM, of course! (: 15:31:15 L4::Pistachio is just one of the implementations (original L4 was written in pure assembly) 15:31:15 haha, obviously 15:31:17 densem: It's just a lil' forth system ;) 15:31:27 haha, obviously (bis) 15:31:39 -!- alinp [n=alinp@86.122.9.2] has left #lisp 15:31:45 p_l: I see. 15:31:50 I planned to dream about fantasizing coding something like that too ¬¬ 15:32:03 Interesting. So *nix forces C on everyone :p 15:32:09 Madness. 15:32:12 fvw [n=sdfpme@59.39.243.199] has joined #lisp 15:32:12 -!- dv_ [n=dv@83-64-248-68.inzersdorf.xdsl-line.inode.at] has quit ["Verlassend"] 15:32:15 -!- froydnj [n=froydnj@gateway.codesourcery.com] has quit [Remote closed the connection] 15:32:20 densem: It's a beagleboard to be precise. It's a really lovely thing. Right now running the linux. I'm trying to clear up time to dig deep into sbcl so I can maybe start porting it. 15:32:29 or a bare-metal binary repl , to play with hardware in debugger-style 15:32:33 *awe* 15:32:35 artagnon: VMS, otoh, is unholy mess of VAX assembly (MACRO32), BLISS, C (in newer parts), Alpha assembly (MACRO64), IA-64 Assembly and possibly some more :P 15:32:37 But sbcl seems a big nut to crack for me. 15:33:19 p_l: Sounds worse than you describe it now :p 15:33:23 Oh I just read a bit more about the specs of that board, to make a tiny 1080p :p 15:33:30 schme: What CPU type is that? 15:33:30 *decoder 15:33:44 -!- Pepe___ is now known as Pepe_ 15:33:50 Oh, ARM? 15:33:50 Ti OMAP3 I think 15:33:56 artagnon: oh, and all of those exist in one system image on IA-64 (I'm not sure if they removed MACRO-32 code on IA-64) 15:34:00 nyef: cortex a-8@600MHz 15:34:06 Hrm. 15:34:12 thumb2 and everything. 15:34:20 with 256MB of RAM sweetness. 15:34:22 fisxoj [n=fisxoj@149.43.252.13] has joined #lisp 15:34:26 hehe 15:34:37 schme: beagleboard? 15:34:40 p_l: yup. 15:34:43 schme: nice thingy :) 15:34:48 I've got part of an arm port of SBCL, and part of a no-host-OS x86 port of SBCL... 15:34:50 very! 15:34:52 personally i'd love to do that with a tegra soc 15:34:55 hawkbill [n=pradyus@117.192.6.120] has joined #lisp 15:35:06 nyef: cools! 15:35:06 Which ring of Hell am I going to if I start my own text processing library? 15:35:14 Hm. Is LLVM a step to eliminate this C madness (ie. every compiler's desire to compile to C for speed etc.)? 15:35:16 -!- SandGorgon [n=OmNomNom@122.160.41.129] has quit [Read error: 110 (Connection timed out)] 15:35:29 artagnon: Why would compiling to C == speed? 15:35:34 artagnon: people compile to C for convenience. 15:35:41 tmh: what do you want that for? 15:35:46 speed of portability maybe :fff 15:35:46 artagnon: I tell you I have written some really slow C ;) 15:36:12 tmh: I think the one where you get frozen in a block of ice. 15:36:14 funny that lisp told me the fact that C != fastness 15:36:16 schme: well, because it's the closest to bare metal you can get, without touching x86 assembly. 15:36:27 Actually, compiling to (legal) C is pretty hard. 15:36:29 artagnon: That just makes no sense whatsovere. 15:36:29 pkhuong: Right. 15:36:52 http://www.cs.virginia.edu/papers/Hitting_Memory_Wall-wulf94.pdf 15:36:55 rsynnott: Because I don't want to use regular expressions(cl-ppcre) and I don't want to go through the trouble of using an actual parser-generator. 15:37:02 not that generating x86 assembly is that hard. 15:37:25 schme: Ok, I don't mean C, it could be x86 assembly too, but I think that's harder. 15:37:30 pkhuong: what's the convinience? 15:37:33 runtime libs? 15:37:50 artagnon: I'm more thinking that C being close to bare metal was odd :) 15:37:54 artagnon: but oh well! 15:38:08 As far as I'm concered you're all crazy 15:38:20 schme: C is the aprox equivalent of portable assembly I'd say. 15:38:32 Actually, spending time when you fetch your operands to check their types, and once you complete an operation, to check the result (overflow, etc), could be faster, since you'd introduce cpu cycles in the pipelines without a need for memory access, therefore without stalling. 15:38:47 schme: which is why it's so easy for compilers to optimize C and produce efficient assembly code. 15:38:56 so much harder in the case of Ruby, for example. 15:39:05 no, just more convenient. You avoid some busywork wrt reg alloc, handling branches, calling conventions, etc. 15:39:05 The FreeBSD people are looking into replacing gcc with LLVM. 15:39:18 Because it's so high level and cute that it's a convinience trade-off 15:39:22 artagnon: what's the point of producing code that's close to the machine? You want code that's close to your problem! 15:39:37 C is nowhere near easy to optimise. 15:39:40 clang website says linux is almost compilable 15:39:49 matimago: ofcourse. Use Ruby then. Optimization is impossible. 15:39:52 pkhuong: it isnt'? 15:39:53 isn't C just more cache friendly ? 15:40:05 pkhuong: I thought it was designed to be easy to optimize 15:40:07 Why is optimization or ruby impossible? 15:40:12 * locality-wise 15:40:22 It's just that nobody wrote a ruby compiler. 15:40:34 matimago: That's what I thought too. 15:40:35 artagnon: look at the hairy analyses C compilers have to pull off. 15:40:37 densem: that'd depend on your algorithm, not on your language. 15:41:01 matimago: and your ability to lay your data out as needed. 15:41:10 that was my point 15:41:18 schme && matimago: I'm guessing it's because of the language's design 15:41:37 artagnon: I think it is very much more related to that no one has bothered to write a compiler to do it. 15:41:42 For example, Python3 is significantly faster than the previous version. 15:41:43 Why? 15:41:52 I think C / Os / Cpu vendors all influenced themselves to be in the state they're all optimized together 15:42:09 schme: That's very odd. It's very popular, but MUCH slower than even Python. 15:42:16 schme: people haven't done it because it's hard. 15:42:29 artagnon: I think maybe people don't do it because it claims to be a scripting language. 15:42:44 schme: well, the interpreter is slow as hell. 15:42:53 it's not important 15:42:59 there's other gains 15:43:05 right. 15:43:06 schme: My point is simple. Ruby is SLOW. Why? 15:43:22 Fare [n=Fare@c-98-216-111-110.hsd1.ma.comcast.net] has joined #lisp 15:43:22 *sigh* 15:43:27 densem: you're probably right there. 15:43:28 artagnon: nobody bothered to make it less slow. 15:43:36 pkhuong: is that the only reason? 15:43:37 artagnon: no one has decided to write a compiler or interpreter to make it all go very fast. Because no feels they need it faster. 15:43:40 artagnon : exec speed is not the problem nowadays 15:43:50 that's the main reason, yes. 15:43:59 Historically confirmed on all the PL we invented. 15:44:06 -!- blandest [n=user@softhouse.is.ew.ro] has quit [Success] 15:44:10 artagnon: there's little that heroic efforts can't overcome. 15:44:11 oh, I see. 15:44:20 Today, Fortran is the fastest, because a lot of people spent a lot of time optimizing it. 15:44:21 there's unladen swallow for python (llvm compiling, gil is going to be removed) 15:44:24 artagnon: I have not used ruby much, but I haven't seen anything related to it that would make it impossible to generate fast code from it. 15:45:01 stuart71 [n=user@75-145-221-229-Memphis.hfc.comcastbusiness.net] has joined #lisp 15:45:07 schme, artagnon: there's a closed-source and quick ruby implementation called maglev 15:45:18 *artagnon* looks 15:45:22 btw, VM backend for scripting languages seems to be rewarding 15:45:23 schme: I see. 15:45:44 php, ruby .. their JVM port are all pretty neat speed-wiz 15:45:47 wise* 15:46:05 stuff like type-checking at runtime makes it impossible to produce more efficient code, right? It's a conviniece-speed trade-off then 15:46:13 huh? 15:46:17 *beach* Thinks it has to do with available funds. Java is pretty hard to compile to efficient code, but with enough money from Sun, they managed. 15:46:27 artagnon: run-time type checking is free. 15:46:34 artagnon: see the above url about the memory wall. 15:46:35 er. 15:46:40 *artagnon* looks 15:46:41 artagnon: You are having quite a few naive opinions today. 15:46:57 at the end it all comes down to communication bandwith 15:47:03 *artagnon* shrugs 15:47:04 *densem* try to be smart ^^ 15:47:36 fine, forget type-checking. 15:47:53 Memory Latency in 1991 was 14 CPU clocks; in 2003 was 400 CPU clocks... 15:47:58 matimago: not really, it costs you instruction bandwidth and space. 15:48:01 stuart71` [n=user@75.145.221.229] has joined #lisp 15:48:18 artagnon: C was designed to be dead simple while retaining similarity to its ancestors 15:48:36 abeaumont_ [n=abeaumon@84.76.48.250] has joined #lisp 15:48:38 pkhuong: right, it would be better with a different microcode, such as a LispMachine one! :-) 15:48:45 That it is. The C Programming Language is not a big book. 15:48:45 -!- abeaumont_ [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 15:48:46 matimago: more like 200, worst case, these days. 15:49:14 carlocci [n=nes@93.37.211.201] has joined #lisp 15:49:21 It's ancestor was B, which didn't have types other than byte AFAIR. 15:49:34 minion: thwap for pbusser3 15:49:35 pbusser3: have a look at thwap: THWAP! http://www.angryflower.com/bobsqu.gif and http://www.angryflower.com/itsits.gif (see also: http://www.unmutual.info/misc/sb_itsits.mp3 ) 15:49:46 schme: you know what was the common answer every time someone at bell labs whined for more complex features directly in the language? "If you want PL/I, you know where to find it" :D 15:49:49 -!- stuart71 [n=user@75-145-221-229-Memphis.hfc.comcastbusiness.net] has quit [simmons.freenode.net irc.freenode.net] 15:49:49 -!- pem [n=pem@159.226.35.246] has quit [simmons.freenode.net irc.freenode.net] 15:49:49 -!- zophy [n=sy@host-242-6-111-24-static.midco.net] has quit [simmons.freenode.net irc.freenode.net] 15:49:49 -!- Edico [n=Edico@unaffiliated/edico] has quit [simmons.freenode.net irc.freenode.net] 15:49:49 -!- holly_ [n=holly@ronaldann.demon.co.uk] has quit [simmons.freenode.net irc.freenode.net] 15:49:49 -!- peddie [n=matthew@67.169.49.37] has quit [simmons.freenode.net irc.freenode.net] 15:49:49 -!- ineiros [n=ineiros@james.ics.hut.fi] has quit [simmons.freenode.net irc.freenode.net] 15:49:50 -!- Buganini [n=buganini@security-hole.info] has quit [simmons.freenode.net irc.freenode.net] 15:49:50 -!- hefner [n=hefner@c-69-140-128-97.hsd1.md.comcast.net] has quit [simmons.freenode.net irc.freenode.net] 15:49:50 -!- phadthai [i=mmondor@ginseng.pulsar-zone.net] has quit [simmons.freenode.net irc.freenode.net] 15:49:50 -!- koollman [n=samson_t@sd-10510.dedibox.fr] has quit [simmons.freenode.net irc.freenode.net] 15:49:50 -!- Orest^bnc [n=Orest@hates-script-kiddies.ath.cx] has quit [simmons.freenode.net irc.freenode.net] 15:49:50 -!- moesenle [n=moesenle@atradig124.informatik.tu-muenchen.de] has quit [simmons.freenode.net irc.freenode.net] 15:49:50 -!- Bootvis [i=bob@baltar.lan.endoria.net] has quit [simmons.freenode.net irc.freenode.net] 15:49:50 -!- rbancroft [n=rumble@S01060014bf54b5eb.cg.shawcable.net] has quit [simmons.freenode.net irc.freenode.net] 15:49:50 -!- mbishop [n=martin@unaffiliated/mbishop] has quit [simmons.freenode.net irc.freenode.net] 15:49:50 -!- michaelw [i=michaelw@lambda.foldr.org] has quit [simmons.freenode.net irc.freenode.net] 15:49:50 -!- specbot [n=specbot@common-lisp.net] has quit [simmons.freenode.net irc.freenode.net] 15:49:54 abeaumont_ [n=abeaumon@84.76.48.250] has joined #lisp 15:50:00 p_l: hahaha. 15:50:03 -!- abeaumont_ is now known as abeaumont 15:50:04 :( So C being close to Assembly doesn't make it any easier for it to produce efficient assembly code? Someone could write a compiler to produce equally efficient assembly code from say, Ruby with the same effort? 15:50:08 Lol! 15:50:13 artagnon: I have no idea how C is close to assembly. 15:50:14 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 15:50:23 artagnon: actually it's easier to optimize higher-level languages 15:50:38 artagnon: But indeed. Anyone can write a compiler for any language to produce efficient assembly for any CPU. 15:50:42 artagnon: because you can optimize whole algorithms, cases etc. easily 15:50:46 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 15:51:08 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 15:51:16 rread [n=rread@nat/sun/x-f3b97c55669f4ac8] has joined #lisp 15:51:25 Fantastic. Then why is the world stuck with producing faster compilers just for C!? Why aren't there super-fast compilers for say, Java? 15:51:28 Most C constructions translate 1:1 to assembly constructions. 15:51:39 pbusser3: correct. 15:51:41 In anycase, that memory wall problem will be worsened by multicores. 15:51:41 schme: ^^ 15:51:45 artagnon: The thing with C is that the programmer has to know how to write efficient assembly code. The higher-level the language (the more you can express "intent"), the more leeway the compiler has to optimize and do things that may seem non-intuitive. 15:51:46 artagnon: actually, Java is quite close to C in terms of speed 15:51:49 artagnon: modern ARM processors have a Jazelle state for running java byte code more efficient. 15:51:54 yeah but C permit side effects everywhere.. then compilation goes crazy right ? 15:51:59 artagnon: C shifts the burden to the programmer. It's more that a naive C compiler can still be coerced into generating decent code. If you want to move some of the work to the compiler's side, C becomes harder to *optimise*. 15:52:03 p_l: So are good Lisp compilers. 15:52:05 schme: and nearly no-one uses that - Jazelle was a flop, iirc 15:52:13 artagnon: and there are CPUs designed for running forth. 