00:00:20 No idea. :o 00:01:14 well, you can always add some -- there are ways to #! lisp 00:03:10 quek [n=read_eva@router1.gpy1.ms246.net] has joined #lisp 00:03:46 *myrkraverk* is now wondering why there is no unix lisp shell, after all, there is a c shell 00:04:01 myrkraverk: there is. 00:04:08 there is? 00:04:08 -!- metawilm [n=willem@e179022050.adsl.alicedsl.de] has quit [] 00:04:10 i just tried ABLE lisp editor, seems to be good enough 00:04:15 of course there is ;) 00:04:16 There's SCSH, for starters. 00:04:21 more than one 00:04:30 http://clisp.cons.org/clash.html for example 00:04:49 hrm 00:05:27 i think i'm going to call it a day at this point, i've got my frequency matrix generator, my frequency-to-probability converter, and tomorrow i can write the matcher and be done with this week's code :D 00:05:58 thanks for the help regarding the division/float thing, i'll keep an eye on that when i run it with real code and not small test snippets :D 00:07:51 bakkdoor [n=bakkdoor@xdslct066.osnanet.de] has joined #lisp 00:08:05 then again, i'm probably going to have to deal with the pain of finding a specific machine to test this on, given that the code is using a function i don't seem to have in clisp (and it's not in the hyperspec...) 00:08:23 JoshJ: and what is that function? 00:08:28 while 00:08:42 -!- pstickne is now known as peaches` 00:08:49 i'll look into it tomorrow, it may be that i missed a utility macro somewhere 00:08:56 that looks like a PGism. You could always replace that macro with the standard (loop while ... do ...). 00:09:09 well, it'd be an Allegro-ism if anything 00:09:12 chavo_ [n=user@c-24-118-166-62.hsd1.mn.comcast.net] has joined #lisp 00:10:07 regardless, i'll look into it tomorrow 00:10:15 -!- JoshJ [n=JoshJ@r38h107.res.gatech.edu] has quit ["Leaving"] 00:10:22 AshyIsMe [n=User@202.176.4.21] has joined #lisp 00:11:31 -!- schoppenhauer [n=css@unaffiliated/schoppenhauer] has quit [Remote closed the connection] 00:11:49 -!- rajesh [n=rajesh@nylug/member/rajesh] has quit ["leaving"] 00:12:25 nyef pasted "Encoding immediate data for ARM data-processing instructions" at http://paste.lisp.org/display/69609 00:12:53 gigamonkey`: Sorry that didn't work, the error must be building up somewhere else as well. 00:12:53 tmh, memo from gigamonkey`: sadly, that didn't help. Thanks though. 00:12:59 Thanks minion 00:13:04 So, feedback request time: Can anyone improve this function for me? 00:13:13 (Does this function even work?) 00:13:43 -!- metawilm___ [n=willem@e179029034.adsl.alicedsl.de] has quit [Read error: 110 (Connection timed out)] 00:16:44 -!- tritchey [n=tritchey@c-98-226-113-211.hsd1.in.comcast.net] has quit [] 00:16:50 minion: thanks 00:16:50 np 00:16:51 -!- ``Erik__ is now known as ``Erik_ 00:17:09 -!- milanj [n=milan@212.200.217.28] has quit ["Leaving"] 00:18:21 nyef: pretty sure you can do test if you can encode it as a regular shift, and then try for an encoding that uses the roll to have data in both the low and hit bits. 00:21:35 tritchey [n=tritchey@c-98-226-113-211.hsd1.in.comcast.net] has joined #lisp 00:26:03 -!- chavo_ [n=user@c-24-118-166-62.hsd1.mn.comcast.net] has quit [Remote closed the connection] 00:26:26 -!- Odin- [n=sbkhh@s121-302.gardur.hi.is] has quit [] 00:27:49 Something based on (integer-length operand) and (integer-length (logand operand (1- operand))), maybe? 00:28:02 Hrm... 00:28:11 Is that what I mean? 00:28:16 minion: memo for gigamonkey`: My experiment must have been flawed and not captured enough of the operations. 00:28:16 Remembered. I'll tell gigamonkey` when he/she/it next speaks. 00:28:48 nyef: i was more thinking integer-length and mask to see if you only need contiguous bits 00:29:17 and integer length on two parts (e.g. the first byte and the rest) 00:29:35 rpg [n=rpg@216.243.156.16.real-time.com] has joined #lisp 00:30:00 Mmm... I'm actually not sure that would be cheaper, given integer-length on 32-bit integers. 00:30:09 xreyes__ [n=xreyes@8.84-48-175.nextgentel.com] has joined #lisp 00:31:11 nyef: doesn't ARM have that instruction? 00:31:48 I don't know, I'm not that familiar with the ARM instruction set. 00:32:13 I'm fairly sure that x86 doesn't, though, and guess what my host platform is? 00:33:13 x86 certainly does, but arm seems not to. 00:33:54 -!- xreyes_ [n=xreyes@8.84-48-175.nextgentel.com] has quit [Read error: 113 (No route to host)] 00:35:03 -!- chris2 [n=chris@ppp-88-217-78-62.dynamic.mnet-online.de] has quit [Read error: 54 (Connection reset by peer)] 00:36:51 lisp looks so archaic. :/ 00:37:12 Quadrescence: So does breathing. 00:37:33 No it doesn't. :> 00:37:40 Exactly. 00:38:50 Or maybe "Practical Common Lisp" makes it look old. 00:39:14 Quadrescence: Maybe you could give something a little less hand-wavy? 00:39:41 ... Wait, x86 has an integer-length instruction? 00:40:02 sellout: What do you mean? 00:40:11 -!- njsg_ [n=njsg@unaffiliated/njsg] has quit ["gone"] 00:40:35 Quadrescence: what looks archaic about Lisp? 00:40:36 gigamonkey`, memo from tmh: My experiment must have been flawed and not captured enough of the operations. 00:41:01 Quadrescence: What gigamonkey said. 00:41:22 gigamonkey`: I don't know. I shouldn't really judge it until I learn some of it. 00:42:40 nyef: bsr 00:42:50 Bit Scan Reverse. Right. 00:42:53 gigamonkey: perhaps the way it shouts back at the user 00:43:00 Quadrescence, by that standard, you can also claim C# as old, since it borrows heavily from C and related languages 00:43:07 Are you sure that's not Branch to SubRoutine? 00:43:46 *nyef* has clearly been unduly influenced by 65816 ASM. 00:44:16 wouldn't that be JSR? (maybe not, I never did 65816, so I don't know anything they added after 6502) 00:45:36 No, JSR takes an absolute address. BSR takes a (signed-byte 8) offset. 00:45:43 oh, clever. 00:45:55 Might be a C89 or 6280 instruction instead. 00:46:26 I don't see it in my 6502 assembler 00:46:36 Right, it wouldn't be there. 00:46:54 Even if your 6502 assembler supports the "undocumented" opcode space. 00:47:27 -!- froog_ [n=david@87.192.28.247] has quit [Remote closed the connection] 00:48:04 froog_ [n=david@87.192.28.247] has joined #lisp 00:48:09 There was some stupid trick in viewing the 6502 opcode space that made it all make sense. 00:48:15 And I've managed to forget what it was. 00:48:23 *ushdf* wrote a new song 00:48:27 Do any of you use lisp more than any other language (excluding careers, possibly)? :o 00:48:39 *nyef* uses english more than any other language. 00:48:39 i almost never use lisp 00:48:42 nyef: i hear you were supposed to read it as horizontal microcode. 00:48:42 i'm really terrible at it 00:48:59 Quadrescence: ? I use lisp almost exclusively in my career... 00:49:12 pkhuong: ISTR that it was some specific bit-grouping like 3:3:2. 00:49:34 I know that z80 opcodes make sense on a 2:3:3 basis. 00:49:50 rpg: I mean, using lisp where you don't need to. 00:50:02 (mostly obvious with the LD reg,reg instruction group.) 00:50:45 Quadrescence: everyone needs to use lisp, always. 00:50:53 excluding people using lisp on the job is a curious twist on an old question 00:51:06 (the more common question being "does anyone use lisp for their work?") 00:51:12 And I remember when a bunch of "us" determined that the SPC700 (SPC900?) in the SNES was a messed-about-with 6502, primarily with a funky remapping of the opcode space. 00:51:43 hefner: Right. I mean the opposite. I really don't know much about lisp, so whatever I hear, I'm probably learning something new. 00:52:04 jtoy [n=jtoy@58.63.90.61] has joined #lisp 00:52:05 I don't really know what lisp can and can't do, and with what ease it can do certain things compared to other languages. 00:52:07 if you can count reading rather than writing, my primary language at my last job was x86 assembly =p 00:52:22 pkhuong annotated #69609 with "This is actually more code..." at http://paste.lisp.org/display/69609#1 00:52:31 -!- AshyIsMe [n=User@202.176.4.21] has quit ["Leaving"] 00:52:44 gonzojive_ [n=red@fun.Stanford.EDU] has joined #lisp 00:52:52 Heh. I mostly get paid for server-side javascript work. 00:53:16 BrianRice-mb [n=briantri@c-98-225-51-174.hsd1.wa.comcast.net] has joined #lisp 00:53:26 nyef: I guess it's a bad sign that I just slowly learned the ones I used often enough by heart without trying to find any pattern. 00:53:32 -!- gonzojive_ [n=red@fun.Stanford.EDU] has quit [Client Quit] 00:53:37 SB thread question --- if one invokes interrupt-thread, does the thread resume its current computation after the interrupting function returns? 00:53:54 rpg: hopefully. 00:54:02 rpg: If you're lucky, yes. 00:54:07 More particularly, could I use this to implement a very rudimentary scheduler? 00:54:13 rpg: In generall, interrupt-thread is Not Safe. 00:54:29 *rpg* is trying to write an ACL multiprocessing compat pkg... 00:54:30 rpg: sure you don't want to have your threads wait on a wait queue? 00:55:11 pkhuong: Well, I did too, but as an emulator author I started looking for patterns like that... I'd probably have looked for the same if I was an assembler author as well... 00:55:12 pkhuong: Yes, for now the only piece of code I'm porting I have translated to wait on a gate --- and we've implemented ACL-like gates on top of SBCL mutexes. 00:55:26 rpg: Or hook serve-event. That's always fun. 00:55:55 pkhuong: That really wasn't clear --- what I meant was "the only piece of code I'm porting *that uses acl's process-enable and process-disable*..." 00:56:07 Hrm... 00:56:40 "ADC{cond}{S} , , "... 00:56:44 -!- b4|hraban [n=b4@0brg.xs4all.nl] has quit [Remote closed the connection] 00:58:12 How the heck do I sanely implement this as SBCL instruction definitions? 00:58:44 via combinatorial explosion! 00:59:09 nyef: oh right, and I didn't make sure to minimize the shift count, sorry. 00:59:29 I think that what I want is to have the condition as an optional first argument prior to Rd. 00:59:42 I'll be interested to see if I can recode my entire Fortran program in lisp with ease. :o 01:00:02 Quadrescence: Have a look for f2cl. 01:00:07 Quadrescence: very much, with f2cl (assuming the fortran isn't too modern). 01:00:29 nyef: pkhuong: I'll take a look at it. :O 01:00:30 -!- Nshag [i=user@Mix-Orleans-105-3-144.w193-250.abo.wanadoo.fr] has quit ["Quitte"] 01:00:50 jaoswald [n=user@74.73.49.134] has joined #lisp 01:01:09 I assume operations on lists are... well, they exist. e.g., I can add (1 2 3) to (4 5 6) easily. :o 01:01:14 -!- bakkdoor [n=bakkdoor@xdslct066.osnanet.de] has quit [] 01:01:17 nyef: I don't want to interrupt your SBCL/ARM hacking, but is your port log publicly readable anywhere? 01:01:36 Quadrescence: why would you do use that instead of specialised vectors? 01:01:37 Just don't read the Lisp afterward ... it's been known to drive men insane. 01:01:49 nyef: in particular, I was interested if you had any thoughts on the calling convention and whether the ARM exception standard made sense for Lisp? 01:02:19 jaoswald: No, but I was thinking about either posting it in my web space, or converting it to blog posts. 01:02:32 pkhuong: I have no idea. I don't know any lisp. 01:03:32 jaoswald: No real thoughts as to calling conventions, I'm afraid. 01:03:42 And the exception handling thing... 01:03:48 I still remember Win32 exceptions. 01:04:43 -!- ivansto is now known as ivanst 01:05:48 nyef: I ask because I was thinking of getting a crude Lisp implementation going on ARM, but I know little about ARM assembly, and less about Lisp implementations. I had thought about keeping things as close as possible to the ARM standards, but have no idea where I would run into trouble getting a GC to work, or iplementing CL non-local exits. 01:06:25 unfortunately, most Lisp compiler examples I have seen are geared toward Scheme, or are mature compilers like SBCL. 01:06:28 jaoswald: http://www.lisphacker.com/projects/sbcl-arm/port-log.txt 01:06:38 nyef: I will read it with much interest. thanks. 01:06:47 Wait, why would non-local exits be hard? 01:07:25 tripwyre [n=tripwyre@117.193.162.10] has joined #lisp 01:07:45 jaoswald: scheme is even worse than CL wrt NLX. 01:07:55 -!- krumholt [n=krumholt@port-92-202-39-177.dynamic.qsc.de] has quit ["Ex-Chat"] 01:07:56 So, how would you add (1 2 3) to (4 5 6) ? 01:08:12 For SBCL/Win32, what we do is store a copy of the current SEH frame in each NLX target block when its instantiated, and force a stack unwind at each NLX. 01:08:31 Then we hook into the system stack unwind process for UWP. 01:08:50 I guess there might be trouble if the unwinder doesn't like being unwound... 01:08:50 bakkdoor [n=bakkdoor@xdslct066.osnanet.de] has joined #lisp 01:09:09 Quadrescence: are you interested in learning lisp? 01:09:12 -!- bakkdoor [n=bakkdoor@xdslct066.osnanet.de] has quit [Remote closed the connection] 01:09:21 sykopomp: Very much so, for some reason. 01:09:37 minion: tell Quadrescence about that-dead-sexy-book 01:09:37 Quadrescence: please see that-dead-sexy-book: pcl-book: "Practical Common Lisp", an introduction to Common Lisp by Peter Seibel, available at http://www.gigamonkeys.com/book/ and in dead-tree form from Apress (as of 11 April 2005). 01:09:39 Quadrescence: map[-into] and +. 01:10:43 sykopomp: Any books that are a little less... hum, "fluffed up" I guess? 01:10:51 fluffed up?... 01:11:12 minion: paip 01:11:12 paip: Paradigms of Artificial Intelligence Programming: Case Studies in Common Lisp by Peter Norvig. http://www.cliki.net/paip 01:11:18 PCL does ramble on a bit 01:11:19 Are you looking for the K&R of the Common Lisp world? 01:11:19 blah2345 [i=a@c-67-185-17-84.hsd1.wa.comcast.net] has joined #lisp 01:11:22 maybe that? 01:11:38 The reference at the last of ANSI Common Lisp has been very helpful. 01:11:43 Quadrescence: you could also try Gentle Intro 01:11:54 minion: gentle introduction 01:11:54 you speak nonsense 01:11:58 minion: liar. 01:11:58 please stop playing with me... i am not a toy 01:12:09 sykopomp: I want to download this into my brain. I want the most dense thing I can get. :D 01:12:47 -!- Adamant [n=Adamant@unaffiliated/adamant] has quit [] 01:12:48 (I've programmed in Haskell, C, C++, Fortran, lolBASIC, and others, if that helps) 01:12:52 -!- jlouis [n=jlouis@port1394.ds1-vby.adsl.cybercity.dk] has quit [Read error: 110 (Connection timed out)] 01:13:22 nyef: fwiw, I think I can get a core that write (2) "Hello, world" down to a couple dozen KB by doing an aggressive GC with only the toplevel function as a root (I only have a semi-automatically generated heap groveller). 01:13:25 The most dense thing you can get would be the hyperspec. 01:13:26 Quadrescence: PCL covers a lot of common lisp, and I wouldn't call it fluffed up. If you want the densest thing possible, try just reading through the hyperspec and then going through sbcl's source code. 01:13:54 Quadrescence: but you won't really learn idiomatic lisp from the hyperspec. Maybe a bit from sbcl. 01:14:07 -!- gonzojive [n=red@fun.Stanford.EDU] has quit [Connection timed out] 01:14:29 sykopomp: I guess I shouldn't say "densest". But PCL seems like it's teaching to someone who's never done any programming before. :/ 01:14:41 Quadrescence: and PCL is meant for people who already know how to program. I learned that the hard way :P 01:14:44 Quadrescence: that's exactly not the point. 01:14:52 -!- epoch [n=K-I-S-S@p3m/member/epoch] has quit ["  "] 01:15:09 One moment please. :S 01:15:11 krumholt [n=krumholt@port-92-202-39-177.dynamic.qsc.de] has joined #lisp 01:15:16 Adamant [n=Adamant@c-98-244-152-196.hsd1.ga.comcast.net] has joined #lisp 01:15:19 Quadrescence: PCL is more about teaching you the neat stuff about lisp with actual examples of what it's useful, from a perspective of someone who already has experience in other languages. 01:15:22 -!- Quadrescence [n=quad@c-24-118-118-178.hsd1.mn.comcast.net] has quit [Read error: 104 (Connection reset by peer)] 01:15:23 pkhuong: Cool. I've been thinking about taking the problem of minimal cores from the perspective of RPC/debugging/etc. a live core and then sort of "stuffing the genie back into the bottle" to produce a native dll or exe. 01:15:59 nyef, pkhuong: Like I said, I know little about Lisp implementation; most toy examples deal with scheme, and I worry about details like exactly what needs to be wound onto the stack for CL dynamic binding, how to properly wind & unwind, while maintaining any invariants needed for (multithreaded) GC. But I don't want to clutter the channel more with my ignorance. 01:17:06 hi what can i do if a website is not loading in drakma? 01:17:13 So, for example, genesis would start with an "empty" core in a separate process, and would load stuff in and run top-level forms as if via debugger, with a hook for treating undefined-functions in the target core as RPC calls. 01:17:33 krumholt: Well, the first thing to do is check to see if it loads in firefox... 01:17:45 nyef, no it doesnt 01:17:50 -!- quek [n=read_eva@router1.gpy1.ms246.net] has left #lisp 01:18:09 nyef, i mean how can i solve that problem? there is just nothing happening 01:18:24 nyef: my actual application is rpc; I just realized that I could write a core instead of sending a vector of words + fixup bitmap. 01:18:34 pkhuong: Ah, cool. 01:18:41 nyef, pkhuong: oh, and one more thing, how to deal with CL multiple values. 01:18:54 jaoswald: Ah, now -that's- a tricky subject. 01:19:03 nyef, try this site for example: http://www.informatik.uni-frankfurt.de it won't even load in the browser 01:19:35 krumholt: Umm... Aside from the familiarity of the URL, I don't see how this would be a drakma (or lisp) problem at all... 01:20:43 nyef, what i want to know is if there is a way to tell drakma that there is a timeout or something. if my program encounters an url like that it will hang indefinetly 01:21:33 krumholt: Look for a "TIMEOUT" parameter to the various drakma APIs. If you're on SBCL, maybe look at the deadline stuff. Beyond that, you might just be looking for trouble. 01:21:37 -!- O_4_ [n=souchan@ip-118-90-42-91.xdsl.xnet.co.nz] has quit [] 01:22:12 krumholt: maybe one of the three foo-timeout keyword arguments for http-requestç 01:22:53 jaoswald: What I've found is that you can get interesting and useful results even when you disregard host ABI integration for various things. 01:23:36 pkhuong, it says in the api they are only available for lispworks 01:24:18 Really, I was probably the only one to run into the stack unwind problems on SBCL/Win32 relating to native exception handling, and it was broken in public releases for months. 01:24:39 nyef: I think the real issue is that I think about this stuff for roughly thirty minutes every other day while bored, and that is not enough to learn anything. 01:24:56 jaoswald: Ah. Then try implementing some subset of it? 01:25:16 Somewhere around I have my original set of tests for Win32 exception handling behavior. 01:25:58 nyef: I will probably start by porting my CL implementation of COMFY to ARM to learn ARM assembly. By then you should be well on your way to porting a real CL to ARM, and I'll avoid the hard thinking that way. 01:26:02 -!- Yuuhi [n=user@p5483DED1.dip.t-dialin.net] has quit ["ERC Version 5.3 (IRC client for Emacs)"] 01:26:52 Heh. "should". 01:27:15 Honestly, ARM exception handling scares me. 01:27:27 For the same reason that x86-64 exception handling scares me, actually., 01:27:44 I expect that SBCL falls down completely, there. 01:27:58 -!- ebzzry [n=rmm@115.147.100.214] has quit [Remote closed the connection] 01:28:21 ebzzry [n=rmm@115.147.100.214] has joined #lisp 01:28:27 Nshag [n=shagoune@99.a2c-250-158.astra2connect.com] has joined #lisp 01:28:58 quek [n=read_eva@router1.gpy1.ms246.net] has joined #lisp 01:29:04 are there any alternatives to drakma? 01:29:12 nyef: ABI compliant exception handling? pfah (: 01:29:32 krumholt: just use a deadline if you're on sbcl, or whichever feature that your implementation provides 01:29:43 pkhuong: Absolutely necessary on those platforms where the UI ABI uses it around callbacks. 01:30:00 krumholt, you could wrap the call to drakma:http-request in sb-ext:with-timeout .. or just add support for timeout in drakma for sbcl 01:30:08 *Xach* gets a chance to make a zpb-ttf release, 2 years after the last one 01:30:18 pkhuong: Which will be more and more of them as time goes by, provided they don't just fall back and provide C++-only APIs, the bastards. 01:30:26 *nyef* is still bitter about BeOS. 01:30:30 hm ok will do the sb-ext:with-timeout 01:30:33 O_4 [n=souchan@ip-118-90-42-91.xdsl.xnet.co.nz] has joined #lisp 01:30:43 thanks 01:30:49 krumholt: No, don't do that! That's where the monster is hiding! 01:30:52 nyef: seems like it would be quite the task. 01:31:01 krumholt: yeah, I hope this is for a quick hack 01:31:17 no i need a good solution 01:31:36 whats bad about the sb-ext:with-timeout? 01:31:37 krumholt: minion uses with-timeout around its cliki lookup. It took us ages to sort out where the file-descriptor leak was and plug it. 01:32:04 usocket has or had a leak, yes 01:32:15 i reported it, and it should be fixed by now 01:32:15 -!- Adamant [n=Adamant@unaffiliated/adamant] has quit [No route to host] 01:32:17 great :( 01:32:59 pkhuong: Yeah, it was at least possible on Win32, but I really don't know about these platforms that use the DWARF debugging tables as a basis for frame unwinding. 01:33:14 Even though I know it's something that people have been asking for in SBCL. 01:33:50 krumholt, you can use lsof to see if you have a leak .. just try a couple of times in a loop with a very short timetout .. and see if you have a leak 01:34:04 krumholt, (lsof is a unix/linux-tool) 01:34:13 nyef: right, I was thinking of linux/x86-64. 