15:52:20 And I'm thinking switching back to 100 Hz insteaed of 1000 Hz for my linux scheduler... (to keep the caches loaded longer on each job)... 15:52:22 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 15:52:23 pkhuong: Right. 15:52:25 p_l: Well I dunno. I'd think it'd be nice for cellphones :) 15:52:29 ineiros_ [n=ineiros@james.ics.hut.fi] has joined #lisp 15:52:38 p_l: I sure don't use it :) 15:52:39 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 15:52:40 pbusser3: Yes, though we don't have compiler using similar methods as HotSpot 15:52:42 schme: yeah, then marketing did a s/forth/java/g. 15:52:56 pkhuong: ? 15:53:07 matimago: will break your system 15:53:13 stuart71 [n=user@75-145-221-229-Memphis.hfc.comcastbusiness.net] has joined #lisp 15:53:13 zophy [n=sy@host-242-6-111-24-static.midco.net] has joined #lisp 15:53:13 Edico [n=Edico@unaffiliated/edico] has joined #lisp 15:53:13 holly_ [n=holly@ronaldann.demon.co.uk] has joined #lisp 15:53:13 peddie [n=matthew@67.169.49.37] has joined #lisp 15:53:13 hefner [n=hefner@c-69-140-128-97.hsd1.md.comcast.net] has joined #lisp 15:53:13 ineiros [n=ineiros@james.ics.hut.fi] has joined #lisp 15:53:13 Buganini [n=buganini@security-hole.info] has joined #lisp 15:53:13 phadthai [i=mmondor@ginseng.pulsar-zone.net] has joined #lisp 15:53:13 koollman [n=samson_t@sd-10510.dedibox.fr] has joined #lisp 15:53:13 Orest^bnc [n=Orest@hates-script-kiddies.ath.cx] has joined #lisp 15:53:13 moesenle [n=moesenle@atradig124.informatik.tu-muenchen.de] has joined #lisp 15:53:13 Bootvis [i=bob@baltar.lan.endoria.net] has joined #lisp 15:53:13 mbishop [n=martin@unaffiliated/mbishop] has joined #lisp 15:53:13 rbancroft [n=rumble@S01060014bf54b5eb.cg.shawcable.net] has joined #lisp 15:53:13 specbot [n=specbot@common-lisp.net] has joined #lisp 15:53:13 michaelw [i=michaelw@lambda.foldr.org] has joined #lisp 15:53:15 Yes! I got it! Then the C compiler does very little. It's the programmer that has to write efficient code. In the case of HLLs, the compiler has to do all the weightlifting (and is therefore harder to write). 15:53:15 schme: those JVM chips pretty much all started out as forth systems. 15:53:37 pkhuong: I see. 15:53:38 Bootvis_ [i=bob@baltar.lan.endoria.net] has joined #lisp 15:53:38 p_l: anyways, I don't have much more than 100 processes, so once a sec should be enough processing time for them :-) 15:53:56 Buganini_ [n=buganini@security-hole.info] has joined #lisp 15:53:58 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 15:54:02 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 15:54:03 if you want to put optimize a C compiler, it's much harder than optimizing a HLL compiler 15:54:05 peddie_ [n=matthew@67.169.49.37] has joined #lisp 15:54:08 kooll [n=samson_t@sd-10510.dedibox.fr] has joined #lisp 15:54:10 well I should write some more code instead of IRC. 15:54:15 rbancrof1 [n=rumble@S01060014bf54b5eb.cg.shawcable.net] has joined #lisp 15:54:15 pkhuong : you mean back in the days of the embedded pre-java ( which I forgot the name ) 15:54:18 ? 15:54:19 matimago: 1000 Hz was placed because of time to respond to "interactive" interrupts. And interrupt code is still not completely wrinkle-free with regards to HZ constant 15:54:32 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 15:54:33 matimago: The 1msec timer was introduced to improve stuff like audio processing and interactive things. 15:54:36 schme: I see. 15:54:45 schme: yeah, I've heard of those Lisp machines. 15:54:56 machines that directly ran Lisp. 15:55:05 matimago: when they finish, you'll probably see much less than 1000Hz even if it will be set at 1000Hz :P 15:55:06 Yep. It depends on the kind of processes you have. 15:55:08 Almost directly I bet. 15:55:09 Orest814 [n=Orest@hates-script-kiddies.ath.cx] has joined #lisp 15:55:32 pbusser3: there was microcode 15:55:32 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 15:55:43 -!- Orest814 is now known as Orest 15:55:49 "lambda the ultimate opcode" was about designing a scheme cpu, awesome paper 15:55:54 sellout, pkhuong: Thanks. I understand finally :) 15:55:55 The classical LispM architectures didn't run Lisp directly, they ran microcode directly and the microcode interpreted lisp "macrocode", which these days we'd call "bytecode". 15:56:06 densem: hm. Did it ever take off? 15:56:11 nyef: Right. 15:56:28 nyef: ah I see. 15:56:44 nyef: I'm not sure it would really count as bytecode. You'd have to categorise PDP-10 (at least KS-10) as bytecode machines :D 15:57:06 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 15:57:07 I don't remember that ( annoying memory of mine ) 15:57:08 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 15:57:22 Thanks for everything, all. I should be off now. 15:57:24 *artagnon* waves 15:57:25 -!- artagnon [n=artagnon@unaffiliated/artagnon] has left #lisp 15:57:29 *pbusser3* squares 15:57:40 -!- jlf` [n=user@unaffiliated/jlf] has quit [Read error: 110 (Connection timed out)] 15:57:53 Meh. Whichever. 15:57:57 Is there a real difference between bytecodes and native opcodes? I can't think of any. 15:58:04 -!- moesenle [n=moesenle@atradig124.informatik.tu-muenchen.de] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- hefner [n=hefner@c-69-140-128-97.hsd1.md.comcast.net] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- stuart71 [n=user@75-145-221-229-Memphis.hfc.comcastbusiness.net] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- rbancroft [n=rumble@S01060014bf54b5eb.cg.shawcable.net] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- holly_ [n=holly@ronaldann.demon.co.uk] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- michaelw [i=michaelw@lambda.foldr.org] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- specbot [n=specbot@common-lisp.net] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- Bootvis [i=bob@baltar.lan.endoria.net] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- Orest^bnc [n=Orest@hates-script-kiddies.ath.cx] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- ineiros [n=ineiros@james.ics.hut.fi] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- mbishop [n=martin@unaffiliated/mbishop] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- koollman [n=samson_t@sd-10510.dedibox.fr] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- zophy [n=sy@host-242-6-111-24-static.midco.net] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- Buganini [n=buganini@security-hole.info] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- phadthai [i=mmondor@ginseng.pulsar-zone.net] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- peddie [n=matthew@67.169.49.37] has quit [simmons.freenode.net irc.freenode.net] 15:58:04 -!- Edico [n=Edico@unaffiliated/edico] has quit [simmons.freenode.net irc.freenode.net] 15:58:18 bytecodes are designed for an abstract machine 15:58:21 It would be an interesting exercise to write byte code for a modern CPU in order to support the execution of Lisp stuff. 15:58:22 beach: Screwy timing profile is about it. 15:58:30 -!- dialtone [n=dialtone@unaffiliated/dialtone] has quit [Success] 15:58:32 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 15:58:38 pbusser3: check clisp sources ;-) 15:58:41 that can differ a lot from the physical cpu 15:58:49 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 15:58:52 After all, modern CPUs accept micro code uploads for patching CPU bugs. 15:58:59 densem: So what happens if someone builds a real processor to execute some existing bytecodes? 15:59:07 matimago: What is there to check in the CLISP sources? 15:59:10 well, yes, it goes full speed 15:59:15 -!- sohail [n=Sohail@unaffiliated/sohail] has quit [Remote closed the connection] 15:59:16 pbusser3: the VM. 15:59:21 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 15:59:29 densem: So then we have bytecodes that are also native opcodes. 15:59:29 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 15:59:37 pbusser3: it's a bytecode VM. You could try to implement it in Intel microcode... 15:59:42 but that may be impossible to do so, like the scheme cpu, they couldnt make a real hardware garbage collector 15:59:49 matimago: won't fit 15:59:49 matimago: Ah I see. 15:59:59 p_l: I guessed so :-) 16:00:12 pbusser3 : is that true ? cpu are patchable ??? 16:00:15 densem: At the same time, some processors interpret part of the instruction set. Are they then bytecodes? 16:00:24 haha right 16:00:28 densem: I am not sure, but I heard that they are yes. 16:00:29 densem: they were since *long* time. 16:00:36 there are so many subtle layers 16:00:50 densem, to a point 16:00:51 Fare, memo from fe[nl]ix: http://twistedmatrix.com/trac/ticket/2967 16:00:56 pardon my ignorance there, I thought they were really .. static piece of hardware 16:01:02 I am willing to bet we cannot find a significant distinction between bytecodes and native opcodes. 16:01:08 densem: actually first more "complex" cpus were microcoded cpus built from smaller, less potent units 16:01:13 beach: No, those are called micro-code or nano-code (depending on where the interpretation takes place). 16:01:17 ignas [n=ignas@ctv-79-132-160-221.vinita.lt] has joined #lisp 16:01:17 stuart71 [n=user@75-145-221-229-Memphis.hfc.comcastbusiness.net] has joined #lisp 16:01:17 zophy [n=sy@host-242-6-111-24-static.midco.net] has joined #lisp 16:01:17 Edico [n=Edico@unaffiliated/edico] has joined #lisp 16:01:17 holly_ [n=holly@ronaldann.demon.co.uk] has joined #lisp 16:01:17 hefner [n=hefner@c-69-140-128-97.hsd1.md.comcast.net] has joined #lisp 16:01:17 phadthai [i=mmondor@ginseng.pulsar-zone.net] has joined #lisp 16:01:17 moesenle [n=moesenle@atradig124.informatik.tu-muenchen.de] has joined #lisp 16:01:17 mbishop [n=martin@unaffiliated/mbishop] has joined #lisp 16:01:17 specbot [n=specbot@common-lisp.net] has joined #lisp 16:01:17 michaelw [i=michaelw@lambda.foldr.org] has joined #lisp 16:01:21 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:01:38 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:01:43 sohail [n=Sohail@d207-81-121-15.bchsia.telus.net] has joined #lisp 16:01:56 beach: Micro code usually has a direct control over various parts of the CPU. That is something machine code doesn't even have. 16:01:57 plus the big media articles about fpu bugs .. made me think that way 16:02:29 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:02:35 densem: current x86/x86-64 designs implement only some of the simpler instructions in silicon (and even those go through a rather big change), with some of the complex ones microcoded 16:02:36 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:02:36 pbusser3: It is not very confortable to a scientist to have call the same codes bytecodes if they run on one machine, native opcodes if run on a different one, and microcode if run on a third one. 16:03:16 pbusser3: because the names suggest something intrinsic to the codes themselves, but in fact it depends on the machine on which they are executed. 16:03:22 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:03:23 Please excuse my ignorance, I'm not a CS guy. Does the general design of processors facilitate the use of C over lisp? Would the benefit from having lisp specific hardware justify the expense? (Probably not, because if it did, there still would be lisp specific hardware) Does it even matter anymore? Can hardware be designed in a language neutral manner? 16:03:27 -!- aerique [i=euqirea@xs2.xs4all.nl] has quit ["..."] 16:03:28 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:03:34 beach : vocabulary is such a unbeatable monster :D 16:03:39 densem: iirc, P6 (also known as Pentium Pro/II/III/M and possibly Core 1/2) used 112-bit long instructions 16:03:42 beach: Well, you can say the same about a bike, a car, and a truck. All three have wheels. But they are still different beasts which are not easily mistaken for one another. 16:03:53 tmh: yes, no, a bit, no 16:04:14 densem: Except that it is worth trying to get it right, because otherwise we will argue over things we have different definitions of. 16:04:22 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:04:34 *tmh* nods to Fare 16:04:42 beach : i'm totally with you on that subject 16:04:43 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:04:47 http://img.thedailywtf.com/images/200907/errord/aberdeenbus.jpg <--- pity it wasn't ( and ) 16:04:51 pbusser3: A bike doesn't turn into a car if driven on a highway. 16:04:54 beach : except that the world seems to disagree ^^ 16:05:00 densem: cpu microcode may not be able to change the whole instruction decoding process -- it may only allow to somehow modify complex instructions already marked as such in the decoding process. 16:05:22 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:05:48 tmh: Well, C can almost be translated 1:1 to machine code. It is almost as if you substitute C with assembly code. That doesn't take a lot of intelligence nor processing power. 16:05:48 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:06:01 usually microcode looks like list of input registers, list of execution units etc. and list of outputs 16:06:07 tmh: Lisp requires a lot more intelligence to be implemented efficiently. 16:06:09 Xach: you wrote Vecto, am I right? 16:06:13 -!- ASau [n=user@193.138.70.52] has quit [No route to host] 16:06:14 sometimes also settings for some of the units 16:06:15 pbusser3: However, if you did that, the resulting code wouldn't be very fast. 16:06:16 stuart71`: yeah 16:06:20 C seems to be assembly with ~symbolic~ names 16:06:22 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:06:29 df_aldur [n=df@81.171.44.83] has joined #lisp 16:07:07 C is assembly without a good memory model, without the ability to tail call, without control on your temporary variables. 