01:34:22 Exactly. 01:34:39 How on -earth- did they expect to be able to support runtime code generation on such an ABI? 01:34:50 -!- Nshag [n=shagoune@99.a2c-250-158.astra2connect.com] has quit ["Quitte"] 01:35:02 the JVMs must do it somehow. 01:35:15 Yeah... Separate FFI threads, maybe? 01:35:21 right. 01:35:43 Oh, god. You're serious? Separate FFI threads? 01:36:47 nyef: Maybe I'm just completely ignorant, but I had been planning to emit code decorated with the appropriate exception information at run-time. 01:38:46 nyef: I believe JNI works that way (or similarly, but IC definitely BW). 01:39:09 jaoswald: The problem, if you're doing this at run-time instead of binding a new shared object or something, is hooking the exception tables into the runtime and keeping things from getting confused when the GC runs. 01:40:06 Of course, if you're GCing, what you can do is unhook the exception tables, GC, rebuild the exception tables with all of the new function locations and re-hook the tables. 01:40:12 It's just a pain. 01:40:23 do you mean GC with respect to moving/GC'ing function objects, or more generally? 01:40:31 And that's assuming that there's an API to let you do that, instead of having to monkey about with a shared object. 01:40:40 Yes, moving/GCing function objects. 01:40:52 Or, more specifically, their underlying code sections. 01:41:46 Now, if all of your functions are implemented as unrelocatable code in an executable or dynamic shared object, you don't really have to worry about GC effects causing code motion. 01:42:09 -!- quek [n=read_eva@router1.gpy1.ms246.net] has left #lisp 01:42:24 nyef: I think Sun's did something like that, but that could definitely have changed when they stopped doing green threading. 01:42:30 But then you're looking at a new shared object created on disk and loaded for each call to COMPILE. 01:42:47 ok, I think that makes sense. 01:43:51 This is one of those "I'll burn that bridge when I get to it" things. 01:44:41 Really, the main reason I did the full integration for Win32 exceptions is that I ended up in the debugger from my window function callback when I was messing around with GUI programming. 01:45:19 ths_ [n=ths@X4d78.x.pppool.de] has joined #lisp 01:45:49 Nshag [i=user@Mix-Orleans-106-3-209.w193-248.abo.wanadoo.fr] has joined #lisp 01:46:02 And the Win32 exception stack does its own sanity checking and will kill processes outright if its sanity check bounces. 01:47:35 ARM and x86-64 exceptions seem to be based on walking stack frames as described by the debugging information, which doesn't have nearly the same failure modes. 01:47:50 Forced unwind is nasty, though. 01:49:07 But most of this only applies when dealing with alien code, so a lisp system can end up being perfectly usable without proper exception integration. 01:51:28 It's just one of those things, like how SBCL doesn't come with a good cocoa bridge on OSX. 01:51:57 Well, maybe not that, but... 01:54:13 nyef: I was sort of aiming, long term, for some CL-like Lisp environment that would live well in environments like the iPhone or similar ARM devices. Playing by the platform rules felt like a good idea. 01:54:42 jaoswald: you would want to do that just to have smaller images, anyway. 01:55:14 -!- O_4 [n=souchan@ip-118-90-42-91.xdsl.xnet.co.nz] has quit [] 01:56:15 -!- ths [n=ths@X6c8d.x.pppool.de] has quit [Read error: 110 (Connection timed out)] 01:56:19 I'm not convinced those environments need a typical CL-like environment 01:57:38 pkhuong, hefner: I don't quite follow. 01:57:57 -!- krumholt [n=krumholt@port-92-202-39-177.dynamic.qsc.de] has quit ["Ex-Chat"] 01:58:16 devices, rather. but anyway you're not likely to be using slime on your iphone to develop apps. 01:58:52 and you wouldn't want to bring your compiler/evaluator on the iphone for a random app. 02:01:34 So the idea is to still get as much mileage out of lambda as one can, without the compiler around at run-time? 02:02:03 that's a curious way to put it. 02:04:22 Well, a certain fraction of Lisp programs get by with all their code visible at compile/development time, and lambda just serves to produce closures of known code with new instances of the bindings, or am I completely wrong? 02:06:06 And pkhuong's statement is that for this style of program, you leave the compiler behind at execution time. 02:06:14 -!- rvirding [n=chatzill@h235n3c1o1034.bredband.skanova.com] has quit ["ChatZilla 0.9.83 [Firefox 3.0.3/2008092417]"] 02:07:06 *hefner* notes the channel has come full circle, back to embedding CL (maybe it never really left the topic) 02:10:06 echo-area [n=user@nat/yahoo/x-7544ccafdf9ee778] has joined #lisp 02:12:58 haiwei [n=haiwei@192.9.202.3] has joined #lisp 02:14:10 well, good night, all. And, nyef, just wanted to let you know there are folks cheering from the sidelines for SBCL/ARM. 02:14:21 jaoswald: Sleep well, and thanks. 02:14:40 -!- jaoswald [n=user@74.73.49.134] has quit ["ERC Version 5.1.3 (IRC client for Emacs)"] 02:14:45 Maybe soon I'll type up my notes on WDL/NIL. Most of them are hand-written. 02:15:45 Might be amusing to have a lashup with an embryonic core running in an emulator strapped to a host-side RPC/debugger thing. 02:15:58 (for the record, I think SBCL/ARM sounds like great fun, I just wonder how long it'll be until there's a widespread ARM device big enough to be interest) 02:16:01 But that's for much later. 02:16:24 hefner: Does the NSLU2 count? Only 32 megs of RAM, but it's a 266MHz device and takes external USB hard drives. 02:17:11 widespread? 02:17:36 Xach: I threw that in so that some smartass couldn't counter with "my ARM development board has 800 MB of ram on it" 02:17:42 elurin [n=user@81.213.203.109] has joined #lisp 02:19:16 nyef: I'll leave it to your expert judgement whether or not SBCL is really usable for anything besides printing its startup banner in 32 MB :) 02:19:38 xreyes_ [n=xreyes@8.84-48-175.nextgentel.com] has joined #lisp 02:20:27 Truly, usability on such a platform is irrelevant. 02:20:44 Functioning at all on such a platform is the goal. 02:21:03 -!- xreyes__ [n=xreyes@8.84-48-175.nextgentel.com] has quit [Read error: 110 (Connection timed out)] 02:21:23 that's the spirit. plan for the future, I guess. 02:38:32 Okay, I'm going to pretend to go to bed now. I'll be back tomorrow, unless I'm not. 02:38:34 fusss [n=chatzill@pool-72-66-40-106.washdc.east.verizon.net] has joined #lisp 02:38:36 G'night all. 02:38:41 nite nyef 02:39:02 -!- wol [n=wol@c-98-207-129-115.hsd1.ca.comcast.net] has quit [Remote closed the connection] 02:39:59 mch [n=bitsurge@S01060004e24ffe41.cg.shawcable.net] has joined #lisp 02:40:10 AshyIsMe [n=User@202.176.4.21] has joined #lisp 02:41:38 with regards to beach's talk about hosting mirrors of cliki/cl.net hosted libraries 02:42:01 it would be nice to have cliki/cl.net cache the _latest.tar.gz copies of each 02:42:39 if cliki.net/foo?download return 404, then asdf-install/clbuild can fall back to retrieving the last good copy 02:43:30 tarball caches are also good for fetching projects that don't have any releases but ask users to check it out from git/darcs/svn/cvs, etc. 02:46:18 I'm still fond of my idea of using version control for mirroring 02:46:19 O_4 [n=souchan@ip-118-90-42-91.xdsl.xnet.co.nz] has joined #lisp 02:46:51 like clbuild, but instead of pulling from 50 repositories and some mirrored tarballs, you pull the whole thing from one repository 02:47:22 tricky, though 02:47:49 crod [n=cmell@p4186-ipbf4108marunouchi.tokyo.ocn.ne.jp] has joined #lisp 02:47:53 hmmm 02:48:06 EtFb [n=etfb@mail.hatrix.com] has joined #lisp 02:48:08 people would have to g 02:48:26 either your vcs has to be "nestable", you put exported copies rather than version control checkouts into it (which is no fun, really), or you use a vcs that none of the subprojects use 02:48:32 people would have to give up control of their code base and let one repository to host them? 02:48:37 no, not at all 02:49:27 -!- jtoy [n=jtoy@58.63.90.61] has left #lisp 02:49:29 -!- jestocost [n=cmell@p4186-ipbf4108marunouchi.tokyo.ocn.ne.jp] has quit [Read error: 110 (Connection timed out)] 02:49:42 put aside vcs for a minute and let's stick to asdf-install/clbuild. maybe we add vcs checkout functionality by specializing ASDF methods and using cl-darcs or similar mixins 02:49:46 the nice thing there is that you could tag or branch or whatever across the entire collection to coordinate versions across packages for things with trick dependencies 02:49:59 that's good 02:52:46 jtoy [n=jtoy@58.63.90.61] has joined #lisp 02:57:17 -!- froog_ [n=david@87.192.28.247] has quit [] 02:59:16 maybe this is what git-submodule is for. 03:00:04 gjvc_ [n=gjvc@chardonnay.extremis.net] has joined #lisp 03:00:42 -!- Piranha__ [i=jabber-i@number-41.thoughtcrime.us] has quit [Read error: 104 (Connection reset by peer)] 03:01:12 yahooooo3 [n=yahooooo@c-24-19-6-19.hsd1.wa.comcast.net] has joined #lisp 03:01:20 -!- H4ns2 [n=Hans@p57BBABCD.dip0.t-ipconnect.de] has quit [Read error: 104 (Connection reset by peer)] 03:01:20 -!- gjvc [n=gjvc@chardonnay.extremis.net] has quit [Remote closed the connection] 03:01:20 -!- yahooooo [n=yahooooo@c-24-19-6-19.hsd1.wa.comcast.net] has quit [Remote closed the connection] 03:01:21 -!- fihi09 [n=user@pool-96-224-164-180.nycmny.east.verizon.net] has quit [Remote closed the connection] 03:01:21 -!- kidd [n=kidd@167.Red-79-147-12.staticIP.rima-tde.net] has quit [Remote closed the connection] 03:01:21 fihi09 [n=user@pool-96-224-164-180.nycmny.east.verizon.net] has joined #lisp 03:01:26 H4ns [n=Hans@p57BBABCD.dip0.t-ipconnect.de] has joined #lisp 03:01:37 -!- mogunus [n=marco@173.9.7.10] has quit [Remote closed the connection] 03:01:43 froog_ [n=david@87.192.28.247] has joined #lisp 03:01:47 mogunus [n=marco@173.9.7.10] has joined #lisp 03:01:56 kidd [n=kidd@167.Red-79-147-12.staticIP.rima-tde.net] has joined #lisp 03:03:12 -!- Bzek [n=SK_sj@mcc-dyn-31-225.kosnet.ru] has quit ["L"] 03:04:32 karpar [n=zhili@2001:da8:8000:d010:0:5efe:7dd9:f6d9] has joined #lisp 03:06:28 replor_ [n=replor@ntkngw375028.kngw.nt.ftth.ppp.infoweb.ne.jp] has joined #lisp 03:07:05 -!- karpar [n=zhili@2001:da8:8000:d010:0:5efe:7dd9:f6d9] has quit [Client Quit] 03:07:34 Is there a portable predicate/test that would highlight "shared/shareable" values (those that are referenced) vs atomics stored within the slot itself? For example, fixnum, t and nil are stored in the reference. 03:08:35 With fixnum, I definitely have no idea what you mean. 03:09:17 -!- nasloc__ [i=tim@kalug.ks.edu.tw] has quit [Remote closed the connection] 03:09:29 timchen1` [i=tim@kalug.ks.edu.tw] has joined #lisp 03:09:50 Piranha__ [i=jabber-i@number-41.thoughtcrime.us] has joined #lisp 03:10:13 -!- fihi09 [n=user@pool-96-224-164-180.nycmny.east.verizon.net] has quit [Remote closed the connection] 03:10:43 Can I get a small favor from somebody with access to SBCL on multicore/multi-CPU hardware? 