16:07:21 Xach: This is a stupid question, I'm sure, but I would like to write the file out to a string in memory instead of a proper file. Can you tell me how to do that? 16:07:21 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:07:27 Right. Micro code has a bit to enable the add register. Another bit enables the shifter. Another enables a data fetcher from the data bus. Etc. 16:07:29 without the ability to return multiple values 16:07:41 -!- df_aldur [n=df@81.171.44.83] has left #lisp 16:07:55 pbusser3: sometimes it's much crazier, like Transmeta's cpus 16:07:59 yeah, the "linguistic" features are pretty limited 16:08:08 all the abstraction level of assembly, none of the power 16:08:15 ^^ 16:08:17 Fare: Good one! 16:08:19 stuart71`: it's not really a stupid question, but there is no direct support for it in vecto. it might be easiest with flexi-streams and fiddling with the internals of save-png. 16:08:22 So in theory you write a microcode to enable the data fetcher. Then you enable the adder in the next instruction. That makes that the ALU register is added to the memory value. Then you enable the store result in ALU bit, etc. 16:08:26 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:08:36 p_l: I'm just painting a general picture here. 16:08:55 i think lisp is more close the assembly that it seems, you have as much control as you can get 16:08:59 For a better introduction, read a computer architecture book, like the one written by Tannenbaum for instance. 16:08:59 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:09:11 *pbusser3* bbl 16:09:12 Fare: you can return structs by value. 16:09:22 pbusser3: I've got overload of cpu design literature right now :D 16:09:22 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:09:31 Xach: okay, that sounds good. I'm just playing around with sending images to emacs through swank. Thanks. 16:09:57 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:10:13 stuart71`: it might be easiest to just use temporary files and either send the filename to emacs or load & delete it afterwards. 16:10:15 pkhuong, you can??? since when? 16:10:26 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:10:28 what happens if the struct is large? 16:10:37 Fare: It has always been possible I think. 16:10:39 *Xach* strets his stuff 16:11:10 is it some magic where the caller pre-allocates a fixed-sized struct on stack? 16:11:30 Fare: That is the way it was done in GCC last time I looked. 16:11:31 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:11:33 -!- fvw [n=sdfpme@59.39.243.199] has quit [Remote closed the connection] 16:11:42 OK, I stand corrected on multiple values, then. 16:11:47 Xach: I thought of that, but would that still work with a lisp running remotely? 16:11:55 (though I've never seen it used) 16:11:59 Fare: right. 16:12:02 Fare: But as I recall, there was something wrong with pre-GCC compilers that got this wrong. 16:12:17 Maybe it just had to do with thread safety. 16:12:32 unsafe stack manipulations? 16:12:36 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:12:43 stuart71`: i don't know. that's why the other option is to load it into the CL with something like (prog1 (file-octets file) (delete-file file)) 16:12:43 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:12:45 Fare: It wasn't implemented that way in pre-GCC compilers. 16:12:50 stack frobbing and signals don't play well 16:12:52 Small structs are passed/returned directly in registers according to the ABI, and large ones get a pointer to a pre-allocated auto struct. 16:12:52 charitwo [n=charitwo@wikia/Charitwo] has joined #lisp 16:13:15 is that a C9x addition? 16:13:16 kenyao [n=kenyao@58.62.5.184] has joined #lisp 16:13:22 *beach* laments his bad memory 16:13:23 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:13:28 Fare: no. 16:13:30 Fare: no 16:13:31 whoa 16:13:44 I should have read my K&R more carefully 16:13:45 -!- kenyao [n=kenyao@58.62.5.184] has quit [Client Quit] 16:13:50 Fare: Yep! :) 16:14:38 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:14:39 sykopomp [n=sykopomp@unaffiliated/sykopomp] has joined #lisp 16:14:51 *Fare* is trying filesystem frobbing with PLT Scheme. Some of it is very nice, other stuff is a pain. 16:15:53 would be so much pain just trying to download and package the necessary libraries, in semi-portable CL. 16:16:00 elliotstern [n=chatzill@cpe-67-240-162-85.rochester.res.rr.com] has joined #lisp 16:16:16 isn't the struct passing thing architecture-dependant? 16:16:43 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:16:46 anybody tried that nix-os ? 16:16:47 p_l: the implementation, definitely. That's how it seems to be on x86oids and IA-64 and maybe SPARC IIRC. 16:17:03 pkhuong: cause PDP-11 didn't have many registers, afaik :) 16:17:03 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:17:05 densem, I haven't, but I plan to for my next installation, if ever 16:17:38 df_arctor [n=ben@bspencer.plus.com] has joined #lisp 16:17:44 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:17:47 -!- mathrick [n=mathrick@wireless.its.sdu.dk] has quit [Read error: 110 (Connection timed out)] 16:17:51 Fare : great 16:17:52 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:18:10 -!- df_arctor [n=ben@bspencer.plus.com] has quit [Client Quit] 16:18:30 Xach: I'm sorry, where is file-octets defined? 16:18:46 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:18:53 -!- stuart71 [n=user@75-145-221-229-Memphis.hfc.comcastbusiness.net] has quit [Connection timed out] 16:18:59 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:19:21 -!- benny [n=benny@i577A1ACD.versanet.de] has quit [Success] 16:19:40 nha [n=prefect@137-64.105-92.cust.bluewin.ch] has joined #lisp 16:19:44 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:19:52 df_arctor [n=ben@bspencer.plus.com] has joined #lisp 16:20:33 pkhuong: with only five registers, of which one would be quite possibly the address of called function, I doubt you could fit much :) 16:20:48 So I typed up my problem on the XKCD fora: http://echochamber.me/viewtopic.php?f=11&t=4052&sid=bcc59938561ee79c1b269a821dc6c3de&start=720#p1680653 I won't waste space retyping it up. Basically, I'm running into the problem that "*** - (SETF FUNCALL) is only defined for functions of the form #'symbol.", and I want to do it on a function that's passed in 16:20:49 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:20:56 p_l: not particularly worse than x86. 16:21:06 mbishop_ [n=martin@unaffiliated/mbishop] has joined #lisp 16:21:17 elliotstern: you can't portably setf funcall. 16:21:40 Right 16:21:47 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:22:01 but what should I do to fix my code in a non-ugly way? 16:22:01 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:22:03 (Especially with places that expand into something non-trivial) 16:22:20 You'll have to pass a setter around and/or use a macro. 16:22:26 amnesiac [n=amnesiac@p3m/member/Amnesiac] has joined #lisp 16:22:32 phf [n=phf@host.icnfull.com] has joined #lisp 16:22:48 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:23:07 get-setf-expansion makes it simple enough to generate forms (which can then be wrapped as closures) to get/set the values in a place without evaluating temporary values multiple times. 16:23:07 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:23:56 Ok 16:24:27 Xach pasted "file-octets" at http://paste.lisp.org/display/83693 16:24:29 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:24:34 The form for the setf-ing of a variable of a CLOS defined class will be constant, right? 16:24:38 stuart71`: ^ up there 16:24:49 oops 16:24:52 error already. 16:25:14 Xach annotated #83693 "file-octets corrected" at http://paste.lisp.org/display/83693#1 16:25:14 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:25:23 That is to say, I can get-setf-expansion on the REPL, and copy and paste (more or less) it into a function to do the same thing, right? 16:25:37 elliotstern: the forms only depend on the s-expression (and environment) that defines the place. get-setf-expansion is meant to be called during macro expansion. 16:25:52 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:26:16 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:26:32 Ok. So I have to wrap this into a macro? 16:26:49 alinp [n=alinp@89.137.98.94] has joined #lisp 16:26:52 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:28:27 Does anyone have experience with Michael Parker's REGEX/CLAWK/LEXER packages? Does anyone know Michael Parker? Maybe he lurks? 16:28:27 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:28:55 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:28:58 Xach: I see, I was trying to do something similar with with-output-to-string. Would that not be possible? 16:29:05 -!- Guest47226 [n=lexa_@seonet.ru] has left #lisp 16:29:44 stuart71`: maybe. you will have to keep encodings carefully in mind, in that case. 16:29:50 mathrick [n=mathrick@users177.kollegienet.dk] has joined #lisp 16:30:08 kami- [n=user@unaffiliated/kami-] has joined #lisp 16:30:14 hello 16:30:56 Hun [n=hun@p50993726.dip0.t-ipconnect.de] has joined #lisp 16:31:02 -!- serichse1 is now known as serichsen 16:31:18 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:31:55 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:32:31 Xach: this might take me a while to figure out, but I'm just doing it for my own edification. Thank you. I appreciate your help. 16:32:45 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:33:02 -!- phf [n=phf@host.icnfull.com] has quit ["Leaving..."] 16:33:28 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:33:33 is the sbcl boinkmarks available somewhere? .. the first hits on google yield dead links 16:33:47 -!- stuart71` [n=user@75.145.221.229] has quit ["lunch time"] 16:33:49 bhyde [n=Adium@c-76-19-135-167.hsd1.ma.comcast.net] has joined #lisp 16:33:51 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:33:54 -!- jewel [n=jewel@dsl-247-203-169.telkomadsl.co.za] has quit [Read error: 60 (Operation timed out)] 16:34:06 -!- sykopomp [n=sykopomp@unaffiliated/sykopomp] has quit [Read error: 104 (Connection reset by peer)] 16:35:05 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:35:12 sykopomp [n=sykopomp@unaffiliated/sykopomp] has joined #lisp 16:35:14 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:35:41 stassats [n=stassats@wikipedia/stassats] has joined #lisp 16:35:49 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:36:37 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:37:12 lnostdal: is the new URL 16:37:21 -!- mbishop [n=martin@unaffiliated/mbishop] has quit [Read error: 113 (No route to host)] 16:37:22 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:37:27 thanks, pkhuong 16:38:31 sykopomp` [n=sykopomp@unaffiliated/sykopomp] has joined #lisp 16:38:47 -!- Davse_Bamse [n=davse@4306ds4-soeb.0.fullrate.dk] has quit [Remote closed the connection] 16:38:59 Davse_Bamse [n=davse@4306ds4-soeb.0.fullrate.dk] has joined #lisp 16:40:09 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:44:09 -!- moghar` is now known as moghar 16:44:49 -!- mstevens [n=mstevens@zazen.etla.org] has quit ["leaving"] 16:46:06 -!- df_arctor [n=ben@bspencer.plus.com] has quit ["leaving"] 16:47:03 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 16:47:18 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 54 (Connection reset by peer)] 16:47:53 -!- hawkbill [n=pradyus@117.192.6.120] has quit [Read error: 113 (No route to host)] 16:53:09 -!- sykopomp [n=sykopomp@unaffiliated/sykopomp] has quit [Read error: 110 (Connection timed out)] 16:54:05 Hmm, Michael Parker was last seen on c.l.l in 2006. Michael Weber used CLAWK in earlier 2008 and noted that MP's email address bounced, so MP is missing in action. 16:54:31 -!- sykopomp` is now known as sykopomp 16:58:33 rouslan [n=Rouslan@unaffiliated/rouslan] has joined #lisp 17:00:17 -!- kiuma [n=kiuma@85-18-55-37.ip.fastwebnet.it] has quit ["Bye bye ppl"] 17:04:05 -!- envi^home [n=envi@220.121.234.156] has quit [Read error: 104 (Connection reset by peer)] 17:05:22 -!- acieroid` is now known as acieroid 17:06:49 mrsolo [n=mrsolo@nat/yahoo/x-662147a15b03ca69] has joined #lisp 17:07:48 -!- manuel_ [n=manuel@213.144.1.98] has quit [] 17:08:53 saikat_ [n=saikat@c-98-210-13-214.hsd1.ca.comcast.net] has joined #lisp 17:10:44 -!- bgs100 [n=ian@unaffiliated/bgs100] has quit [] 17:10:51 If someone was kept busy all the time, would they be missing inaction? 17:11:03 -!- Fare [n=Fare@c-98-216-111-110.hsd1.ma.comcast.net] has quit [Read error: 110 (Connection timed out)] 17:12:00 nyef: that's really a philosophical question 17:13:22 nyef: Perhaps, unless they liked to be busy. I found his Facebook page, so he appears to still be *here*, but I think that his CL days are in the past. 17:14:32 -!- segv [n=mb@ita4fw1.itasoftware.com] has quit [Read error: 110 (Connection timed out)] 17:17:21 Jabberwockey [n=jens@port-93801.pppoe.wtnet.de] has joined #lisp 17:17:40 Anyway, what about the META parser, anyone have experience with that? 17:19:19 Could you tell me what can I use to parse HTML website? 17:19:37 mrSpec: Closure HTML is a decent starting place. 17:19:55 -!- charitwo [n=charitwo@wikia/Charitwo] has quit [Read error: 104 (Connection reset by peer)] 17:19:58 spilman [n=spilman@ARennes-552-1-36-239.w92-135.abo.wanadoo.fr] has joined #lisp 17:20:21 *tmh* decided to make an honest effort to use an existing text parser and hopefully avoid one of the rings of Hell. 17:20:32 -!