03:10:43 fihi09 [n=user@pool-96-224-164-180.nycmny.east.verizon.net] has joined #lisp 03:10:46 -!- thomas001 [n=thomas@p54B75281.dip.t-dialin.net] has quit [Remote closed the connection] 03:11:25 Piranha__: just ask. 03:11:33 I have a test program which should consume CPU for a few seconds with two threads. The threads contend for a shared lock. For some reason, on my machine, the program causes kipmi0 to go nuts. 03:11:45 I couldn't reproduce the behavior on a single-CPU, 64-bit machine. 03:12:10 I don't have an IPMI card in my multicore machine, so it's doubtful I can reproduce that. 03:12:38 I've tried recreating the problem with a few different test programs, but couldn't. So now I'm trying to whittle down the program as best I can. 03:12:40 -!- adeht [n=death@nessers.org] has quit [Read error: 113 (No route to host)] 03:12:42 pkhuong: I assume fixnum is the number of bits that are stored in the reference vs stored on the heap 03:12:44 What kind of firmware is the IPMI running? Do you get ssh access to it? 03:13:10 I guess what I'm asking is if there's a way to determine if a variable's value is stored on the heap or not. 03:13:13 chandler: I own the machine. I don't know anything about its IPMI setup; I didn't even know what that was until I saw this problem. 03:13:17 Modius__: so what are `atomics', if fixnum aren't one of them. 03:13:28 Piranha__: well now you know the problem. 03:14:01 pkhuong: But huge numbers pass the atom test as well 03:14:11 Modius__: sure. They're not conses. 03:14:19 pkhuong: I don't. My guess is that it's some esoteric kernel-related interaction with multithreading. Like I said, I can't reproduce it with simple cases. 03:14:20 clhs atom 03:14:20 http://www.lispworks.com/reference/HyperSpec/Body/a_atom.htm 03:14:52 Is there a way to differentiate allocated values from those that are stored in with the reference (like nil and fixnum)? 03:15:30 Since Atom doesn't provide this information I am asking for something else, or confirmation that there is no such determination that would be remotely portable. 03:15:47 I don't know why you're putting NIL in that category. (The usual term for what you're talking about is "immediate", FYI) 03:15:47 Modius__: how is nil more ``stored in with the reference'' than any symbol? 03:16:03 I just don't understand how IPMI could have any interaction with a program that burns CPU time with two threads, a lock, and a condition variable. 03:16:26 Piranha__: I don't either. This seems like something that's worth taking up with the machine vendor, frankly. 03:17:03 Piranha__: I don't see how that's not, a priori, a bug in your box's IPMI suite. 03:17:12 In the implementation that I am using, it does not appear to allocate for nil or fixnums; but allocates for large numbers. I was asking if there was a way to query values for belonging to one of these groups or the other. 03:17:36 b4|hraban [n=b4@0brg.xs4all.nl] has joined #lisp 03:17:38 adeht [n=death@nessers.org] has joined #lisp 03:18:05 Modius__: If memory serves, SBCL calls such things "immediate values", though I have no idea if it offers a test for them. So far as I know, CL does not distinguish between an immediate value and one which goes through a pointer. 03:18:16 pinterface: Thanks 03:18:23 Modius__: There appears to be some confusion on your part. NIL is already allocated; it's a pointer. Fixnums are immediates, and are differentiated from pointers by the tag in the value. Bignums, like most other kinds of numbers under most circumstances, are pointers and must be allocated. 03:18:39 chandler: Thanks. 03:18:59 Computing a bignum is no different than creating a cons cell. 03:19:28 Reason I asked as I'm in a situation where I want (trivial garbage) weak hash tables for the keys that are allocated/non-immediate only. I realize that this skirts nonportability and will probably not work. 03:20:12 Modius__: and what would it do for immediate keys? 03:20:28 pkhuong: Not store them in the hashtable 03:21:35 -!- EtFb [n=etfb@mail.hatrix.com] has quit [Read error: 110 (Connection timed out)] 03:21:39 chandler: Is the fixnum/larger-than-fixnum divide considered, in the standard, the point between immediate and allocated? Or is this, while being a common implementation detail, technically an assumption. 03:22:07 On SBCL, I guess you could (logbitp 0 (sb-kernel:lowtag-of x)). 03:23:08 -!- photon [n=photon@unaffiliated/photon] has quit [Read error: 113 (No route to host)] 03:23:56 Modius__: The standard doesn't say anything about immediates. 03:24:10 -!- replor [n=replor@ntkngw375028.kngw.nt.ftth.ppp.infoweb.ne.jp] has quit [Connection timed out] 03:24:30 In essence, it allows any value not to be an immediate, because numbers are not guaranteed to be EQ. 03:24:49 tiesje [n=user@202.51.72.181] has joined #lisp 03:24:58 pkhuong's solution works, though it screams "break me" to an SBCL maintainer :-) 03:25:22 pkhuong: Thanks 03:25:27 Modius__: pretty much the only thing I can think of for fixnums is that array indices (and row-major-indices) must be fixnums and that a-t-s-l must be at least 1024 or something tiny like that, which gives you a lower bound on m-p-f. 03:25:47 there's an explicit bound on m-p-f, since fixnum is a supertype of (signed-byte 16) 03:25:57 chandler: Yeah, and I was looking for something portable anyway. Going that low level I might as well have #+s for different implementations with type lists 03:26:30 chandler: i think that assumption is pretty deeply wired in various bits of the code base. 03:26:43 pkhuong: true 03:27:18 I just shudder when I see code outside the compiler dealing in concepts like lowtags :-) 03:27:52 I can probably err the other way, merely filter "in" bignum, symbol, sequence, cons 03:28:02 Modius__: I do understand what you're trying to do, but I'm afraid there's not much help beyond pkhuong's solution. When I ran into a similar situation, I just forbid numbers and characters from the hash. 03:28:05 chandler: I hope I can have something truly shudder worthy for christmas, then. 03:28:21 uh oh. what's this? :-) 03:28:39 chandler: Yeah, I need to back away from detecting immediacy and err in the other direction. 03:29:26 I don't think it makes much sense to have numbers or characters in a head-weak hash table anyway. 03:29:27 chandler: pass partial heaps around with a copying function generated mostly via primitive object definitions. 03:29:34 chandler: It's probably safe "enough" to assume bignum is allocated 03:29:47 pkhuong: Aiee, that's good hacking. :-) 03:30:08 chandler: it's mostly much less code than my previous attempt. 03:30:15 Modius__: what do you do with cached bignums? 03:30:19 chandler: It's a weird situation, I'd want to once-only huge numbers of undetermined size 03:30:25 Modius__: ... but in an implementation like SBCL where head-weak hash tables can't be EQL hashes, how would you retrieve it? 03:30:45 The implementation is permitted to copy those whenver it wants to. 03:30:53 In essence, bignums are the worst of both worlds. 03:31:01 chandler: Bignums can be copied? I did not know that 03:31:17 chandler: Only on mutation or at any time? 03:31:20 Modius__: numbers and eq => lose. 03:31:23 At any time. 03:31:41 Anyone have a quick link to the SBCL weak hash table restriction? 03:31:44 An implementation is permitted to always return NIL for EQ of any numbers or characters. 03:31:51 No. Try it and see. 03:32:02 I'm in Lispworks so I have to dig a bit to see what TG is doing 03:35:43 ... actually, I may have misremenbered. 03:35:47 "misremembered" 03:36:20 It's Clozure CL I was thinking of. 03:36:27 epoch [n=K-I-S-S@p3m/member/epoch] has joined #lisp 03:36:43 -!- bighouse [n=bighouse@modemcable012.160-131-66.mc.videotron.ca] has quit [Connection timed out] 03:37:09 All i find on SBCL is "weak hash tables don't work" 03:37:11 are you SBCL guys going to stand idly by while CCL outdoes you in obnoxious pedantry like that? 03:37:20 yes 03:37:31 sbcl is falling behind on the fundraising drive 03:37:49 pkhuong: I was not attempting to eql numbers - I've been on lispworks weak eql-tables have been working 03:38:01 pkhuoing: I mean, I know not to eq numbers 03:38:33 $20,000 seems like a bargain for developing a fancy IDE, even with the fantastic powers of lisp 03:38:35 Modius__: then you know that they can be copied arbitrarily by the implementation. 03:38:58 although if you'd asked me in 2003, I'd have assured you that I could have something running by the end of the summer 03:39:11 in clim? 03:39:16 Modius__: weak hash tables don't work? 03:39:19 What I didn't know was the state of the weak hash tables across major implementations. Trivial-garbage doesn't seem to enshrine such limitations 03:39:49 Somebody said SBCL supported eq-only, and my searching on the subject just finds posts from 2005 etc complaining that they don't really work. 03:40:27 Xach: yes, but I'm only 25% serious, as writing an IDE was never a goal of mine, and various pieces (beginnings of my inspector and debugger) now rot in the dustbin of history. 03:40:33 hefner: what was missing from the Planet Lisp announcement is that $20k is half the real number 03:40:47 thus the only logical conclusion: SBCL's weak hash tables don't work, in late 2009. 03:40:55 Clozure will put in the other half 03:41:24 -!- fusss [n=chatzill@pool-72-66-40-106.washdc.east.verizon.net] has quit [Remote closed the connection] 03:41:29 pkhuong, time travelling hacker extrairdinaire! 03:41:50 Well, I'm having trouble finding anything more than that and that someone here said they were eq-only. Confidence is really inspired 03:41:51 ouch. I blame the hour change. 03:42:20 Daylight Savings Year 03:42:58 anyway, my typos-per-line threshold was hit a minute ago. good night 03:43:16 Is there some way one can grab a piece of the latest SBCL without having to check out the whole thing? I'm grinding through mp and sockets compat programming, and figure I might make others' lives easier if I update the manual as I go... 03:44:01 cvs accepts subdirectories in a checkout command 03:44:18 chandler: OK, will pull that.. 03:45:10 funny enough, I pondered blogging some reflections on mcclim, the listener and other components, and how none of it turns out to be as useful as I imagined it would be, and chanced via reddit upon evidence that the outside world has finally reinvented the notion of a graphical repl 03:45:14 photon2 [n=photon@unaffiliated/photon] has joined #lisp 03:45:17 Here's one thing I would put in --- in sockets, does anyone know the format of the address and port arguments? Both integers? 03:45:59 http://coding.derkeiler.com/Archive/Lisp/comp.lang.lisp/2006-06/msg00399.html <-- Does anyone have a link to anything about SBCL weak hash tables other than anecdotals? 