- asksol [n=ask@pat-tdc.opera.com] has quit [Read error: 110 (Connection timed out)] 17:20:38 ok thanks :D 17:21:12 tmh: what are you parsing? And why? 17:21:18 charitwo [n=charitwo@wikia/Charitwo] has joined #lisp 17:21:37 tmh: How does using an existing text parser help you avoid one of the rings of Hell? 17:22:37 nyef: writing parsers seems to be a lost art these days 17:23:01 splittist: parser is probably generous. I have test data in various formats. I've been writing one-off parsers for each set of data and am starting to accumulate more general functions. I pondered creating my own text utilities library, but am also thinking about converting everything to use an existing one. 17:23:35 sounds similar to me, i'm parsing via regexps to throw into sql 17:23:57 nyef: Because if I create yet another library for processing text, I'll probably end up there. 17:24:40 That kindof depends on how annoying you find it to maintain and use said library and whether or not you inflict it upon anyone else. 17:25:03 dialtone [n=dialtone@adsl-99-136-101-166.dsl.pltn13.sbcglobal.net] has joined #lisp 17:25:32 mdavid [n=mdavid@ita4fw1.itasoftware.com] has joined #lisp 17:26:18 parsing or lexing? I sometimes think the general purpose lexer-creators are more obscuring than liberating. 17:26:53 -!- alinp [n=alinp@89.137.98.94] has quit [] 17:27:30 -!- silenius [n=jl@yian-ho03.nir.cronon.net] has quit [] 17:28:18 splittist: parsing. Now I'm thinking about learning how to use some parser generator. I always end up at this phase with every language I use. I guess it was only a matter of time with CL. 17:28:23 Mmm... I've always had better luck with hand-coded recursive-descent things than anything else. 17:29:06 Though, I thing I was just about at the point of writing a macro to deal with some of the matching for my last serious parser. 17:29:13 nyef: That's basically what I have now. But some of the support functions are generic text processing, split a string on a delimiter, look for blank lines, end of file, etc. 17:29:13 parsec FTW 17:30:22 manuel_ [n=manuel@rzdu-ub-141-125.rz.uni-karlsruhe.de] has joined #lisp 17:30:42 -!- schoene [n=ms@cpe-76-188-180-22.neo.res.rr.com] has quit ["Leaving"] 17:31:04 -!- mega1 [n=mega@pool-00839.externet.hu] has quit [Read error: 110 (Connection timed out)] 17:31:16 cl-yacc isn't bad. Good enough for cl-python, which is pretty cool by all accounts. 17:31:17 I'm intrigued by META because the functionality-to-size ratio looks large. 17:31:40 tmh: META is fun, too (: 17:31:48 The problem with META is that it is, effectively, line noise, and was designed to emulate an older system. 17:32:05 -!- ThomasI [n=thomas@unaffiliated/thomasi] has quit ["Bye Bye!"] 17:32:11 Which amounts to an interesting approach, but not something I'd want to use in modern practice. 17:32:45 nyef: Line noise in terms of use {} []], etc? 17:33:08 sledge [n=chris@pdpc/supporter/student/sledge] has joined #lisp 17:34:49 Rob Warnock's post about the original Meta was very interesting. 17:34:53 -!- jamief [n=jamie@158.223.53.104] has quit ["Leaving"] 17:35:55 Xach: On c.l.l? If so, can you point me in the general direction? 17:38:05 let me see. 17:38:39 Just let me know if it is on c.l.l., I'll do the searching. 17:38:59 -!- ace4016 [i=ace4016@cpe-76-168-248-118.socal.res.rr.com] has quit [Nick collision from services.] 17:39:00 17:39:08 ace4016 [i=ace4016@cpe-76-168-248-118.socal.res.rr.com] has joined #lisp 17:39:17 bighouse [n=bighouse@x-132-204-252-252.xtpr.umontreal.ca] has joined #lisp 17:39:18 *Xach* had it saved from gnus 17:39:26 ...from 2006...was it really that long ago? ugh. 17:39:30 ragel is quite interesting, though it lacks CL backend :/ 17:39:41 What is ragel? 17:39:53 pbusser3: Finite State Machine generator 17:40:02 can be used to build parsers as well 17:40:17 I see. 17:40:23 -!- Reav__ [n=Reaver@212.88.117.162] has quit [Read error: 104 (Connection reset by peer)] 17:40:39 ew, ew. recurse, descend, and be happy. 17:40:54 Or packrat. :-) 17:41:05 hefner: So you advocate roll your own RDP as needed? 17:41:25 don't take my advice, I avoid text parsing chores. 17:41:31 for example, mongrel's HTTP support was written using it for the whole HTTP logic and they claim that they have better HTTP protocol compliance than apache :D 17:41:32 DeusExPikachu [n=DeusExPi@pool-173-58-94-2.lsanca.fios.verizon.net] has joined #lisp 17:42:16 p_l: Does HTTP require a finite state machine? 17:42:51 pbusser3: both parsing of the protocol as well as protocol logic can be implemented this way 17:43:03 I see. 17:43:38 fraktil [n=fraktil@pool-96-248-228-113.snloca.dsl-w.verizon.net] has joined #lisp 17:43:46 cl-yacc is fabulous, in particular if you get a correct BNF handed to you, which you just need to translate to Lisp syntax then. 17:43:57 Its error reporting though... 17:44:03 Heh! 17:44:09 pbusser3: so you can have state machine parsing requests and state machine that implements the protocol itself as state transitions 17:45:07 -!- ruediger [n=ruediger@91-115-20-76.adsl.highway.telekom.at] has quit [Read error: 110 (Connection timed out)] 17:45:47 blandest [n=user@79.112.98.100] has joined #lisp 17:46:04 ruediger [n=ruediger@91-115-177-141.adsl.highway.telekom.at] has joined #lisp 17:46:52 levy [n=levy@apn-94-44-11-166.vodafone.hu] has joined #lisp 17:47:14 sepult [n=user@xdsl-87-78-24-88.netcologne.de] has joined #lisp 17:47:19 üdv 17:47:20 levy, memo from pkhuong: connectivity isn't very hot these days. Your best bet is to send me an email (pvk@pvk.ca) if I'm not immediately available. 17:50:15 -!- fraktil [n=fraktil@pool-96-248-228-113.snloca.dsl-w.verizon.net] has left #lisp 17:51:29 *p_l* wonders if SSE intrinsics could be used slightly unconventionally on x86-64 17:51:37 -!- davazp [n=user@56.Red-79-153-148.dynamicIP.rima-tde.net] has quit [Remote closed the connection] 17:54:05 -!- gko [n=gko@122-116-15-138.HINET-IP.hinet.net] has quit [Read error: 113 (No route to host)] 17:54:18 tmh: was that sufficient? 17:54:59 Xach: Don't know, couldn't find it using Google and am currently waiting for my news client to download older articles. 17:55:17 2006... 17:55:22 Currently 3545 of 10984 17:55:55 http://groups.google.com/group/comp.lang.lisp/msg/9e0caf0c11fa2729 17:57:05 Xach: Much obliged. I tried to get by with Google groups until very recently. Then, I only downloaded recent articles to my news client. 17:57:48 google has a message-id search 17:58:20 Xach: Ah, missed that. Anway, at least this motivated me to finally download the articles. 17:59:26 -!- sepult [n=user@xdsl-87-78-24-88.netcologne.de] has quit ["ERC Version 5.3 (IRC client for Emacs)"] 17:59:43 -!- jsoft [n=user@unaffiliated/jsoft] has quit [Remote closed the connection] 18:01:46 trebor_home [n=user@dslb-084-059-005-168.pools.arcor-ip.net] has joined #lisp 18:02:54 -!- kidd [n=kidd@119.Red-79-152-42.dynamicIP.rima-tde.net] has quit [Read error: 104 (Connection reset by peer)] 18:02:59 asksol [n=ask@250.246.16.62.customer.cdi.no] has joined #lisp 18:03:20 -!- blandest [n=user@79.112.98.100] has quit [Read error: 104 (Connection reset by peer)] 18:03:29 "That's what Peter Siebel's interpretation of META is doing." Is interpretation in PCL? 18:03:47 -!- elliotstern [n=chatzill@cpe-67-240-162-85.rochester.res.rr.com] has quit [Read error: 113 (No route to host)] 18:04:08 jewel [n=jewel@dsl-247-203-169.telkomadsl.co.za] has joined #lisp 18:05:19 -!- koning_robot [n=aap@88.159.107.70] has quit [Read error: 104 (Connection reset by peer)] 18:07:02 kidd [n=kidd@119.Red-79-152-42.dynamicIP.rima-tde.net] has joined #lisp 18:07:18 -!- ace4016 [i=ace4016@cpe-76-168-248-118.socal.res.rr.com] has quit [Nick collision from services.] 18:07:26 _ace4016_ [i=ace4016@cpe-76-168-248-118.socal.res.rr.com] has joined #lisp 18:08:21 -!- _ace4016_ is now known as ace4016 18:08:58 koning_robot [n=aap@88.159.107.70] has joined #lisp 18:09:49 p_l: What do you mean with unconventionally? 18:10:06 elliotstern [n=chatzill@129.21.82.198] has joined #lisp 18:12:09 -!- rouslan [n=Rouslan@unaffiliated/rouslan] has quit [Read error: 104 (Connection reset by peer)] 18:12:29 Demosthenex pasted "Macro Help" at http://paste.lisp.org/display/83696 18:12:56 blandest [n=user@79.112.98.100] has joined #lisp 18:12:59 so, i'm having difficulty using a macro to create an object using a list for its parameters 18:13:58 Demosthenex annotated #83696 "Output" at http://paste.lisp.org/display/83696#1 18:14:30 <_3b> you are using a symbol as parameter to the macro, not a list 18:15:29 <_3b> `(apply 'make-instance 'test ,args) 18:15:42 segv [n=mb@ita4fw1.itasoftware.com] has joined #lisp 18:15:43 there's apply again.... 18:16:16 Demosthenex: any time you have a list of parameters that you wish to use as function args, you're going to want APPLY. 18:16:22 right, so apply is making it evaluate the argument? 18:16:31 <_3b> no, being evaluated it 18:16:46 <_3b> apply is making it pass the elements of the list as individual arguments 18:16:58 -!- manuel_ [n=manuel@rzdu-ub-141-125.rz.uni-karlsruhe.de] has quit [] 18:17:01 right. 18:17:07 somehow i got that confused with ,@ 18:17:11 the problem is you're trying to splice a list in at compile time, but passing your macro a symbol 18:17:13 thanks for the input ;] 18:17:15 <_3b> ,@ would work on an actual list 18:17:16 -!- mattrepl [n=mattrepl@76.104.2.182] has quit [] 18:17:20 not a symbol. 18:18:16 Demosthenex: remember that a macro does not evaluate the arguments it's given, so ,@ is operating on the symbol *obj-args* 18:18:24 which is an error, of course 18:18:36 thats getting clearer ;] 18:19:03 you can't splice in the list of args at compile time, bacause you won't be evaluating the variable which holds the list of args until run-time 18:19:29 dnm_ [n=dnm@c-68-49-47-248.hsd1.va.comcast.net] has joined #lisp 18:19:37 Fare [n=Fare@guest-fw.dc4.itasoftware.com] has joined #lisp 18:19:37 fe[nl]ix [n=algidus@88-149-210-67.dynamic.ngi.it] has joined #lisp 18:20:33 -!- Jabberwockey [n=jens@port-93801.pppoe.wtnet.de] has quit [Remote closed the connection] 18:20:58 Demosthenex: contrast with `(make-instance 'foo ,@*obj-args*), for example. 18:21:06 Soulman [n=kvirc@154.80-202-254.nextgentel.com] has joined #lisp 18:22:08 hello 18:23:00 fe[nl]ix, hi 18:23:08 hi fe[nl]ix 18:23:18 hi Fare, densem 18:23:25 Fare: read it ? 18:23:37 I think that if you design the library carefully, you can use waitpid(-1), and provide a decent API for other process-forkers to use 18:23:40 sure 18:24:04 Fare: that bug report covers nicely the current brokenness :( 18:24:07 drewc: yeah, that had been failing on me complaining of a symbol 18:24:20 fe[nl]ix, it's not so much "brokenness" as "unmodularity" 18:24:57 the unix API is not modular -- you may provide a modular API on top of it, but you can't use the unix API in a modular way. 18:25:06 I don't know what you're referring to, but it's basically always broken to use waitpid(-1). 18:25:14 foom: why so? 18:25:30 because someone's going to call popen. 18:25:43 -!- Davse_Bamse [n=davse@4306ds4-soeb.0.fullrate.dk] has quit ["Lost terminal"] 18:25:53 what if you control a faithful table of all the forked processes, including things forked by any library? 18:25:59 we refer to this: http://twistedmatrix.com/trac/ticket/2967 18:26:26 Fare: how do you do that ? 18:26:27 how can you control a faithful table of all forked processes including those forked by any library? 18:26:29 foom: how does that one person intend to wait for the result? 18:26:50 Fare: pclose 18:27:16 salex [n=user@216.80.147.206] has joined #lisp 18:27:28 'afternoon 18:27:41 hi salex 18:27:50 Davse_Bamse [n=davse@4306ds4-soeb.0.fullrate.dk] has joined #lisp 18:27:52 hum. You might redirect popen and pclose with some LD_LIBRARY_PATH magic. 18:28:06 anyone know what the timeline for next sbcl release is? I have to roll out to a bunch of machines .... want to know if it's best to hold off a few days 18:28:10 but OK, you have to cope with libraries that won't play fair and square. 18:28:29 (or at least, you have to provide a way out in this case) 18:28:47 What does "playing fair and square" even mean. There's *no way* to write library code, since as you said unix doesn't provide a modular way to wait for processes. 18:30:35 -!- splittist [n=dmurray@63-55.5-85.cust.bluewin.ch] has quit ["you can't win; you can't quit"] 18:31:10 salex: I have to commit one more bugfix 18:32:31 Fare: could you do a little investigating wrt. the use of SIGCHLD in the various implementations(or lack thereof) ? 18:33:47 roygbiv [n=blank@pdpc/supporter/active/roygbiv] has joined #lisp 18:34:16 pkhuong: ah, ok. I just wanted to avoid doubling up work. thanks. 18:36:04 grouzen [n=grouzen@91.214.124.2] has joined #lisp 18:36:35 slyrus_ [n=slyrus@dsl092-019-253.sfo1.dsl.speakeasy.net] has joined #lisp 18:36:43 Odin- [n=sbkhh@s121-302.gardur.hi.is] has joined #lisp 18:38:25 pitui [n=pitui@135.207.174.197] has joined #lisp 18:38:30 fe[nl]ix, in SBCL, it's ignored by default, until you run "run-program" that installs its own SIGCHLD handler. 18:38:40 other than the video on slime, has anyone seen a good tutorial on using the debugger in sbcl or slime? 18:38:42 I've been bitten by that in the past. 18:38:44 bgs100 [n=ian@unaffiliated/bgs100] has joined #lisp 18:40:03 foom: if someone provided a modular API that became standard, things could change. At least, all your pure-lisp libraries sitting on top of that could be modular. 18:40:29 and if you wanted to mix lisp and non-lisp libraries, you'd have them live in separate processes 18:41:07 (that both wanted to run subprocesses) 18:42:00 xinming_ [n=hyy@125.109.246.66] has joined #lisp 18:42:20 ultimately, if your API is that good, it could make it to the libc and/or kernel in a few years. 18:42:26 Fare: there's another(but ugly) option: use a daemon subprocess for run-program 18:42:29 Fare: dream on :D 18:43:14 fe[nl]ix, that's a special case of what I said with "put them in separate processes" 18:43:31 -!- spilman [n=spilman@ARennes-552-1-36-239.w92-135.abo.wanadoo.fr] has quit ["Quitte"] 18:43:36 sure 18:43:54 (and that was also our solution with POIU -- not use run-program until POIU is done with its own forking and waiting) 18:44:17 (although here, it was same process, different phases) 18:45:25 foom: "playing fair and square" would basically mean using the new API. 18:45:36 or something implemented on top of it 18:46:02 whatever that API be -- I suspect it will be in terms of an event loop and callbacks. 18:46:27 -!- moghar [n=user@unaffiliated/moghar] has quit [Read error: 113 (No route to host)] 18:48:10 moghar [n=user@157.185.jawnet.pl] has joined #lisp 18:52:08 so you'd register a callback for when your process has been reaped. 18:52:30 to be called by whoever does the waitpid and catches your process 18:52:50 Demosthenex: the slime manual is excellent for explaining the slime debugger. I also like M-x describe-mode. Make sure your ramped up in the (declare (optimize (debug ...))) department as well. 18:52:54 Fare: a half-solution might be to convince implementors to add a *sigchld-hook*, so that at least our run-program be able to cooperate with the host's 18:53:01 (optionally, you'd specify whether or not your callback is signal-safe) 18:53:28 drewc: i'm reading in it atm 18:53:31 fe[nl]ix, the full solution will be to KILL KILL KILL the host run-program and replace it with yours. 18:53:37 especially on SBCL and CMUCL 18:53:38 Demosthenex: and build slime manual yourself, the one at the website isn't current 18:53:48 Fare: no way 18:53:53 why not? 18:54:03 -!- syamajala [n=syamajal@140.232.182.22] has quit ["Leaving..."] 18:54:18 Speaking of the SLIME debugger, the other day I was asked by a student how to get the value of a local variable, and I was unable to answer. 18:54:25 well updating my environment is another hurdle... i was going to update sbcl to latest and slime/swank, but it blew up, and i had to get back to coding so i reverted 18:54:38 Demosthenex: you using clbuild? 18:54:43 I would then have to replicate all idiosyncrasies of the host's run-program in order no to break code that relies on them 18:54:49 beach: it usually lists locals for each frame 18:54:49 note that hooks are problematic: how do you safely remove the hook? Your hook-list might grow infinite as you spawn things that exit 18:54:56 such as (at least) asdf and cffi-grovel 18:54:58 (you need to RET on the frame0 18:54:59 ) 18:55:20 what you want is have some hash-table, balanced tree, etc., in which you register pids atomically. 18:55:22 ... and hope they haven't been elided by (speed 3) or something 18:55:27 each with a callback 18:55:39 stassats: "usually", really? I find it often doesn't 18:55:56 that gets called after the process is atomically reaped and its callback unregistered 18:56:10 Fare: no. I'd add a single function to that hook, one that reaped all PIDs spawned by our run-program 18:56:11 beach: if it doesn't, recompile with higher debug settings --- C-u C-c C-c on the frame 18:56:13 stassats: Even though I have (debug 3) on all the time. 18:56:13 ASau [n=user@193.138.70.52] has joined #lisp 18:56:40 drewc: no. just trying to compile the latest and make a local (ie: nonroot) installation... 18:56:44 i got bit by some lib issues 18:56:56 fe[nl]ix, OK, so only one hook per library. Sucks a bit, but not too much. 18:57:15 Demosthenex: i suggest using clbuild, it makes updating the world slightly less painful. 18:57:51 Demosthenex: one of the nice things about clbuild is that it creates a local, non-root installation of SBCL :) 18:58:18 -!- Buganini_ is now known as Buganini 18:58:19 fe[nl]ix, although I can foresee atomicity issues with such a hook. 18:58:32 Fare: what issues ? 18:58:45 -!- yango [n=yango@unaffiliated/yango] has quit [Remote closed the connection] 18:58:46 -!- legumbre [n=user@r190-135-3-234.dialup.adsl.anteldata.net.uy] has quit [Read error: 104 (Connection reset by peer)] 18:59:00 yango [n=yango@unaffiliated/yango] has joined #lisp 18:59:06 Guys, I've compiled a library using `clbuild install foo`. Apparently it is in ~/clbuild/source/foo now. How can I use it using sbcl? Linking the folder to ~/.sbcl/site/? 18:59:28 -!- xinming [n=hyy@125.109.246.206] has quit [Read error: 110 (Connection timed out)] 18:59:37 fe[nl]ix, basically, all the handlers that register into the hook better be signal-safe and/or return quick. Because you can't afford anyone to spawn while there are more registered callbacks to call. 19:00:37 so if you don't want to put the atomic PID-management inside your API, you'll have to require very strong constraints from your API users. 19:00:47 konr: There's some option to clbuild to start sbcl with access to all of the projects installed via itself. 19:00:49 which is OK, I suppose. 19:01:01 legumbre [n=user@r190-135-16-107.dialup.adsl.anteldata.net.uy] has joined #lisp 19:02:17 -!- fisxoj [n=fisxoj@149.43.252.13] has quit [Remote closed the connection] 19:02:51 in any case, either below or above your API, someone WILL have to do the hard work of managing a PID table with atomic registration/callback. 19:03:27 (which every run-program implementation does in its own way) 19:04:28 drewc: i'll go look at it! ;] 19:04:35 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 19:04:54 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 104 (Connection reset by peer)] 19:05:07 (and atomic across both threads and signals, at that) 19:06:01 (plus all the wonders of acquiring locks in the parent and cleaning state in the child if you're using fork w/o a prompt exec) 19:07:33 (just because it's so hard to do right, I'd suggest that whatever does handle this pid table be made a library too, whether layered on top of a lower-level one or not) 19:10:44 SandGorgon [n=OmNomNom@122.163.212.113] has joined #lisp 19:14:41 francogrex [n=franco@91.177.16.59] has joined #lisp 19:15:52 -!- sykopomp [n=sykopomp@unaffiliated/sykopomp] has quit [Connection timed out] 19:15:59 -!- francogrex [n=franco@91.177.16.59] has left #lisp 19:17:52 -!- bhyde [n=Adium@c-76-19-135-167.hsd1.ma.comcast.net] has quit ["Leaving."] 19:22:50 how does one get the function object for '(setf slot-value-using-class) 19:22:57 i mean: ? 19:23:18 #'(setf slot-value-using-class), perhaps? 19:23:25 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 19:23:45 or (function (setf slot-value-using-class)) 19:23:47 ah, fdefinition works 19:23:51 just wrong package 19:24:14 would anyone happen to know how best to retrieve the first weekday and week numbering info for a given locale? 19:24:35 ie. what nl_langinfo (_NL_TIME_FIRST_WEEKDAY) does in POSIX 19:24:41 the usual answer to that is to use ICU. 19:25:02 foom: tell me more, do we have a standard way to read that? 19:25:34 ICU is a C library...I don't know of a already-made lisp binding to it 19:26:09 CURLICUE would be a good name for such a library 19:26:11 foom: yeah, I know ICU in C, I was asking about ready-made :) 19:26:12 Urfin [i=foobar@85.65.93.74.dynamic.barak-online.net] has joined #lisp 19:26:15 Xach: hah 19:26:42 I'll probably go cheap and just nick the weekday info and ignore the rest 19:26:53 cl-l10n does most of everything else I need 19:29:11 it's really too bad the posix locale APIs are so terrible. 19:29:32 peddie [n=matthew@67.169.49.37] has joined #lisp 19:29:48 foom: and the C locale APIs are incomplete 19:29:49 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 104 (Connection reset by peer)] 19:29:52 C++'s locale class is a little bit better, but not much 19:30:11 and I'd really like to avoid bringing in ICU for my webapps 19:30:19 it's gonna be a bitch to build anyway 19:30:27 -!- peddie_ [n=matthew@67.169.49.37] has quit [Read error: 60 (Operation timed out)] 19:30:41 minion: cl-icu? 19:30:42 cl-icu: cl-icu is a common lisp interface to the ICU library. http://www.cliki.net/cl-icu 19:30:43 build? apt-get install libicu38. :) 19:30:47 serichsen: oooh 19:30:54 benny` [n=benny@i577A2C29.versanet.de] has joined #lisp 19:30:55 foom: I mean the app, not ICU 19:31:08 ah, can't help you there. :) 19:31:39 also I shall await the reign of XCVB eagerly, all the mysterious build failures that weren't there the last time I dumped an image are getting on my nerves 19:32:45 holly__ [n=holly@ronaldann.demon.co.uk] has joined #lisp 19:33:35 -!- benny` is now known as benny 19:34:49 mathrick, you can hasten the coming of the saviour by coding, you know 19:34:58 -!- zophy [n=sy@host-242-6-111-24-static.midco.net] has quit [Read error: 104 (Connection reset by peer)] 19:35:58 Fare: yes, but I also have other code to work on (it's not like I get build failures in vacuum :) 19:37:00 dstatyvka [i=ejabberd@pepelaz.jabber.od.ua] has joined #lisp 19:38:18 phf [n=phf@host.icnfull.com] has joined #lisp 19:38:23 ehu` [i=52aa21ad@gateway/web/freenode/x-7a2ab5e6302be006] has joined #lisp 19:43:23 -!- billstclair [n=billstcl@unaffiliated/billstclair] has quit [Read error: 145 (Connection timed out)] 19:43:37 Samy [n=sbahra@c-68-50-176-174.hsd1.dc.comcast.net] has joined #lisp 19:44:22 billstclair [n=billstcl@unaffiliated/billstclair] has joined #lisp 19:47:07 -!- SandGorgon [n=OmNomNom@122.163.212.113] has quit [Connection timed out] 19:47:25 -!- holly_ [n=holly@ronaldann.demon.co.uk] has quit [Read error: 110 (Connection timed out)] 19:48:14 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 19:48:29 bad_alloc [n=marvin@HSI-KBW-085-216-108-080.hsi.kabelbw.de] has joined #lisp 19:49:40 fe[nl]ix pasted "raw run-program" at http://paste.lisp.org/display/83701 19:49:54 -!- rdd [n=user@c83-250-157-93.bredband.comhem.se] has quit [Success] 19:50:29 Fare: #83701 is a crude prototype I wrote some while ago with the help of nyef 19:51:48 evening, i've just recently started to learn common lisp and now i'm stuck with this function which is supposed to calculate the wallis product. lisp tells me that i has no value, although i've assigned it in the second line. whats goning wrong? 19:51:48 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 104 (Connection reset by peer)] 19:51:49 http://paste.lisp.org/display/83702 19:51:50 odd, watching the slime video, he's demonstrating things like electric parens, and i don't see that :P 19:51:55 maybe i should update slime 19:52:01 why does run-program have to be portable? this is something the implementations should get right themselves. 19:52:40 bad_alloc: too many i's 19:53:06 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 19:53:26 Demosthenex: my program or my sentence? 19:53:27 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 104 (Connection reset by peer)] 19:53:34 -!- Davse_Bamse [n=davse@4306ds4-soeb.0.fullrate.dk] has quit ["leaving"] 19:53:53 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 19:54:02 Alvar [n=alvar@h49n2fls303o885.telia.com] has joined #lisp 19:54:14 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 104 (Connection reset by peer)] 19:54:21 hefner: because it's hard to convince implementors to expose the exact same interface if you want to do anything more than just running a command in a shell 19:54:45 bad_alloc: prg 19:54:52 ah in the for loop? 19:54:58 manuel_ [n=manuel@HSI-KBW-091-089-172-094.hsi2.kabel-badenwuerttemberg.de] has joined #lisp 19:55:26 And, in the possibly unanticipated use-case department, how well will this portable run-program interact with ptrace debugging? 19:55:41 bad_alloc: i dont' get what exactly you'r etrying to do 19:55:48 fe[nl]ix: well, I'll be impressed if you can write such a thing portably and have it actually work reliably. 19:55:48 chris2_ [n=chris@dslb-094-216-084-022.pools.arcor-ip.net] has joined #lisp 19:56:15 hefner: I'm a dreamer :) 19:56:24 nyef: I have no idea 19:56:33 bad_alloc: http://paste.lisp.org/display/83702#1 19:56:41 -!- sbahra [n=sbahra@c-68-50-176-174.hsd1.dc.comcast.net] has quit [Read error: 110 (Connection timed out)] 19:56:59 wedgeV [n=wedge@cm56-176-196.liwest.at] has joined #lisp 19:57:04 -!- chris2_ is now known as chris2 19:57:11 Demosthenex: i'm trying to calculate the wallis product found here: http://en.wikipedia.org/wiki/Wallis_product 19:57:13 and if you're asking implementations to add hooks to support what you're doing, you might as well just fix their run program to add the features you want 19:57:22 I'm wondering about the thread-friendliness of ignore-sigchld... 19:57:24 schoppenhauer [n=christop@unaffiliated/schoppenhauer] has joined #lisp 19:57:28 fe[nl]ix: I essentially had to wrap waitpid() myself in order to get access to all the status bits needed. 19:57:40 Fare: very bad indeed. 19:57:45 the loop is suppodes to 19:57:58 run through a simpler form 19:58:29 unless you're the one thread that has sigchld enables already 19:58:29 it's best to essentially never modify signal handlers except at the start of the program. 19:58:50 -!- saikat_ [n=saikat@c-98-210-13-214.hsd1.ca.comcast.net] has quit [] 19:58:52 bad_alloc: *ponder* i'm poor at reading calculus symbols 19:58:58 oh, inheritance of signal actions across fork/exec is always fun too. :) 19:59:25 foom: too bad that the implementations themselves don't follow that rule 19:59:26 foom: how do you specify the initial signal handling status of new threads? 