03:46:30 hefner: it's actually one of the important criteria we've been looking at for a teaching language. 03:48:08 pkhuong: interesting. for the sake of being able to do graphics in it, or just general user friendliness? 03:49:06 -!- photon2 is now known as photon 03:49:14 hefner: to have (near-) immediate graphical (read interesting) feedback. 03:49:33 Modius__: how about trying it out? 03:49:44 pkhuong: I'm not really set up in SBCL 03:49:56 I'd really like to see any docs - like the one that says it's eq-only 03:49:58 maybe if nobody complains, it's because it works. 03:50:04 Modius__: so would I. 03:50:07 especially that one. 03:50:15 pkhuong: awesome, you can teach them to implement logo as an exercise. 03:51:22 chandler: If you could point me at wherever you found about SBCL weak hash table limitations I'd sure appreciate it 03:52:05 chandler: ah. Not so clever of me. I didn't realize that the real documentation is in the source, in the docstrings. 03:52:14 Modius__: you'll have to look in ccl's documentations for that, since chandler wrote shortly after that he'd misremembered. 03:53:27 Clozure's docs work for SBCL? 03:54:11 no, but they work very well for Clozure CL; " It's Clozure CL I was thinking of". 03:54:47 -!- elurin [n=user@81.213.203.109] has quit [Remote closed the connection] 03:56:32 -!- ushdf [n=ushdf@syru217-183.syr.edu] has quit ["Lost terminal"] 03:59:23 borism [n=boris@195-50-199-64-dsl.krw.estpak.ee] has joined #lisp 04:01:14 -!- echo-area [n=user@nat/yahoo/x-7544ccafdf9ee778] has quit [Remote closed the connection] 04:06:11 -!- borism_ [n=boris@195-50-200-185-dsl.krw.estpak.ee] has quit [Read error: 145 (Connection timed out)] 04:10:53 -!- rpg [n=rpg@216.243.156.16.real-time.com] has quit [] 04:10:57 tritchey_ [n=tritchey@c-98-226-113-211.hsd1.in.comcast.net] has joined #lisp 04:13:36 hello 04:14:02 how can i put common lisp code to my web site in a way it respects the indentation? 04:15:01 -!- ivan4th [n=ivan4th@212.1.228.48] has quit [Remote closed the connection] 04:20:11 -!- photon [n=photon@unaffiliated/photon] has quit [Read error: 110 (Connection timed out)] 04:20:53 photon2 [n=photon@unaffiliated/photon] has joined #lisp 04:22:39 -!- tritchey [n=tritchey@c-98-226-113-211.hsd1.in.comcast.net] has quit [Read error: 110 (Connection timed out)] 04:27:07 -!- Nshag [i=user@Mix-Orleans-106-3-209.w193-248.abo.wanadoo.fr] has quit ["Quitte"] 04:27:46 stassats` [n=stassats@ppp91-122-105-57.pppoe.avangarddsl.ru] has joined #lisp 04:28:48 wol [n=wol@c-98-207-129-115.hsd1.ca.comcast.net] has joined #lisp 04:30:06 evening all 04:31:33 V-ille [n=ville@dsl-olubrasgw1-febddf00-91.dhcp.inet.fi] has joined #lisp 04:36:15 ushdf [n=ushdf@syru217-183.syr.edu] has joined #lisp 04:41:02 -!- stassats [n=stassats@wikipedia/stassats] has quit [Read error: 104 (Connection reset by peer)] 04:42:50 xreyes__ [n=xreyes@8.84-48-175.nextgentel.com] has joined #lisp 04:44:55 -!- beach``` is now known as beach 04:45:26 Good morning. 04:46:44 -!- blah2345 [i=a@c-67-185-17-84.hsd1.wa.comcast.net] has quit [] 04:47:51 -!- tritchey_ [n=tritchey@c-98-226-113-211.hsd1.in.comcast.net] has quit [] 04:47:51 jso [n=user@host-154-148-107-208.midco.net] has joined #lisp 04:47:59 -!- xreyes_ [n=xreyes@8.84-48-175.nextgentel.com] has quit [Read error: 110 (Connection timed out)] 04:50:12 morning beach 04:53:36 jestocost [n=cmell@p4186-ipbf4108marunouchi.tokyo.ocn.ne.jp] has joined #lisp 04:55:23 -!- crod [n=cmell@p4186-ipbf4108marunouchi.tokyo.ocn.ne.jp] has quit [Read error: 110 (Connection timed out)] 04:55:58 -!- benny [n=benny@i577A0519.versanet.de] has quit [Read error: 104 (Connection reset by peer)] 04:56:27 -!- vcgomes is now known as vcgomes[away] 04:57:50 -!- abeaumont [n=abeaumon@30.pool85-49-182.dynamic.orange.es] has quit [Read error: 104 (Connection reset by peer)] 04:58:38 abeaumont [n=abeaumon@30.pool85-49-182.dynamic.orange.es] has joined #lisp 04:59:03 -!- tiesje [n=user@202.51.72.181] has quit [Read error: 113 (No route to host)] 05:02:58 tiesje [n=user@202.63.242.211] has joined #lisp 05:04:03 Arrgh. How do you walk a spaghetti tree? 05:05:13 Corporate entity org chart. Half the 400 entities have multiple related shareholders. 05:13:18 fusss [n=chatzill@pool-72-66-40-106.washdc.east.verizon.net] has joined #lisp 05:13:29 -!- photon2 [n=photon@unaffiliated/photon] has quit [Read error: 110 (Connection timed out)] 05:17:53 gonzojive_ [n=red@c-24-5-14-234.hsd1.ca.comcast.net] has joined #lisp 05:20:12 -!- jestocost [n=cmell@p4186-ipbf4108marunouchi.tokyo.ocn.ne.jp] has quit [Connection timed out] 05:20:31 -!- tsuru [n=user@c-69-245-36-64.hsd1.tn.comcast.net] has quit [Remote closed the connection] 05:20:37 jestocost [n=cmell@p4186-ipbf4108marunouchi.tokyo.ocn.ne.jp] has joined #lisp 05:22:57 quek [n=read_eva@router1.gpy1.ms246.net] has joined #lisp 05:23:45 xreyes_ [n=xreyes@8.84-48-175.nextgentel.com] has joined #lisp 05:24:04 Last year, for the programming-project course, I suggested two projects: a specialized editor for Lisp and "Stamp", a tag-based email reader. Nobody chose Stamp, but the editor group did a decent job. 05:24:25 did they? where is it? 05:24:48 hefner: Somewhere in my hierarchy I guess. I'll try to dig it up. 05:25:18 This year, I could suggest Stamp again, but I need at least one more. 05:27:00 hefner: It is on my computer at work. I'll find it later. 05:27:28 no hurry, I was only curious to glance through it 05:27:30 Also, it would be nice if I could plan it so that I can get them to produce some reusable code. It is always hard to make that happen. 05:28:53 Perhaps I should suggest the word processor that I have been using in various other student projects. 05:29:01 spacebat_ [n=akhasha@124-171-97-115.dyn.iinet.net.au] has joined #lisp 05:29:06 -!- AshyIsMe [n=User@202.176.4.21] has quit ["Leaving"] 05:29:39 -!- xreyes__ [n=xreyes@8.84-48-175.nextgentel.com] has quit [No route to host] 05:29:55 I was going to suggest a high performance 2d drawing and compositing library, but the former would be redundant several times over (modulo the high performance bit), the latter trivial 05:30:09 holycow [n=biteme@S01060016b6b53675.vf.shawcable.net] has joined #lisp 05:33:36 hefner: Yeah, that's not going to fly I think. The guy running the course is going to question the usefulness of re-implementing something that already exists. 05:38:09 I'm just thinking of your proposed framebuffer mcclim backend, and the possibility that taking over rendering in software might be the best (most reliable and consistent, versus opengl and xrender) approach to getting a medium supporting modern graphical effects (alpha blending, antialiasing, alpha blended gadgets even) 05:39:09 and I am, for reasons unknown, unfairly biased against cairo 05:39:11 hefner: Sure, yes. I might be able to turn that idea into a student project. I'll think about it. 05:39:16 x6j8x [n=x6j8x@77.21.43.186] has joined #lisp 05:39:32 (why I think about these things with no intention of working on them is another question) 05:39:39 hefner: We use Cairo through FFI, right? 05:39:54 we do 05:40:08 hefner: so the bias is not that unfair :) 05:40:08 hjpark [n=user@61.109.28.117] has joined #lisp 05:40:33 *fusss* currently enjoying http://www.discontinuity.info/~pkhuong/LightBot.swf 05:40:34 -!- quek [n=read_eva@router1.gpy1.ms246.net] has left #lisp 05:41:43 -!- spacebat [n=akhasha@124-171-105-231.dyn.iinet.net.au] has quit [Read error: 110 (Connection timed out)] 05:41:47 hefner: We already know how to render polygons, right? I can't remember the name of that library. Is it Common Lisp-Vectors? 05:42:04 oops, my abbrev kicked in. CL-Vectors. 05:42:14 that's the one 05:43:36 hefner: So an interesting idea could be to polygonalize shapes defined as cubic Bézier curves, and render them with CL-Vectors. 05:44:03 cl-vectors renders curves of some variety 05:44:44 hefner: ah, that's what I was just checking. 05:45:01 quek [n=read_eva@router1.gpy1.ms246.net] has joined #lisp 05:45:03 -!- arbscht [n=arbscht@unaffiliated/arbscht] has quit ["leaving"] 05:47:07 hefner: indeed, it can handle Bézier curves. 05:47:58 -!- cmm [n=cmm@bzq-79-182-117-83.red.bezeqint.net] has quit [Remote closed the connection] 05:48:11 cmm [n=cmm@bzq-79-182-117-83.red.bezeqint.net] has joined #lisp 05:48:53 hefner: Thanks for your input. I am not sure I can turn this idea into a rasonable student project, but I'll certainly give it a thought. 05:49:48 Bzek [n=SK_sj@mcc-dyn-31-225.kosnet.ru] has joined #lisp 05:56:03 -!- ebzzry [n=rmm@115.147.100.214] has quit [Remote closed the connection] 05:56:25 ebzzry [n=rmm@115.147.100.214] has joined #lisp 05:56:39 -!- tripwyre [n=tripwyre@117.193.162.10] has quit [] 05:57:14 -!- x6j8x [n=x6j8x@77.21.43.186] has quit [] 06:01:16 spacebat [n=akhasha@124-171-80-206.dyn.iinet.net.au] has joined #lisp 06:03:21 -!- peaches` [n=pstickne@c-71-236-177-238.hsd1.wa.comcast.net] has quit [Remote closed the connection] 06:03:59 arbscht [n=arbscht@unaffiliated/arbscht] has joined #lisp 06:04:22 oudeis [n=oudeis@p11811120.orange.net.il] has joined #lisp 06:06:04 -!- wol [n=wol@c-98-207-129-115.hsd1.ca.comcast.net] has quit [Remote closed the connection] 06:08:26 -!- quek is now known as yoshinori 06:12:24 -!- mch [n=bitsurge@S01060004e24ffe41.cg.shawcable.net] has quit [Client Quit] 06:12:25 -!- haiwei [n=haiwei@192.9.202.3] has quit ["Leaving."] 06:12:40 haiwei [n=haiwei@192.9.202.3] has joined #lisp 06:14:03 -!- spacebat_ [n=akhasha@124-171-97-115.dyn.iinet.net.au] has quit [Read error: 110 (Connection timed out)] 06:16:50 -!- haiwei [n=haiwei@192.9.202.3] has quit [Client Quit] 06:17:02 haiwei [n=haiwei@192.9.202.3] has joined #lisp 06:21:02 hefner: the documentation for CL-Vectors indicates that the path library needs work. 06:22:08 hefner: http://projects.tuxee.net/cl-vectors/section-limitations#limitations 06:23:29 harshrc [n=harshrc@c-65-96-221-139.hsd1.ma.comcast.net] has joined #lisp 06:23:33 that's humility. If I wrote cl-vectors, I'd have implemented 15% of what's there, took screenshots, then patted myself on the back and called it a day. 06:33:20 echo-area [n=user@nat/yahoo/x-e1bb2192b6186648] has joined #lisp 06:35:03 -!- arbscht [n=arbscht@unaffiliated/arbscht] has quit ["leaving"] 06:43:55 then again, I've always been skeptical of the benefit of sophisticated vector graphics, at least from the perspective of wanting to implement everything from scratch and get the most for your time spent 06:56:44 TravisD [n=user@S01060016b6218db2.ed.shawcable.net] has joined #lisp 06:57:28 Does anyone feel like having an off-topic discussion about statistics? (This channel is usually pretty dead, yet I've had some of my most helpful non-lisp help here :)) 06:57:47 (more specifically, its about conditional independence, and proving a couple properties) 07:01:35 uhhhh 07:03:15 gigamonkey was doing some statistics recently, but this is way past his hours 07:03:20 -!