19:59:34 Demosthenex: i recon the line on which it claculates is correct, just the varaible declaration etc. should be wrong 19:59:52 foom: modifying signal handlers can be OK if you're single-threaded. 20:00:06 Fare: sigaction is process-wide 20:00:07 bad_alloc: are you tring to add successive iterations together? 20:00:10 Fare: it's not per thread at all 20:00:16 useful, ICU's weekday-number-to-name mapping is 0-based, but the initial weekday is 1-based 20:00:31 foom: oh, OK then. 20:00:43 the signal *mask* is per thread, though 20:00:47 Demosthenex: yes, kindof "i = result of algorithm)" 20:00:50 sykopomp [n=sykopomp@unaffiliated/sykopomp] has joined #lisp 20:01:02 -!- blandest [n=user@79.112.98.100] has quit [Read error: 110 (Connection timed out)] 20:01:04 -!- ehu` [i=52aa21ad@gateway/web/freenode/x-7a2ab5e6302be006] has quit [Ping timeout: 180 seconds] 20:01:07 that's what the let construct there is supposed to do 20:01:08 -!- xristos [n=x@dns.suspicious.org] has quit ["Terminated with extreme prejudice - dircproxy 1.2.0"] 20:01:11 bad_alloc: so when i is 10, it should run 10 times and give the product of all iterations? 20:01:36 bad_alloc: let is not assignment 20:01:47 no, iters should determine the number of runs 20:02:06 mathrick: it isn't? would defvar work? 20:02:15 -!- phf [n=phf@host.icnfull.com] has quit ["Leaving..."] 20:02:21 xristos [n=x@dns.suspicious.org] has joined #lisp 20:02:27 Demosthenex: i just stores the value to be used in the next run 20:02:29 bad_alloc: it creates a *new* binding, and you want to update the old 20:02:36 bad_alloc: no, defvar also defines a new variable 20:02:40 how do i do that? 20:02:41 oh, i see 20:02:41 bad_alloc: setf is what you're after 20:02:50 geat thanks! 20:03:14 bad_alloc: also, please note that defvar creates special variables, whereas let creates lexical ones. Usually special variables are not what you want 20:03:55 as long as lexical stores _long_ floating points its fine by me 20:04:02 so unless you know you want defvar, don't use it. Any use of defvar not at the top-level is to be treated with great suspiciousness 20:04:11 bad_alloc: no, you don't understand what they are about 20:04:16 mathrick: let creates bindings and they can be special 20:04:26 stassats: yes, it was a simplification 20:04:31 bad_alloc: http://paste.lisp.org/display/83702#2 20:04:34 bad_alloc: what book are you using as your handbook? 20:04:57 Practical common lisp (since yesterday, i'm being abitious) 20:05:04 by handbook he means "your tome of arcane knowledge" 20:05:26 bad_alloc: ah, that's a good book, though might be a bit steep if you aren't a fairly good programmer already 20:05:31 -!- elliotstern [n=chatzill@129.21.82.198] has quit [Read error: 110 (Connection timed out)] 20:05:54 kmcorbett [n=Keith@c-76-119-215-57.hsd1.ma.comcast.net] has joined #lisp 20:05:57 mathick: c++, python and a bit of java 20:06:08 i hope that will do 20:06:13 Slowbro [n=slowpoke@174-21-46-162.tukw.qwest.net] has joined #lisp 20:06:14 bad_alloc: but still, the bindings let creates don't influence in any way what values you give them. They don't store long floats any more than they store hashtables 20:06:16 bad_alloc: your output is a fraction ;] 20:06:46 bad_alloc: and since you are operating on integers, the default output format will be rationals, which are exact, not floating point 20:07:20 mathrick: i'll need to use some format t stuff? 20:07:31 bad_alloc: eh, you're going too fast 20:07:42 bad_alloc: where are you in PCL? 20:07:58 do any of the free Windows implementations support threading? I've done a bit of looking around and I'm afraid it looks like only the expensive commercial Windows-based Common LISP implementations support threading so far, but I really hope I'm wrong 20:08:06 ccl does support 20:08:20 ECL does too 20:08:26 mathrick: you may be right, chapter one 20:08:34 SBCL almost supported, but the code was never completed and has long since bitrotted. 20:08:35 SBCL does with sufficient input of live kittens 20:08:47 dmiles_afk [n=dmiles@c-71-197-210-170.hsd1.wa.comcast.net] has joined #lisp 20:09:06 Slowbro: ccl should 20:09:10 since i managed to do it in c++ quite soon i figured that lisp may be the same 20:09:11 bad_alloc: way too fast for your own good, wait until at least chapter 4 20:09:18 okay 20:09:30 Yeah I've been using SBCL so far but I think I'll give CCL a try. didn't know about that one. thank you 20:09:35 Slowbro: clisp does too, it's experimental though 20:09:38 well i'll be leaving you now, it's kina late over here. bye 20:09:41 -!- bad_alloc [n=marvin@HSI-KBW-085-216-108-080.hsi.kabelbw.de] has quit ["You shoot yourself in somebody else's foot."] 20:09:46 bad_alloc: you need a lot of groundwork because lisp is very much unlike C++, so what you think you know will hurt you 20:09:48 eh 20:10:02 *nyef* whimpers. 20:10:19 Either this keyboard can't send C-c or it's not the break character. :-( 20:11:04 nyef: xev, assuming you're in X 20:11:15 I'm not in X. 20:11:24 I think he mentioned a ps3 or something earlier 20:11:28 I have -one- VT, and am running from an initrd. 20:11:30 ps2. 20:11:31 Is it really so unlike? Isn't it more the case that when you're learning C++, you're learning C++. And when you're learning Lisp, you're learning programming? ;) 20:11:47 -!- Alvar [n=alvar@h49n2fls303o885.telia.com] has quit [Read error: 110 (Connection timed out)] 20:11:48 tcr: please tell me the last time you used LET in C++ 20:11:55 -!- Elench is now known as JHVH 20:12:20 I did that maybe once in my entire C life 20:12:38 C has let? 20:12:42 yes 20:12:49 {} is a block by default 20:12:52 I like introducing a new scope by {, } 20:13:10 tcr: because you picked it up from lisp I guess :) 20:13:11 The last time I've programmed in C++ has been... like.. never 20:13:35 I've done that. Less so since switching to C99. Handy in macros, though. 20:13:49 serichsen: so if you just go foo; { bar... } baz, you will have bar... in a new scope 20:14:03 also a ; there for aesthetics 20:14:08 hefner: yeah 20:14:30 I wonder if you can have static variable in a new block, and if you can have two such variables with the same name but in different blocks, and what the semantics would be 20:14:35 bhyde [n=Adium@c-66-30-202-56.hsd1.ma.comcast.net] has joined #lisp 20:14:39 but since old compilers tend to be broken, you usually end up with do { ... } while (0); 20:15:16 tcr: sure, C vars are lexicals, you'd just create two with the same name 20:15:43 -!- jewel [n=jewel@dsl-247-203-169.telkomadsl.co.za] has quit [Read error: 113 (No route to host)] 20:15:52 well, it could be specified that the other would refer to the same static space 20:16:04 -!- segv [n=mb@ita4fw1.itasoftware.com] has quit [] 20:17:11 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 20:17:53 mathrick: I wanted to be sarcastic about your comment regarding "because you picked it up from lisp", but thinking about it, that would explain quite nicely why nobody around me seems to have the same habit of introducing scopes with {} for what I consider clearer code :-) 20:18:19 *Maddas* _really_ misses the ability to narrow the scope like that in Python 20:18:29 (Not Python the compiler, unfortunately) 20:19:32 *tcr* is reminded how he's regularly bitten by the non-lisp2y-ness of other languages 20:20:02 what is lisp2? 20:20:04 mathrick pasted "static + scopes in C" at http://paste.lisp.org/display/83705 20:20:09 tcr: ^ 20:20:17 tcr: no, it couldn't 20:20:23 it would make no sense 20:20:33 "static" refers to the extent, not the scope 20:21:00 I was not talking about the scope 20:21:22 yes you were :) 20:21:44 you were confusing the issues, to be exact 20:22:09 -!- manuel_ [n=manuel@HSI-KBW-091-089-172-094.hsi2.kabel-badenwuerttemberg.de] has left #lisp 20:22:11 tcr: I don't actually mind other languages having the same namespace for functions and variables much, but I really do get irritated if types and variables can't share the same name :-) 20:22:36 the semantics of C are perfectly clear and allow for precisely one interpretation, specifying it to do something completely different and contrary to everything that follows from the rest of the language would be terribly dumb 20:23:40 Maddas: I use (list list) all the time, so I do mind 20:23:57 well, not precisely (list list), but you get the idea 20:24:19 Maddas: it's much more common in modern C++, because you often want to control exactly when destructors are invoked. 20:24:19 mathrick: Actually, I guess I would mind too, given that constructors in some languages have the same names as types. 20:24:32 heh, reminds me I really need to spend a bit of time getting back into C sometime..just to understand it again, it's been years. 20:24:59 pkhuong: I didn't think about that use for scopes, that's a good point 20:25:07 TDT: write lisp interpreter in C 20:25:07 pkhuong: Indeed. I just happen to use it in general to avoid "cluttering" the namespace (I find it easier to read that way), and I tend to get puzzled looks more than anything else. 20:25:08 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 104 (Connection reset by peer)] 20:25:08 Maddas: Well, it happens to be if there's a long computation, and I want to break it up into smaller chunks (basically into some kind of ANF); how do you name the return value of abs?, well I'd call it "abs", but I may have to call that function in some later computation, too. 20:25:15 -!- nvoorhies [n=nvoorhie@75.36.204.63] has quit [Read error: 104 (Connection reset by peer)] 20:25:29 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 20:25:45 -!- roygbiv [n=blank@pdpc/supporter/active/roygbiv] has quit [] 20:26:29 mathrick: sorry, I thought you meant that there was something like a "let" keyword, so nevermind 20:26:42 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 104 (Connection reset by peer)] 20:27:04 it's almost like LET, it only misses the actual keyword 20:27:19 Jabberwockey [n=jens@port-93801.pppoe.wtnet.de] has joined #lisp 20:27:34 heh, i just realized that if the partial evaluation of an if's condition does not result in a constant form (due to being side effects or non local exits in it) i can still omit partial evaluating one of the branches if the return value can be determined to be nil or non nil 20:27:43 mathrick: I don't think that you could create closures 20:27:52 mathrick: I was not confusing the issues. 20:28:02 serichsen: no, but that's orthogonal 20:28:26 levy: you might be able to do something even better by performing a local CPS, with nested conditionals. 20:28:59 pkhuong: you scare me shitless sometimes 20:29:03 stassats: That would definitely be a big exercise, really though you bring up a good idea...write my own compiler of sorts, in C, a difficult task but a good learning experience. 20:29:04 (yes, it's a compliment) 20:29:14 mathrick: not entirely -- you can create lexical blocks, but these do not have the same features as in higher level languages 20:29:44 TDT: well, compiler is a big exercise, but lisp interpreter isn't that hard 20:29:50 serichsen: right, but it's mostly related to there being no way to pass blocks around 20:30:17 gcc adds an extension for nested functions, and you can in fact create downwards closures in it 20:30:22 francogrex [n=franco@91.177.16.59] has joined #lisp 20:30:37 just don't try upwards, it's not going to stop you, instead it will explode violently 20:30:50 mathrick: hope you like your stack executable. 20:30:53 pkhuong, you mean with multiple ifs nested into each other... hmm, how would a local (i guess lexical) CPS help on that, i don't see... 20:31:12 pkhuong: anytime! 20:31:34 but IIRC, they do actually tell you it's allright to pass things downwards 20:32:12 -!- tcr [n=tcr@host145.natpool.mwn.de] has quit ["Leaving."] 20:32:15 http://gcc.gnu.org/onlinedocs/gcc/Nested-Functions.html 20:33:33 levy: Optimisation of (if (if ...) ...)-style forms (or short circuiting logic operators) is often done with a switch to and from a two-continuation form. You can do something similar here. Depending on the level of inlining you perform, that can help produce more specialised code. 20:33:35 pkhuong: so yeah, they use trampolines for addressing and they are true closures 20:33:39 stassats: Yeah, I can't say I've really done either that much. An interpreter I can see being easier. A compiler I wish I've done in school before :) 20:33:50 GCC nested functions are nasty. 20:34:33 dunno, "useless" more than anything. This kind of extension is just asking for trouble if you use it and it's not available everywhere you want to go 20:34:40 "everywhere" including g++ 20:35:11 huh. weird, I annotated that wallis produc thing, but it didn't make a note here 20:35:42 salex: it wasn't original sent to #lisp either. 20:35:49 ah, that's prolly why 20:35:59 g++ supports nested functions. :) 20:36:26 "A nested function is a function defined inside another function. (Nested functions are not supported for GNU C++.)" 20:36:59 yeah, you need to use different syntax. :) 20:37:16 that's kind of silly 20:37:27 foom: link? 20:38:51 http://en.wikipedia.org/wiki/C++0x#Lambda_functions_and_expressions 20:39:36 that's C++0x 20:39:39 -!- ikki [n=ikki@201.155.75.146] has quit [Read error: 113 (No route to host)] 20:39:44 awful. I guess they hack this stuff in wherever it fits. 20:39:49 yes, and? 20:39:55 hefner: you just noticed that? 20:40:05 foom: it's not a standard yet, so it's cheating 20:40:15 *hefner* smacks mathrick around a little 20:40:24 http://www.theregister.co.uk/2009/05/07/verity_stob_cplusplus/ - very funny review of some C++0x features :) 20:40:37 I wonder if it's the one I read previously 20:40:42 though surely we should be calling it either C++09 or C++xx at this point? 20:40:50 rsynnott: it's going to be hex 20:40:54 ah 20:40:55 The trampolines result in an executable stack. Which is not good for security. 20:40:56 :) 20:41:04 rsynnott: c++0xa or c++0xb probably 20:41:08 rsynnott: no, they decided to be silly and said "we have up to C++0F" 20:41:15 0xF even 20:41:28 mathrick: I sincerely hope that isn't actually true 20:41:36 it is 20:42:02 rsynnott: http://en.wikipedia.org/wiki/C%2B%2B0x 20:42:04 TDT: the difficulty is all in your head. It's just a tiny bit more indirection. shows how one can derive a simple compiler from an interpreter. for a CL bent on the same ideas (more or less) 20:43:17 rsynnott: wait, wrong link, I can't find the reference now 20:44:14 pkhuong: Yeah, since I haven't done it that's why I'm thinking it's hard. I'll have to read up a bit on how to do it before I try. I should check if there's a class on this all for graduate level here at the UI, but..dunno if there is, if not, still fun to do it on my own. 20:44:33 My latest high has been trying to get cl-mpi working, and some beginning understanding of MPI..this may be next :) 20:44:59 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 20:45:30 levy pasted "partial eval" at http://paste.lisp.org/display/83708 20:45:48 this is my toy :) 20:45:50 UI? UC? I'd be surprised if there weren't a couple very nice implementation (and performance)-oriented PL courses, at least at the UC campus.. 20:46:07 UIUC? 20:47:14 university of illinois @ urbana-champaign. 20:48:10 Apparently there's a "Friendship University of Central Kansas". 20:48:18 University of Iowa. 20:48:29 I work and go to school here, both. 20:48:36 the interesting thing is that it works from the original SBCL sources 20:50:09 "The standard is expected to be ready for final national votes in 2009 -- hopefully yielding C++09 even if the ISO bureaucracy takes some time work through its formalities" <-- wait, so we have national C++ teams now? 20:50:13 TDT: MPI isn't terrible 20:50:25 do we also get to have C++ special olympics and all that? 20:50:58 mathrick: C++ _is_ the special olympics. 20:50:59 iso is an international organization 20:51:40 Harag [n=phil@wbs-196-2-98-168.wbs.co.za] has joined #lisp 20:51:57 -!- Jasko [n=tjasko@c-98-235-105-148.hsd1.pa.comcast.net] has quit [Read error: 110 (Connection timed out)] 20:52:13 rdd [n=user@c83-250-157-93.bredband.comhem.se] has joined #lisp 20:53:03 may have #2 in a moment 20:53:05 Jasko [n=tjasko@c-98-235-105-148.hsd1.pa.comcast.net] has joined #lisp 20:53:13 gack, wrong channel 20:53:42 #9 is better 20:53:48 salex: Nah, it wasn't that bad, I wish the documentation for cl-mpi was better, but I did get it setup, mostly. 20:54:21 saikat_ [n=saikat@c-98-210-13-214.hsd1.ca.comcast.net] has joined #lisp 20:54:29 nyef: There is a Tallaght Institute of Technology here 20:54:39 salex: I'm planning on trying to finish it as much as I can tonight, by having say 4 virtual machines with NFS mounted /home/mpi to a 5th virtual machine. Then get the MPI to work between them. I don't have the machines to set up a real MPI setup yet. 20:54:41 ah, i misread you. I thought you meant MPI in general. I'm not familiar with cl-mpi 20:54:41 sadly, when they got a domain, they went for itt 20:55:12 salex: yeah, I kinda started with cl-mpi, it's pretty cool. I honestly can't find a whole lot I'd use MPI for, so I'm doing it just to learn. 20:55:24 rsynnott: because tit was already taken? 20:56:28 Hmm, heck, maybe it's best I use the MPI setup to learn C/C++ again on, or at least to get a working knowledge of. Kill a few birds with one stone maybe 20:56:45 -!- Jasko [n=tjasko@c-98-235-105-148.hsd1.pa.comcast.net] has quit [Client Quit] 20:56:48 oh, actually they have tit.ie as well, now 20:56:59 if you don't have a use for it, why learn MPI? It's kind of niche-y 20:57:02 lol 20:57:05 Clearly someone needs to start a technology institute in the middle of Louisiana... 20:57:16 -!- Harag [n=phil@wbs-196-2-98-168.wbs.co.za] has left #lisp 20:57:20 aw heck, where do i set my load path in sbcl if i'm moving functions out into a package? 20:57:56 salex: Because I don't know it...I'm weird that way, I picked up lisp not because I had a use for it, or a need...in fact, I didn't expect to use it as much as I do now, but I'm learning it anyways (and now use it a whole lot). 20:58:01 minion: tell Demosthenex about ASDF 20:58:03 Demosthenex: have a look at ASDF: asdf is Another System Definition Facility. http://www.cliki.net/ASDF 20:58:18 asdf:*central-registry* 20:59:29 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 104 (Connection reset by peer)] 20:59:43 TDT: sure, but lisp is general purpose. MPI is only really useful in pretty constrained circumstances 20:59:50 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 21:00:01 TDT: but MPI is really a solution in search of a problem. You know when you need MPI, it's not like go "oh, I wonder what will happen if I use MPI here" 21:00:06 *like you go 21:00:26 you use MPI because you have to, not because you can 21:00:41 indeed 21:01:25 hah, well, I do agree with you that maybe I wouldn't really find a use for it, and I don't intend to spend that much time using it. I figure, something to try out to have a little working knowledge of distributed systems and message passing. 21:01:43 it's not going to be "working" knowledge 21:01:52 who is minion? is he a human? 21:01:58 nope, bot 21:02:33 so are specbot and lisppaste 21:03:02 i see, but i suppose each have his talent? 21:03:15 exactly 21:03:33 mathrick: So, instead of just picking up something just to learn a little about it, what you're saying is find a solid need for something then dive in deep into a subject? 21:03:33 they're quite suitably named, at that 21:03:34 are they in lisp? 21:03:43 on the internets no one knows you are a bot 21:03:46 or C? 21:03:55 in C, are you kidding? 21:04:14 well they're lisp programs? 21:04:23 they are, yes 21:04:35 TDT: well, specifically in things that are sort of specialized like MPI. You can learn the basics of how it works in half an hour, but actually getting a decently parallized algorithm across a cluster will take significant work 21:04:52 (assuming it isn't trivially parallizable) 21:05:13 and there are lots of fiddly details that vary with hardware and problem 21:05:29 salex: Yeah, that I'm not debating and I'm likely not going to go that far into it since that's not my focus. I figure, get it setup, get it "working" to see how it works, maybe listen to the network packets and watch how it communicates, read a bit, and be done with it. 21:06:39 salex: I doubt I'd write much for it, I'm thinking ensemble data mining operations may work though..now that's one use I can see being very helpful. 21:07:08 -!- bighouse [n=bighouse@x-132-204-252-252.xtpr.umontreal.ca] has left #lisp 21:07:48 TDT: yes. MPI is sort of like specialising in marriage court cases. You can have the general idea of that existing, but unless you have a very specific need and a lot of very deep knowledge and experience, "vague idea" is just as good as "some knowledge" 21:08:48 you may be aware that MPI exists and what its general usage is, but you won't get any more than that unless you actually go all the way through and start writing real-world distributed apps 21:09:01 and you won't unless you have a real need and real hardware to back it up 21:09:27 Actually, minion, specbot and the lisppastes aren't lisp programs, they're -a- lisp program. They all live in the same SBCL instance. 21:09:30 It can be fun to try it out. I'm very happy when I can get a simpler, more robust hack with a tiny server and stateless clients. 21:09:38 -!- Fare [n=Fare@guest-fw.dc4.itasoftware.com] has quit [Read error: 110 (Connection timed out)] 21:09:51 nyef: that's almost like bot fetish porn 21:10:15 Yeah, the "need" is the bottleneck in that area actually at the moment. Hardware is easy, given used computers that I have access to. Work offered to let me borrow 4 computers to get started on it, but yeah..the need, that's the hard part on going far in it. 21:10:34 Okay, this is utterly stupid, but... I'm on what looks to be a debian-derived system and it doesn't have ssh host keys. How do I generate said keys? 21:10:44 I always wondered if it was worth using MPI versus hand-rolling something 21:11:03 (I wonder that about pretty much everything, unless there's something scary inside that I don't understand) 21:11:13 nyef: odd, it should've generated them during install and/or startup 21:11:49 I'm a fan of good ol' voodoo "reinstall and try again" myself 21:11:54 The install instructions said "copy them from the other installed image on your system", but I didn't happen to have one. 21:11:56 just remember to purge it 21:12:15 nyef: uh, so it's some kinda special issue ssh server? 21:12:22 nyef: ssh-keygen or something like that, I think...which would copy to your ~/.ssh/, if I'm reading what you're askign right. 21:12:36 TDT: wrong end of ssh keys 21:12:49 it's the keys used to identify the host, not the client 21:12:49 syamajala [n=syamajal@97-95-190-124.dhcp.oxfr.ma.charter.com] has joined #lisp 21:13:04 -!- dstatyvka [i=ejabberd@pepelaz.jabber.od.ua] has quit [Remote closed the connection] 21:13:14 mathrick: No, it just assumed that you had already set up and configured a different distro and wanted to maintain the two in parallel. Utterly wierd. 21:13:27 indeed 21:13:35 mathrick: ah ok 21:13:39 hefner: a few months ago I had the opposite question regarding MPI_Reduce() for an assignment. Why hand-roll something like that? Turns out I should have because the OpenMPI implementation of MPI_Reduce wasn't very good. 21:13:39 I'd classify that as "special issue" 21:13:42 hefner: it depends a bit on what you're doing. 21:13:45 Hrm. This appears to be a 32-bit userland. How odd. 21:14:17 hefner: a lot of thought has been put into it (MPI, MPI-2) by people familiar with large systems 21:14:35 which doesn't mean a) your implementation is good or b) your particular problem fits well 21:14:52 these are good reasons (and anti-reasons), though. 21:14:53 but it at lesat makes it worth looking at before reinventing wheels 21:15:56 and of course, you may need it to plug into existing software 21:16:34 fwiw, most of my parallel stuff done in lisp has been extremely crude handrolled stuff, which worked just fine (on up to, say, 100 node systems) 21:16:45 but that's as much to do with the problems as anything else 21:16:52 Hrm. "R5900 V3.1 FPU V3.0" 21:17:22 -!- mbishop_ is now known as mbishop 21:18:32 nyef: http://unixwiz.net/techtips/openssh.html#hostkeys 21:18:51 I have no idea if -N sets the proper fileperms though 21:18:53 mathrick: Many thanks. 21:19:00 -!- saikat_ [n=saikat@c-98-210-13-214.hsd1.ca.comcast.net] has quit [] 21:19:05 Handrolling has worked well for some of my problem even with peaks at nearly 400 nodes, but we were able to fiddle with the task granularity as we wished. 21:19:19 salex: What do you use parallel systems for, if you don't mind me asking? 21:19:25 right. so with require you have to specify a pathname... 21:19:30 signal processing, mostly 21:19:36 krumholt [n=krumholt@port-92-193-25-29.dynamic.qsc.de] has joined #lisp 21:19:37 nyef: np. Just check if they have the right perms afterwards, nothing is as fun as ssh barfing due to wrong perms somewhere 21:19:54 (images, specifically) 21:19:58 If I can connect, they have the right perms, right? 21:20:33 yes, assuming the keys are readable only by root 21:20:41 I think it'll die and log errors if they aren't 21:21:25 and in the future, pick better debian derivatives :) 21:22:41 pkhuong: fiddling with granularity helps, so does system utilization ... if you've got a cluster you can basically abuse at will, you can do a lot of things that wouldn't be good if you have to play with other users 21:23:14 -!- wedgeV [n=wedge@cm56-176-196.liwest.at] has quit [] 21:24:02 -!- Jabberwockey [n=jens@port-93801.pppoe.wtnet.de] has quit [Remote closed the connection] 21:25:45 alec [n=aberryma@ita4fw1.itasoftware.com] has joined #lisp 21:28:23 -!- konr [n=konrad@201.82.132.33] has quit [Read error: 104 (Connection reset by peer)] 21:28:39 konr [n=konrad@201.82.132.33] has joined #lisp 21:30:20 rlb3 [n=rlb@ng1.cptxoffice.net] has joined #lisp 21:31:21 -!- bombshelter13_ [n=bombshel@toronto-gw.adsl.erx01.mtlcnds.ext.distributel.net] has quit [] 21:31:38 bombshelter13_ [n=bombshel@toronto-gw.adsl.erx01.mtlcnds.ext.distributel.net] has joined #lisp 21:31:58 ikki [n=ikki@201.155.75.146] has joined #lisp 21:32:02 Oh, the order of collect and while in a loop form is significant. 21:32:54 hm, getting there 21:32:57 -!- bombshelter13_ [n=bombshel@toronto-gw.adsl.erx01.mtlcnds.ext.distributel.net] has quit [Client Quit] 21:32:57 levy pasted "Partial eval" at http://paste.lisp.org/display/83712 21:33:15 -!- moghar [n=user@unaffiliated/moghar] has left #lisp 21:33:30 clhs 6.1.1.6 21:33:31 http://www.lispworks.com/reference/HyperSpec/Body/06_aaf.htm 21:33:42 I suppose that should have been clear from that section. 21:33:52 those nasty SB-INT:LIST-OF-LENGTH-AT-LEAST-P calls... argh 21:34:35 ThomasI [n=thomas@unaffiliated/thomasi] has joined #lisp 21:34:57 tmh: that's why ITER is superior 21:35:31 spilman [n=spilman@ARennes-552-1-36-239.w92-135.abo.wanadoo.fr] has joined #lisp 21:35:48 through all the generics of make-instance, initialize-instance, shared-initialize, etc... 21:35:54 levy: that's already better than having to deal with (>= (length ...) ...) 