- daniel_ is now known as daniel 07:03:33 brb cigarettes 07:05:02 Haha, shucks 07:05:17 I think its pretty elementary, but my stats background is lacking 07:08:50 -!- mathrick [n=mathrick@users177.kollegienet.dk] has quit [Read error: 110 (Connection timed out)] 07:15:02 best thing to do would be to get the formulas from wikipedia/mathworld and start coding :-P 07:15:22 basic numerical algorithms can be got from "Numerical Recipes", free online 07:16:54 Ah, I'm trying to give a proof, haha 07:17:51 x6j8x [n=x6j8x@tmo-100-34.customers.d1-online.com] has joined #lisp 07:20:46 mathrick [n=mathrick@wireless.its.sdu.dk] has joined #lisp 07:22:03 -!- oudeis [n=oudeis@p11811120.orange.net.il] has quit ["This computer has gone to sleep"] 07:23:02 hi guys 07:23:12 i've put up some regexp benchmarks 07:23:15 http://common-lisp.net/projects/cl-irregsexp/ 07:23:37 comparing cl-ppcre to perl, ruby, python and pcre in c++ 07:25:10 unsurprisingly i chose a case where cl-irregsexp is much faster 07:25:27 ebzzry_ [n=rmm@115.147.99.238] has joined #lisp 07:25:50 it is ten to twenty times faster than cl-ppcre 07:29:34 ecraven [n=nex@140.78.42.103] has joined #lisp 07:30:47 aerique [n=euqirea@xs2.xs4all.nl] has joined #lisp 07:35:31 -!- TravisD [n=user@S01060016b6218db2.ed.shawcable.net] has quit [Remote closed the connection] 07:35:35 yum 07:39:23 -!- hjpark [n=user@61.109.28.117] has quit [Read error: 110 (Connection timed out)] 07:42:08 -!- gonzojive_ [n=red@c-24-5-14-234.hsd1.ca.comcast.net] has quit [] 07:42:58 -!- ebzzry [n=rmm@115.147.100.214] has quit [Read error: 110 (Connection timed out)] 07:43:44 -!- amnesiac [n=amnesiac@p3m/member/Amnesiac] has quit ["Leaving"] 07:46:36 mikesch [n=axel@xdsl-78-34-236-155.netcologne.de] has joined #lisp 07:48:17 ebzzry__ [n=rmm@124.217.66.227] has joined #lisp 07:48:52 -!- jso [n=user@host-154-148-107-208.midco.net] has quit [Read error: 113 (No route to host)] 07:50:34 -!- cracki [n=cracki@sglty.kawo2.RWTH-Aachen.DE] has quit ["The funniest things in my life are truth and absurdity."] 07:54:55 jdz [n=jdz@85.254.248.11] has joined #lisp 07:56:28 -!- topo [n=topo@200.37.161.41] has left #lisp 07:56:53 woah! 07:57:17 i would've thought nobody would dare to even think of writing a regex library for cl after cl-ppcre :-P 07:58:45 ebzzry [n=rmm@124.217.85.39] has joined #lisp 08:01:05 knobo [n=user@cartman.nextra.no] has joined #lisp 08:01:05 well cl-ppcre was too slow and did not support vectors of unsigned-byte 8 08:03:59 -!- jestocost [n=cmell@p4186-ipbf4108marunouchi.tokyo.ocn.ne.jp] has quit ["Instain to the do way"] 08:04:04 jestocost [n=cmell@p4186-ipbf4108marunouchi.tokyo.ocn.ne.jp] has joined #lisp 08:04:10 -!- jestocost is now known as c|mell 08:04:18 -!- c|mell [n=cmell@p4186-ipbf4108marunouchi.tokyo.ocn.ne.jp] has quit [Client Quit] 08:04:24 c|mell [n=cmell@p4186-ipbf4108marunouchi.tokyo.ocn.ne.jp] has joined #lisp 08:04:30 I'm looking for a package with some skeletons for writing clos classes in emacs. I remember the guy who made it also made a vide as to show how it works. Anyone know what I'm talking about? 08:04:52 knobo: redshank 08:05:23 schme_: yes, thanx :) 08:07:21 sakana [n=itami@h219-110-182-179.catv02.itscom.jp] has joined #lisp 08:07:41 oudeis [n=oudeis@p11811120.orange.net.il] has joined #lisp 08:08:34 -!- ebzzry_ [n=rmm@115.147.99.238] has quit [Read error: 110 (Connection timed out)] 08:13:02 -!- tiesje [n=user@202.63.242.211] has quit [Remote closed the connection] 08:13:10 seelenquell [n=seelenqu@pD9E46D9C.dip.t-dialin.net] has joined #lisp 08:13:36 tiesje [n=user@202.63.242.211] has joined #lisp 08:15:14 bloody hell 08:15:27 that's awesome 08:16:38 -!- ebzzry__ [n=rmm@124.217.66.227] has quit [Read error: 110 (Connection timed out)] 08:21:06 mikesch_ [n=axel@xdsl-78-34-217-150.netcologne.de] has joined #lisp 08:21:28 nostoi [n=nostoi@90.Red-83-61-174.dynamicIP.rima-tde.net] has joined #lisp 08:21:43 -!- bob_f [n=bob_f@mail.phgroup.com] has quit [Remote closed the connection] 08:21:54 eevar2 [n=jalla@106.80-203-27.nextgentel.com] has joined #lisp 08:22:13 good morning 08:22:24 -!- mathrick [n=mathrick@wireless.its.sdu.dk] has quit [Read error: 104 (Connection reset by peer)] 08:22:27 g'day 08:22:39 mathrick [n=mathrick@wireless.its.sdu.dk] has joined #lisp 08:23:35 -!- mikesch [n=axel@xdsl-78-34-236-155.netcologne.de] has quit [Read error: 60 (Operation timed out)] 08:27:14 -!- nostoi [n=nostoi@90.Red-83-61-174.dynamicIP.rima-tde.net] has quit ["Verlassend"] 08:27:30 -!- sakana [n=itami@h219-110-182-179.catv02.itscom.jp] has left #lisp 08:30:27 -!- oudeis [n=oudeis@p11811120.orange.net.il] has quit [Read error: 110 (Connection timed out)] 08:32:25 splittist [n=splittis@213.235.9.122] has joined #lisp 08:32:28 morning 08:38:15 -!- nullwork_ [n=nullwork@c-24-245-23-122.hsd1.mn.comcast.net] has quit [Remote closed the connection] 08:38:22 nullwork_ [n=nullwork@c-24-245-23-122.hsd1.mn.comcast.net] has joined #lisp 08:39:23 -!- sohail [n=Sohail@unaffiliated/sohail] has quit [Read error: 60 (Operation timed out)] 08:40:37 morning. 08:41:52 arbscht [n=arbscht@unaffiliated/arbscht] has joined #lisp 08:42:37 morning 08:45:10 -!- tiesje [n=user@202.63.242.211] has quit [Remote closed the connection] 08:45:49 tiesje [n=user@202.63.242.211] has joined #lisp 08:46:20 -!- mikesch_ [n=axel@xdsl-78-34-217-150.netcologne.de] has quit [] 08:59:47 -!- mathrick [n=mathrick@wireless.its.sdu.dk] has quit [Read error: 60 (Operation timed out)] 09:01:34 -!- harshrc [n=harshrc@c-65-96-221-139.hsd1.ma.comcast.net] has quit [Remote closed the connection] 09:02:27 VityokOrgUa [n=user@193.109.118.130] has joined #lisp 09:06:37 -!- BrianRice-mb [n=briantri@c-98-225-51-174.hsd1.wa.comcast.net] has quit [] 09:07:54 ManateeLazyCat [n=Andy@222.212.136.7] has joined #lisp 09:09:57 -!- xjrn [n=jim@c-69-181-213-99.hsd1.ca.comcast.net] has quit [Read error: 110 (Connection timed out)] 09:10:05 tcr [n=tcr@host145.natpool.mwn.de] has joined #lisp 09:11:08 Odin- [n=sbkhh@s121-302.gardur.hi.is] has joined #lisp 09:15:44 morn' 09:16:10 morning 09:17:30 -!- yangsx [n=yangsx@218.247.244.25] has quit [Read error: 60 (Operation timed out)] 09:18:04 mathrick [n=mathrick@wireless.its.sdu.dk] has joined #lisp 09:18:09 bob_f [n=bob_f@mail.phgroup.com] has joined #lisp 09:19:11 *_8david* shares the bias against cairo 09:19:22 -!- kuwabara [n=Kuwabara@cerbere.qosmos.fr] has quit [Remote closed the connection] 09:19:59 <_8david> that is, I love it when I've got to work with GTK+ anyway, so another bunch of FFI calls doesn't matter. 09:20:01 <_8david> But I don't think FFI to cairo is a good idea in the middle of clim-clx. 09:21:21 <_8david> (OTOH, sometimes the native implementation is too good to not use it. For example, clim-truetype's output is still way too fuzzy for my eyes, so I'm still using clim-freetype until hefner implements hinting.) 09:22:53 xjrn [n=jim@c-69-181-213-99.hsd1.ca.comcast.net] has joined #lisp 09:24:41 that's quite a doubletake-inducing use of "native" :) 09:25:55 sometimes when I'm inspecting a long list with slime-inspector, I want to inspect several of the objects in the list. So I press "enter" on the first object, then "l" to go back, then "tab", then "enter", and so on. It would be nice to be able to do that in one click. 09:26:08 fe[nl]ix [n=algidus@88-149-211-72.dynamic.ngi.it] has joined #lisp 09:26:12 I tried to make a keyboard macro, but it did not work as expected. 09:26:24 hello 09:26:24 or, it worked, but only sometimes. 09:27:15 kib2 [n=chatzill@bd137-1-82-228-159-28.fbx.proxad.net] has joined #lisp 09:27:19 dash_ [n=dash@dslb-084-057-007-110.pools.arcor-ip.net] has joined #lisp 09:27:52 And, also if the list is log, it takes a lot of time, as emacs has to handle quite a lot of data. Would be better to have a short cut through the lisp. 09:28:43 lacedaemon [n=algidus@88-149-210-249.dynamic.ngi.it] has joined #lisp 09:29:08 It would be nice if you could have multiple inspector buffers open. I worked on it, but I didn't finish my work. 09:29:12 -!- fe[nl]ix [n=algidus@88-149-211-72.dynamic.ngi.it] has quit [Nick collision from services.] 09:29:14 -!- lacedaemon is now known as fe[nl]ix 09:29:37 You could then have a command which references an inspector link by opening a new inspector buffer. 09:30:33 mishok13 [n=gdmfsob@dm.sonopia.com] has joined #lisp 09:36:57 edon [n=edon@82.114.94.2] has joined #lisp 09:37:19 -!- dash__ [n=dash@dslb-088-065-143-143.pools.arcor-ip.net] has quit [Read error: 110 (Connection timed out)] 09:37:52 -!- edon [n=edon@albalug/edon] has quit [Read error: 104 (Connection reset by peer)] 09:38:42 -!- Krystof [n=csr21@84-51-132-95.christ977.adsl.metronet.co.uk] has quit [Read error: 113 (No route to host)] 09:43:18 -!- holycow [n=biteme@S01060016b6b53675.vf.shawcable.net] has quit [Read error: 104 (Connection reset by peer)] 09:44:25 ejs [n=eugen@nat.ironport.com] has joined #lisp 09:45:17 -!- froog_ is now known as froog 09:45:18 holycow [n=biteme@S01060016b6b53675.vf.shawcable.net] has joined #lisp 09:50:28 mejja [n=user@c-7d2472d5.023-82-73746f38.cust.bredbandsbolaget.se] has joined #lisp 09:51:18 -!- Odin- [n=sbkhh@s121-302.gardur.hi.is] has quit [] 09:52:32 MS's idea of .NET/Win32 interoperability tutorial. Should I laugh or cry ... 09:52:36 http://msdn.microsoft.com/en-us/library/ms744829.aspx 09:53:36 that's just one C++ shared library calling another C++ shared library (what .NET is, anyway) 09:54:02 pmatos [n=pmatos@pocm06r.ecs.soton.ac.uk] has joined #lisp 09:54:31 you should feel ashamed to go off topic.. 09:54:32 -!- jao [n=user@72.Red-79-155-244.staticIP.rima-tde.net] has quit [Read error: 110 (Connection timed out)] 09:56:46 -!- pchrist [n=spirit@gentoo/developer/pchrist] has quit [Read error: 110 (Connection timed out)] 09:57:04 ramkrsna [n=ramkrsna@unaffiliated/ramkrsna] has joined #lisp 09:57:18 sorry, almost went to the dark side ... 09:58:11 C++? 09:58:15 or MS? 09:58:29 C++ 09:58:45 i'm scavenging the net for what "modern gui" looks like 09:59:00 -!- pmatos [n=pmatos@pocm06r.ecs.soton.ac.uk] has quit ["Leaving"] 09:59:01 ach so 09:59:14 lispbuilder convinced me i can call win32 api from clisp to make gui 09:59:31 i'm also contemplating some kind of "native LTK" 09:59:36 LTK api, implemented natively 10:00:05 there's http://okmij.org/ftp/cpp-digest/Functional-Cpp.html 10:00:06 how much of what i need to do will be calls to FFI and how much self drawn 10:00:52 merlincorey: i don't wanna write C++ code, just FFI it 10:01:02 fusss: I know I just find it quirky 10:01:10 cute one would say 10:01:25 fusss: see also jack unrue's project 10:01:29 i think a minimum gui toolkit would have something like this: http://www.eclipse.org/swt/widgets/ 10:01:49 froog_ [n=david@87.192.28.247] has joined #lisp 10:02:08 Graphics-forms is the name of it 10:02:33 regarding Lispbuilder: why choose that over cl-ppcre? 