21:35:59 -!- trebor_home [n=user@dslb-084-059-005-168.pools.arcor-ip.net] has left #lisp 21:36:01 mathrick: I'm not familiar with ITER and that behavior actually works to my benefit. 21:36:28 rouslan [n=Rouslan@pool-71-161-78-9.cncdnh.east.myfairpoint.net] has joined #lisp 21:36:34 tmh: ITER is LOOP with lispy syntax 21:36:34 -!- rlb3 [n=rlb@ng1.cptxoffice.net] has left #lisp 21:36:55 pkhuong, actually i don't yet know why on earth did not go down on that one 21:36:56 mathrick: and a half-too-clever implementation. 21:36:57 tmh: http://www.cliki.net/iterate 21:37:05 *tmh* is already there. 21:37:09 pkhuong: dunno, it seems to work well for me 21:37:13 Phoodus [i=foo@ip68-231-38-131.ph.ph.cox.net] has joined #lisp 21:37:17 pkhuong: what's too clever about it? 21:37:26 I should postpone this, again. I need to get some stuff done. But I really should look at it. 21:37:35 you should 21:37:38 The use of a codewalker to be able to extend it with regular macros. 21:38:21 ah that's the reason (make-function-lambda-form 'sb-int:list-of-length-at-least-p) -> nil 21:38:30 hm, got lost in space :) 21:39:14 -!- sellout [n=greg@63.107.91.105] has quit [] 21:40:15 pkhuong: well, yes and no. The fact I can just include forms like (collect foo) wherever in the body is really, really nice. Bumping into sharp edges when macros clash isn't 21:41:05 -!- sykopomp [n=sykopomp@unaffiliated/sykopomp] has quit [Remote closed the connection] 21:41:11 overall it works quite well most of the time, and you get less macro clashes than you'd have to spend time on marking everything you might want to use inside as an ITER extension 21:42:12 nvoorhies [n=nvoorhie@67.100.118.26] has joined #lisp 21:42:35 levy: btw, attempts to document sb-pcl:generate-discrimination-net-internal, if you still have plans to specialise generic functions. 21:42:58 pkhuong, actually i'm doing that already 21:43:06 i mean specializing generic functions 21:43:30 but thanks that might help, because i still have some issues in make-generic-function-lambda-form 21:43:48 :after and :before methods not handled correctly 21:45:15 I'm getting a git error when trying to install some package from clbuild (one I added myself). I need to specify the branch I want to merge with, yet I don't know how to set that. I also wonder wether or not clbuild supports something to do that 21:45:21 birdsbite [n=user@75.110.164.231] has joined #lisp 21:45:34 -!- francogrex [n=franco@91.177.16.59] has quit ["Leaving"] 21:48:41 saikat_ [n=saikat@adsl-71-139-210-181.dsl.snfc21.pacbell.net] has joined #lisp 21:50:18 i got 10% speedup for a make-instance 'standard-object 21:51:01 it's kind of surprising, i thought it would be the same 21:51:47 levy: the standard implementation must take redefinitions into account. There's probably slightly more indirection than in your specialised version. 21:51:54 -!- syamajala [n=syamajal@97-95-190-124.dhcp.oxfr.ma.charter.com] has quit ["Leaving..."] 21:52:12 levy: do you intend to propose your evaluator for inclusion in sbcl (eventually) ? 21:53:50 fe[nl]ix, i'm going to put it up on cl-net or something 21:54:19 pkhuong, i tried my best, but certainly there might be differences i'm not aware of 21:54:53 at least an extra indirection to call the specialized ctor 21:55:48 -!- slash_ [n=Unknown@whgeh0138.cip.uni-regensburg.de] has quit [Client Quit] 21:55:53 fe[nl]ix, BTW, there are not so many SBCL specific stuff in there 21:56:11 actually only the part that produces the lambda-forms 21:56:39 and of course when i'm partial evaling i set up an SBCL specific environment, but that can be done for other implementations too 21:59:21 bakkdoor [n=bakkdoor@xdslce219.osnanet.de] has joined #lisp 21:59:26 my problem might have been related to the changing of the git source. Testing now 21:59:58 -!- nha [n=prefect@137-64.105-92.cust.bluewin.ch] has quit [""mv readme.txt readordie.txt ^_^" -- zordan"] 22:01:18 skimpy [n=skimpy@ip24727.lbinternet.se] has joined #lisp 22:02:32 urgh! when was the last time the scbl garbage collector was updated? 22:05:39 -!- Edico [n=Edico@unaffiliated/edico] has quit ["Leaving"] 22:06:16 mark and sweep instead of stop and copy for 0 generation.. no compacting of higher generations. (heap fragmentation) 22:06:44 and no support for symmetric multiprocessing 22:07:09 -!- ruediger [n=ruediger@91-115-177-141.adsl.highway.telekom.at] has quit ["Leaving"] 22:07:43 yeah, it pretty much sucks. 22:10:47 kmels [n=kmels@190.148.177.22] has joined #lisp 22:10:54 younder: there is an opportunity for you to implement all these features 22:11:19 fame and fortune await. 22:12:11 eh? We have a mark and sweep collector? 22:13:47 well it calls it scavenging 22:14:27 ah, it's late and it's not going to work for with setf svuc tonight... 22:14:34 bb 22:15:33 -!- levy [n=levy@apn-94-44-11-166.vodafone.hu] has quit ["..."] 22:17:32 I've always wanted a M/S GC for older generations. GEN[C]GC is based on running cheney on everything as least as young as a given generation. 22:18:34 There's also a good support for multithreaded mutation and allocation, much better than many language implementations that have a lock on the heap or on the allocator. 22:20:02 -!- bhyde [n=Adium@c-66-30-202-56.hsd1.ma.comcast.net] has quit ["Leaving."] 22:20:11 -!- konr [n=konrad@201.82.132.33] has quit [Remote closed the connection] 22:20:44 konr [n=konrad@201.82.132.33] has joined #lisp 22:20:47 saikat__ [n=saikat@adsl-71-139-210-181.dsl.snfc21.pacbell.net] has joined #lisp 22:27:52 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 104 (Connection reset by peer)] 22:30:08 I liked what I read about Clozure's GC recently, but I am not able to compare pros and cons yet 22:32:00 -!- salex [n=user@216.80.147.206] has quit [Remote closed the connection] 22:32:30 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 22:34:39 -!- saikat_ [n=saikat@adsl-71-139-210-181.dsl.snfc21.pacbell.net] has quit [Read error: 110 (Connection timed out)] 22:34:40 -!- ausente is now known as dalton 22:35:15 sykopomp [n=user@unaffiliated/sykopomp] has joined #lisp 22:35:31 serichsen: it seems remarkably simple while avoiding halving the address space 22:36:07 -!- Pepe_ [n=ppjet@78.113.38.65] has quit [Read error: 110 (Connection timed out)] 22:36:34 -!- abeaumont [n=abeaumon@84.76.48.250] has quit [Read error: 104 (Connection reset by peer)] 22:38:09 -!- chris2 [n=chris@dslb-094-216-084-022.pools.arcor-ip.net] has quit ["Leaving"] 22:41:15 -!- skimpy [n=skimpy@ip24727.lbinternet.se] has quit ["Lost terminal"] 22:43:57 -!- billstclair [n=billstcl@unaffiliated/billstclair] has quit [Read error: 113 (No route to host)] 22:44:11 billstclair [n=billstcl@dsl-216-227-57-194.taconic.net] has joined #lisp 22:44:36 -!- willb [n=wibenton@compsci-wifi-39.cs.wisc.edu] has quit [Read error: 110 (Connection timed out)] 22:44:39 -!- rdd [n=user@c83-250-157-93.bredband.comhem.se] has quit [Read error: 104 (Connection reset by peer)] 22:46:03 grouzen_ [n=grouzen@91.214.124.2] has joined #lisp 22:46:15 serichse1 [n=harleqin@xdsl-81-173-147-120.netcologne.de] has joined #lisp 22:51:38 -!- kmels [n=kmels@190.148.177.22] has quit [] 22:51:42 -!- nvoorhies [n=nvoorhie@67.100.118.26] has quit [Read error: 110 (Connection timed out)] 22:54:23 -!- Slowbro [n=slowpoke@174-21-46-162.tukw.qwest.net] has quit [" HydraIRC -> http://www.hydrairc.com <- IRC with a difference"] 22:54:37 -!- grouzen [n=grouzen@91.214.124.2] has quit [Read error: 110 (Connection timed out)] 22:55:14 -!- hkBst [n=hkBst@gentoo/developer/hkbst] has quit [Read error: 104 (Connection reset by peer)] 22:55:25 abeaumont [n=abeaumon@84.76.48.250] has joined #lisp 22:56:04 -!- dalton is now known as ausente 22:56:13 kmels [n=kmels@190.148.177.22] has joined #lisp 22:57:52 -!- kmels is now known as kmels-siesta- 22:58:31 -!- serichsen [n=harleqin@xdsl-81-173-189-89.netcologne.de] has quit [Read error: 110 (Connection timed out)] 23:01:36 -!- Hun [n=hun@p50993726.dip0.t-ipconnect.de] has quit [Remote closed the connection] 23:02:33 -!- sykopomp [n=user@unaffiliated/sykopomp] has quit [Read error: 104 (Connection reset by peer)] 23:08:14 roygbiv [n=blank@pdpc/supporter/active/roygbiv] has joined #lisp 23:09:50 -!- bgs100 [n=ian@unaffiliated/bgs100] has quit [] 23:10:46 half the address space should be enough for everyone 23:11:00 :-D 23:11:55 syamajala [n=syamajal@97-95-190-124.dhcp.oxfr.ma.charter.com] has joined #lisp 23:12:38 -!- mrSpec [n=NoOne@88.208.105.1] has quit ["bb"] 23:13:25 -!- yahooooo [n=yahooooo@c-76-104-183-185.hsd1.wa.comcast.net] has quit [Connection timed out] 23:13:31 -!- stassats [n=stassats@wikipedia/stassats] has quit [Read error: 113 (No route to host)] 23:14:19 younder: SBCL's GC has had some important optimizations in the last few years. 23:15:45 -!- billstclair [n=billstcl@unaffiliated/billstclair] has quit [Read error: 145 (Connection timed out)] 23:17:11 yahooooo [n=yahooooo@c-76-104-183-185.hsd1.wa.comcast.net] has joined #lisp 23:18:00 -!- rouslan [n=Rouslan@unaffiliated/rouslan] has quit [Success] 23:18:45 pkhuong: ISTR you talking about how genesis should generate a bunch of GC pieces. Generating the code for the various specialized vector types seems like an obvious candidate. The scav/trans/size tables perhaps as well. Did you have anything else in mind? 23:19:33 macondo [n=macondo@unaffiliated/macondo] has joined #lisp 23:20:32 -!- Yuuhi [i=benni@p5483EB87.dip.t-dialin.net] has quit ["ERC Version 5.3 (IRC client for Emacs)"] 23:20:34 Is there any linux RPL implementation? 23:20:47 -!- ikki [n=ikki@201.155.75.146] has quit [Read error: 113 (No route to host)] 23:21:14 so it appears that default values in structures cannot depend on other slots for that structure, what about object classes? 23:24:49 Demosthenex: usually you specialize one of the initialization generic functions to do that. 23:25:02 like intialize-instance, shared-initialize, etc 23:25:10 -!- ausente [n=user7994@187.10.22.185] has quit ["eject"] 23:25:20 -!- grouzen_ [n=grouzen@91.214.124.2] has quit ["Lost terminal"] 23:25:40 that'd work, just make a parameter optional, and if its nil, generate it 23:25:46 ok, i'll look at that 23:25:56 objs dont seem much worse than a struct in this regard. 23:26:05 and oddly enough, lisp objs remind me alot of perl's moose. 23:26:09 or vice versa 23:26:21 Xach: so do you specialize in protoss, or just quoting them? 23:27:06 -!- ignas [n=ignas@ctv-79-132-160-221.vinita.lt] has quit [Read error: 110 (Connection timed out)] 23:27:51 Demosthenex: i just noodle around with them, so i know the tech tree and can follow along a little better on the korean casts. 23:28:07 *Xach* has never actually played anyone not a roommate 23:30:05 fvw [n=sdfpme@59.39.243.199] has joined #lisp 23:32:32 -!- gonzojive [n=red@condi.Stanford.EDU] has left #lisp 23:32:54 Bah. 23:33:07 *luis* realizes that gencgc copies some large objects, but not others. 23:33:09 Xach: ah ;] 23:33:13 aja [n=aja@S01060018f3ab066e.ed.shawcable.net] has joined #lisp 23:33:33 luis: which "not others" ? 23:34:00 Lorance [n=Lorance@32.134.77.10] has joined #lisp 23:34:08 code objects for instance, even if they're large (> 4 pages, IIRC) are copied around nevertheless. 23:34:30 instances as well, probably. 23:35:12 AFAICT, only large vectors stay still. 23:35:42 nvoorhies [n=nvoorhie@adsl-75-36-204-63.dsl.pltn13.sbcglobal.net] has joined #lisp 23:36:14 -!- Odin- [n=sbkhh@s121-302.gardur.hi.is] has quit [] 23:36:31 -!- Lorance [n=Lorance@32.134.77.10] has quit ["used jmIrc"] 23:37:03 -!- koning_robot [n=aap@88.159.107.70] has quit [Read error: 60 (Operation timed out)] 23:37:08 Let me rephrase that. It was my initial impression that large objects weren't moved around. But it appears this only applies to vectors. 23:38:30 sykopomp [n=sykopomp@unaffiliated/sykopomp] has joined #lisp 23:39:24 -!- yahooooo [n=yahooooo@c-76-104-183-185.hsd1.wa.comcast.net] has quit [Success] 23:39:57 bombshelter13_ [n=bombshel@76-10-149-209.dsl.teksavvy.com] has joined #lisp 23:40:41 bwat74 [n=bwat@host-90-232-99-133.mobileonline.telia.com] has joined #lisp 23:41:08 -!- athos [n=philipp@92.250.250.68] has quit ["leaving"] 23:41:10 Ok, so now I have to figure out a quick "is this a vector?" type check. 23:42:33 -!- claar [n=claar@85-169-63-25.rev.numericable.fr] has quit [Read error: 104 (Connection reset by peer)] 23:43:13 clhs vectorp 23:43:14 http://www.lispworks.com/reference/HyperSpec/Body/f_vecp.htm 23:43:25 claar [n=claar@85-169-63-25.rev.numericable.fr] has joined #lisp 23:43:35 Hello 23:45:04 pjb: hmm, thanks. Unfortunately I don't understand the implementation of VECTORP very well. 23:46:27 koning_robot [n=aap@88.159.107.70] has joined #lisp 23:46:30 luis: anyways, probably the check is not exactly vectorp, it's possible there are other things that are vector-like internally (eg. perhaps structures?) 23:47:14 Better find the exact test made by gencgc. 23:47:27 pjb: AFAICT, structures are not vector-like as far as gencgc is concerned. 23:47:33 ok. 23:48:18 pjb: the trans_vector_* functions call copy_large_object(), that's it. I guess I'll just check every vector widetag with a big switch and try to be smarter later. 23:50:12 -!- densem [n=fasteez2@218.217.72-86.rev.gaoland.net] has quit [Read error: 113 (No route to host)] 23:50:35 -!- bwat74 [n=bwat@host-90-232-99-133.mobileonline.telia.com] has quit [Remote closed the connection] 23:54:17 arbscht [n=arbscht@unaffiliated/arbscht] has joined #lisp 23:55:16 -!- spilman [n=spilman@ARennes-552-1-36-239.w92-135.abo.wanadoo.fr] has quit ["Quitte"] 23:55:41 ajhager [n=ajhager@c-98-223-251-77.hsd1.in.comcast.net] has joined #lisp