10:02:57 (and is lispbuilder actively developed?) 10:02:59 benny [n=benny@i577A0519.versanet.de] has joined #lisp 10:03:17 Win32-ABI-compliant C++ code might be sort-of FFI'able, as long as it doesn't use templates 10:03:41 but that doesn't matter, because thankfully the Win32 API proper is C-level 10:04:00 ffi c++? without flattening it? 10:04:18 (or hardcode the win32-abi and de-munge C++ names? the horror) 10:04:38 nikodemus [n=nikodemu@cs181229041.pp.htv.fi] has joined #lisp 10:04:40 Xach: he is on to something, gotta see screenshots first .. googling 10:04:52 tic: every C++ compiler on Win32 (all 1 of them, realistically) uses the same mangling, and the mangling is pretty stable 10:05:12 cmm, *nod* 10:05:25 still, no need 10:05:40 tic: why choose lispbuilder over cl-ppcre? you can't choose an umbrella graphics programming library over a regular expressions library :-D 10:05:46 maybe if you want to do C++ stuff, cmm? 10:05:56 fusss, lispbuilder-regex.... 10:06:10 you don't. 10:06:30 it also has networking stuff, which i think calls out to SDL_net. usock and iolib is fine by me. 10:06:43 gotta consolidate these stuff 10:07:23 right, so lispbuilder-regex is more-or-less deprecated? 10:08:59 -!- echo-area [n=user@nat/yahoo/x-e1bb2192b6186648] has quit [Read error: 104 (Connection reset by peer)] 10:09:35 dunno, someone is working on it i guess 10:10:16 apparently it is faster in a use case? 10:10:17 mikesch [n=axel@static-87-79-66-80.netcologne.de] has joined #lisp 10:10:19 was that that guy earlier? 10:11:23 different regex lib, that is cl-irregsexp 10:11:33 -!- tcr [n=tcr@host145.natpool.mwn.de] has quit ["Leaving."] 10:15:09 -!- O_4 [n=souchan@ip-118-90-42-91.xdsl.xnet.co.nz] has quit ["Come alive!"] 10:17:34 reaver__ [n=reaver@212.88.117.162] has joined #lisp 10:22:19 -!- dv_ [n=dv@85-127-102-97.dynamic.xdsl-line.inode.at] has quit ["Gone"] 10:22:38 -!- froog [n=david@87.192.28.247] has quit [Read error: 110 (Connection timed out)] 10:25:02 -!- jtoy [n=jtoy@58.63.90.61] has quit [] 10:29:15 jao [n=user@74.Red-80-24-4.staticIP.rima-tde.net] has joined #lisp 10:30:06 pierre_thierry [n=pierre@black.delerce.fr] has joined #lisp 10:32:11 -!- froog_ is now known as froog 10:32:16 rindolf [n=shlomi@bzq-219-139-216.static.bezeqint.net] has joined #lisp 10:34:36 -!- holycow [n=biteme@S01060016b6b53675.vf.shawcable.net] has quit [Read error: 104 (Connection reset by peer)] 10:37:02 -!- mishok13 [n=gdmfsob@dm.sonopia.com] has quit [Read error: 104 (Connection reset by peer)] 10:37:08 gdmfsob [n=gdmfsob@dm.sonopia.com] has joined #lisp 10:41:08 -!- rindolf [n=shlomi@bzq-219-139-216.static.bezeqint.net] has quit ["Yay! I'm a Llama again!"] 10:43:43 -!- xan_ [n=xan@cs78225040.pp.htv.fi] has quit [Read error: 110 (Connection timed out)] 10:44:33 xan_ [n=xan@cs78225040.pp.htv.fi] has joined #lisp 10:53:40 mega1 [n=mega@pool-016d1.externet.hu] has joined #lisp 10:55:46 kuwabara [n=Kuwabara@cerbere.qosmos.fr] has joined #lisp 10:55:50 holycow [n=biteme@S01060016b6b53675.vf.shawcable.net] has joined #lisp 10:56:24 -!- kuwabara [n=Kuwabara@cerbere.qosmos.fr] has quit [Remote closed the connection] 10:56:38 matimago [n=user@cac94-10-88-170-236-224.fbx.proxad.net] has joined #lisp 10:56:47 kuwabara [n=Kuwabara@cerbere.qosmos.eu] has joined #lisp 10:56:54 -!- kuwabara [n=Kuwabara@cerbere.qosmos.eu] has quit [Remote closed the connection] 10:57:21 silenius [n=jl@yian-ho03.nir.cronon.net] has joined #lisp 10:57:39 kuwabara [n=Kuwabara@cerbere.qosmos.eu] has joined #lisp 11:03:44 ths [n=ths@p549AC6B4.dip.t-dialin.net] has joined #lisp 11:07:05 -!- envi^office [i=envi@203.109.25.100] has quit ["Leaving"] 11:08:56 -!- ths_ [n=ths@X4d78.x.pppool.de] has quit [Read error: 110 (Connection timed out)] 11:09:38 Jacob_H [n=jacob@92.3.170.158] has joined #lisp 11:09:59 -!- sqvirt [n=sqvirt@c-24-16-244-102.hsd1.wa.comcast.net] has quit [Read error: 110 (Connection timed out)] 11:10:56 -!- nikodemus [n=nikodemu@cs181229041.pp.htv.fi] has quit ["This computer has gone to sleep"] 11:12:31 ltk:*generate-accessors* is commented out in the sources. wonder why? 11:14:25 oudeis [n=oudeis@bzq-79-176-18-217.red.bezeqint.net] has joined #lisp 11:17:15 ahhh, it uses a CONFIGURE to change slots, instead of (setf slot) 11:21:20 tst___ [n=Tim@p4FD2C829.dip.t-dialin.net] has joined #lisp 11:30:35 valvola [n=fabiovio@salugit.unile.it] has joined #lisp 11:31:32 nikodemus [n=nikodemu@cs27013130.pp.htv.fi] has joined #lisp 11:32:50 -!- valvola [n=fabiovio@salugit.unile.it] has quit [Remote closed the connection] 11:36:47 cracki [n=cracki@sglty.kawo2.RWTH-Aachen.DE] has joined #lisp 11:37:53 kiuma [n=kiuma@83-103-59-156.ip.fastwebnet.it] has joined #lisp 11:38:17 hello lispers 11:42:27 NuMaStresa [n=andreiha@unaffiliated/numastresa] has joined #lisp 11:42:50 vy [n=user@213.139.194.86] has joined #lisp 11:44:30 tcr [n=tcr@host145.natpool.mwn.de] has joined #lisp 11:47:19 Odin- [n=sbkhh@s121-302.gardur.hi.is] has joined #lisp 11:49:09 ephcon [n=ephcon@student165-162.hampshire.edu] has joined #lisp 11:51:49 -!- jdz [n=jdz@85.254.248.11] has quit [Read error: 110 (Connection timed out)] 11:54:46 hello kiuma 11:54:57 hi 11:55:38 *fusss* note to self: before you start debugging, did you forget the second * on your *special-variable? 11:56:00 yowza kiuma 11:56:55 I'm just starting with lisp development with linux ... 11:57:10 any good ide recomandation ? 11:57:20 NuMaStresa: slime 11:57:20 I use vim for all my text editing .. 11:57:23 -!- gjvc_ [n=gjvc@chardonnay.extremis.net] has quit ["leaving"] 11:57:31 NuMaStresa: most people here use emacs with slime 11:58:04 fusss, i do. a lot. must've picked up from perl or something 11:58:05 some of us use slime with emacs 11:58:07 thanks ... I don't think I wanna learn emacs to do lisp ... thanks for the tips 11:58:10 NuMaStresa: Writing program is not the same as text editing. 11:58:15 -!- tiesje [n=user@202.63.242.211] has quit [Remote closed the connection] 11:58:18 NuMaStresa: there is a small group of people who refuse emacs and thus can't leverage all the goodness in emacs and slime. ask tic. 11:58:20 tiesje [n=user@202.63.242.211] has joined #lisp 11:58:41 NuMaStresa: There's also Limp which is a Vim plugin. 11:58:44 NuMaStresa, then start wanting :) 11:58:45 NuMaStresa: emacs isn't hard 11:59:07 NotePad can also be used to write code, but you don't wanna do that 11:59:09 unless you have 5+ years of experience on your belt, but you don't 11:59:13 I know stassats` ... just that I don't need emacs when I already use vim ... 11:59:41 NuMaStresa: time to switch 11:59:43 yes you do :) unless you decide not to learn lisp. or use some commercial vendor's ide 11:59:44 NuMaStresa: ALL commercial Lisps come with builtin IDEs 11:59:45 NuMaStresa: you are assuming that vim can do what slime and emacs do, which is not the case 11:59:46 another option is using a trial version of one of the commercial implementations. they come with IDEs (at least on Windows) 11:59:46 :)) 12:00:31 but commercial IDEs look like native Windows software :-( and they have working drag-and-drop and cut-and paste :-( and they use the mouse :-( 12:00:32 I want to develop under linux ... 12:00:46 NuMaStresa: lispworks comes with a motif based ide 12:00:51 ok ... thanks for the tips ... 12:00:55 yes, even under Linux, they have IDEs 12:00:58 I'll check it on google ... thanks 12:01:08 which is emacs-based (lispworks ide) 12:01:21 even, emacs-like 12:01:24 NuMaStresa: I'd give Limp a try first. There's also Cusp which is a plugin for Eclipse. 12:01:27 stassats`: hey, it does not say "i'm emacs", so it is okay!!eleven 12:01:31 -!- holycow [n=biteme@S01060016b6b53675.vf.shawcable.net] has quit [Read error: 104 (Connection reset by peer)] 12:02:03 under windows at least, LW IDE can be made to act dumb and un-emacsy 12:02:03 I used limp ... for a while .. not to much ... I'm using lisp for some school projects ... although from what I understood it's good to learn programming with lisp ... 12:02:14 I'm not such a good programmer ... just started with python ... 12:02:34 I don't use/want to use windows ... emacs ... although I can try (again) emacs ... 12:02:35 :) 12:03:04 *weirdo* <- an ex-vi user 12:03:05 NuMaStresa: if you are serious about lisp programming, try switching to emacs 12:03:06 NuMaStresa: You may want to read Structure and Interpretation of Computer Programs. It uses a subset of scheme, not Common Lisp. But it's a very good foundation for becomming a programmer. 12:03:11 *NuMaStresa* downloading lisp works ... 12:03:21 fusss: well, emacs can be made to act dumb and un-emacsy 12:03:30 intrados1 [n=intrados@rdrt-164-107-204-170.resnet.ohio-state.edu] has joined #lisp 12:03:39 *stassats`* ex-vim user 12:03:43 NuMaStresa: i am a wizard level vi user, yet i never regret having switched to emacs some 20 years ago. emacs is great for c++ programming, too. 12:03:51 stassats`: believe me, Windows reset my $HOME earlier and emacs started without my .emacs .. ewwww 12:04:11 -!- intrados [n=intrados@rdrt-164-107-204-170.resnet.ohio-state.edu] has quit [Read error: 104 (Connection reset by peer)] 12:04:48 did it go like this? "ay'a 12:04:52 I understand ... I don't know ... learning a new text editor ... I guess I'll give it a try ... 12:04:55 can't even remember the basics :( 12:04:57 thanks for the tips .. 12:05:24 to copy into a selection named "a" from the mark called "a" 12:05:40 NuMaStresa: http://www.xach.com/img/lisp-and-vim.png 12:05:44 NuMaStresa: Slime has a great ascii animation when it starts up. 12:05:46 i actually taught my ma' vi once 12:06:31 fusss: "i don't believe in windows" 12:06:40 :) 12:06:49 windows is a religion? 12:06:59 actually a cult ... :) 12:07:18 stassats`: that's because you flunked out of the sbcl win32 porting project ;-) 12:07:45 "i don't believe in the postman" 12:07:53 should I download lisp works or cancel and use emacs ? ... 12:08:11 NuMaStresa: this is a free countr^Wchannel 12:08:19 lispworks is a trial and afaik there's no crack available :( 12:08:35 NuMaStresa: lispworks is nice, yet if you plan trying to get support here, a free implementation will be better. 12:08:47 NuMaStresa: Considering your level of experience, I'd suggest reading through SICP, and using DrScheme. 12:08:54 +1 12:09:28 can I use drscheme for common lisp ? 12:09:34 it has first class continuations and turtle graphics 12:10:12 -!- pchrist|univ [n=spirit@gateway.hpc.cs.teiath.gr] has quit [Client Quit] 12:10:14 NuMaStresa: Not really. SICP will teach you the foundations of computing, not Common Lisp by any means. 12:10:29 ok 12:10:35 *NuMaStresa* just finished installing emacs ... 12:10:45 pchrist|univ [n=spirit@gateway.hpc.cs.teiath.gr] has joined #lisp 12:11:12 tcr: i believe that sicp is a great book for some, but not for everyone. i find it having a language that will put off people having a more pragmatic approach and are less interested in theory and "computer science" 12:12:14 H4ns: took me months to recover from contrived recursion 12:12:21 antoszka [n=antoszka@unaffiliated/antoszka] has joined #lisp 12:12:22 can I hang around ... and ask questions ... how are newbies welcomed ? ... 12:12:38 good questions are welcome 12:12:42 ok 12:12:59 It depends on the newbie. Act humble, show that you've listened and learned, and you will be cherished. 12:13:00 NuMaStresa: newbies are welcome unless they can't conceal that they have a course assignment and their deadline is slipping 12:13:05 I already read "How to ask smart questions" so I guess it will be ok ... 12:13:46 NuMaStresa: Your level of expertise is not that relevant as long as you exhibit intelligent behaviour. 12:14:04 Although I do have a course, and have to do homeworks ... I want to learn lisp as a foundation for my programming skills ... I just got started with programming and I heard that lisp is a great language ... 12:14:05 NuMaStresa: And good spelling and punctuation are preferred. (: 12:14:12 sorry splittist 12:14:32 NuMaStresa: how did you get into programming (and linux for that matter) 12:14:37 I will keep that in mind ... I am not a native speaker but I guess I can stop using ' and ` ... 12:15:04 -!- Jasko [n=tjasko@c-98-235-21-22.hsd1.pa.comcast.net] has quit ["Leaving"] 12:15:24 jewel [n=jewel@dsl-242-184-103.telkomadsl.co.za] has joined #lisp 12:15:50 ok, i keep forgetting not every linux user is a programmer 12:16:03 H4ns1 [n=Hans@p57BB9CBA.dip0.t-ipconnect.de] has joined #lisp 12:16:03 NuMaStresa: I didn't mean to criticise you at all. You are doing well (except for an overuse of ellipses, of which I am also overfond). 12:16:12 I started with linux because I was getting tired about windows ... and I downloaded, installed, played with a few distributions and finaly used it for more than 3 years without having windows ... 12:16:36 -!- H4ns [n=Hans@p57BBABCD.dip0.t-ipconnect.de] has quit [Nick collision from services.] 12:16:37 -!- H4ns1 is now known as H4ns 12:16:57 programming ... I'm currently doing a faculty (computer science) ... although I know the syntax (lol) of a lot of programming languages I find it hard when I have to "write" something from my head 12:17:14 CMUCL has a builtin emacs like editor and is very "integrated". you will have plenty of fun poking around and writing Motif GUI apps. 12:17:39 -!- jewel [n=jewel@dsl-242-184-103.telkomadsl.co.za] has quit [Remote closed the connection] 12:17:43 "fun" and "Motif" in the same sentence? 12:17:54 cmm: i swear I did 12:18:10 no accounting for some tastes 12:18:16 but then, I came from C++/COM/MFC background when i found Lisp. 12:18:37 augh, MFC! 12:18:39 does somebody actually use hemlock? 12:18:41 yes, that would explain it :) 12:19:08 CMUCL Hemlock doesn't really inspire confidence in me, as so far as I can tell CMUCL hacker's don't use it for their daily CMUCL hacking... 12:19:20 hackers, even 12:19:29 *nikodemus* thwaps himself 12:19:33 -!- tiesje [n=user@202.63.242.211] has quit [Read error: 113 (No route to host)] 12:19:46 stassats`: LW does, sort of 12:19:55 nikodemus: doesn't matter. the whole thing, and the copious manuals, give CMUCL that vintage but "integrated" feel 12:20:14 there's also the portable hemlock project, but I have no idea how live that is 12:20:27 how can I learn/get the algorithmic thinking ? 12:20:35 -!- araujo [n=araujo@gentoo/developer/araujo] has quit [Connection timed out] 12:20:35 practise 12:20:50 go through the basic but essential texts and do all the exercises 12:21:16 knuth? 12:21:24 naah 12:22:35 the SIMPLEST way to test if you will enjoy programming might be this flash game 12:22:37 http://www.discontinuity.info/~pkhuong/LightBot.swf 12:22:46 -!- x6j8x [n=x6j8x@tmo-100-34.customers.d1-online.com] has quit [Read error: 60 (Operation timed out)] 12:23:01 jdz [n=jdz@85.254.248.11] has joined #lisp 12:23:04 you get to use up to two functions later 12:23:20 -!- haiwei [n=haiwei@192.9.202.3] has quit [Remote closed the connection] 12:23:37 but at the highest levels you will have to look at sacked blocks and figure out if any pattern emerges, then factor out 12:23:42 -!- tcr [n=tcr@host145.natpool.mwn.de] has left #lisp 12:23:51 x6j8x [n=x6j8x@tmo-100-35.customers.d1-online.com] has joined #lisp 12:24:04 hint: write down the moves before you make them (L for left, R for right, J for jump and @ for light :-) 12:24:25 sorry, didn't mean to flood. brb. 12:24:55 fusss: nice! 12:24:59 jtoy [n=jtoy@58.63.171.97] has joined #lisp 12:30:07 *NuMaStresa* still thinking if I should invest time in learning emacs ... 12:30:27 NuMaStresa: Why? 12:30:50 I already use eclipse,vim,aptana,netbeans and other ide's/text editors ... 12:31:01 Oh. 12:31:13 Well I'd recommend some emacsen + slime for lisp, for sure. 12:31:14 NuMaStresa: can you use chopsticks? 12:31:24 But you don't need to learn much of emacs for that. 12:31:41 NuMaStresa: did learning to use them mean you're knife and fork skills deteriorated? 12:31:52 s/you're/your/ Sorry. 12:31:53 :) 12:32:51 hmm, limp doesn't look *that* bad 12:32:58 I don't think that this is what I need ... another text editor ... but If you say it will be good for the "long run" 12:33:08 I used limp ... It's ok 12:33:44 Limp? 12:33:53 *tic* should add a highlight on htat. :) 12:33:53 NuMaStresa: please do not repeat this pondering over and over. you'll get no new insights unless you give emacs and slime a try. 12:33:59 that vim ide 12:34:08 NuMaStresa: you will need to consolidate your tools. programmers hardly use more than one text editor. 12:34:13 weirdo, yes, I'm the author. :) 12:34:16 doesn't appear to have xref or autodoc, but it's better than nothing i guess 12:34:18 learn one and learn it well. 12:34:34 fdr- [n=fdr@76-191-209-5.dsl.dynamic.sonic.net] has joined #lisp 12:34:35 weirdo, http://common-lips.net/project/limp - I hope to improve on it. 12:34:46 sorry, i'm not a vi user and i don't know vim-script 12:34:50 thanks fusss ... I guess I'll stick with vim / eclipse 12:35:04 the real problem is lack of employer's acceptance :( but i'm rambling 12:35:12 s/'s/s'/ 12:35:48 fusss, you mean one for each language I hope 12:36:02 they'd sooner accept lua than any lisp for projects 12:36:03 kiuma: nonsense. 12:36:11 bakkdoor [n=bakkdoor@xdsl-92-252-41-80.dip.osnanet.de] has joined #lisp 12:36:44 kiuma: no, i ONLY use one text editor. Emacs. for everything. 12:36:44 -!- bakkdoor [n=bakkdoor@xdsl-92-252-41-80.dip.osnanet.de] has left #lisp 12:37:06 emacs' good for perl, too. sepia attempts to emulate slime, better than any vim plugin for perl 12:37:10 i might double click on a *.txt file and something might open it by default, but never for my own new writing. 12:37:29 too bad gnus is such a hog 12:37:44 fusss, thought I love emacs, I prefer a full bloated ide as many others for java projects 12:37:46 Thunderbird for mail/news/rss 12:38:03 well, i don't do java 12:38:10 -!- a-s [n=user@92.80.109.191] has quit [Remote closed the connection] 12:38:25 H4ns, yes it makes 12:38:48 attila_lendvai [n=ati@business-89-132-61-222.business.broadband.hu] has joined #lisp 12:38:55 fusss, you generalized with 'programmers' :) 12:39:43 if you're good with Eclipse, i think you ought to customize it fully and use for other hacking projects 12:40:32 i go through CL, perl, C, two assembly syntaxes, perl, shell, info, text, html, javascript, actionscipt and CSS .. no editor switching there 12:40:58 not to mention calc and org-mode 12:41:01 this seems the hammer syndrome to me 12:41:36 kiuma: no, text editing is pretty universal 12:42:15 -!- sellout [n=greg@c-24-128-50-176.hsd1.ma.comcast.net] has quit [] 12:42:38 you should see my work environment 12:42:42 H4ns, for writing text `ed` would suffice, but would you use it ? 12:43:05 kiuma: ed is not a good text editor, but emacs is 12:43:28 kiuma: if ed was aware of the syntactic/semantic structure of the documents i was editing and giving good and flexible views of them 12:43:35 tic, how about interfacing with existing swank code? it has compatibility code for autodoc, xref and all the other goodness 12:43:36 not as good as netbeans, for html/css 12:43:49 (Gosh, GNU Emacs is aware of all that, fusss?) 12:43:58 H4ns, and js too 12:44:25 Riastradh: not the semantics, not to the extent i would like, but CL it's pretty kick ass 12:44:30 slime that is 12:44:42 kiuma: that might be true, my argument is: if you know a text editor really well, you don't need so much special-purpose hand-holding that any ide may provide 12:45:19 fusss, aha, so you're talking about state of the Lisp image with which you're interacting, not the structure of documents. 12:45:27 H4ns, yes it depends from a lot of factors indeed 12:45:47 kiuma: i am not trying to convince you that this is a universal argument. i just happen to switch languages and environments a lot, and no special-purpose ide would be unintrusive enough to support the seamless switching that i have with emacs. 12:46:14 and this it true too 12:46:31 Riastradh: someone gave me the enterprise version of Visual Studio.NET with the much tauted "intellisense" ... ummm 12:46:50 how's climacs? 12:46:52 -!- nikodemus [n=nikodemu@cs27013130.pp.htv.fi] has quit [Read error: 110 (Connection timed out)] 12:46:52 :) 12:47:02 goatee ftw! 12:47:29 alright, gotta finish that ltk demo .. 12:47:53 H4ns: seen box2dflash? 12:48:09 fusss: no, what is it? 12:48:23 only the best flash collision detection library :-) 12:49:00 weirdo, swank: that is the plan, indeed. 12:49:02 hjpark [n=user@61.109.28.117] has joined #lisp 12:49:04 actually, entire physics engine 12:49:08 nikodemus [n=nikodemu@cs27013130.pp.htv.fi] has joined #lisp 12:49:30 weirdo, the problem is that to get asynch comms w/ vim you need netbeans so a proxy is required, alas. 12:49:40 but it /should/ be solvable, or at least I hope so. 12:49:42 fusss: nice, but how am i interested in that? 12:49:43 envi^home [n=envi@220.121.234.156] has joined #lisp 12:49:46 *afk* 12:50:05 -!- weirdo [i=sthalik@c144-107.icpnet.pl] has left #lisp 12:50:07 weirdo [i=sthalik@c144-107.icpnet.pl] has joined #lisp 12:50:12 fusss: (i mean, i find it really cute and all, but i don't remember having shopped for a flash physics engine) 12:50:13 H4ns: oh, you're not the one making flash games? must be ahaas 12:50:25 tic, someone once did some work for asynch 12:50:30 there is another cl-flasher here 12:50:36 that ecl-vim stuff 12:50:53 fusss: i'm not even a cl flasher 12:51:01 except that it broke after a few seconds busy-looping in EBADFD 12:51:01 weirdo, the network part of the patch was buggy. could sometimes crash Vim. 12:51:07 but it worked just fine for these few second 12:51:08 s 12:51:16 alright Hans, cheers mate! 12:51:23 fusss: cheers! 12:51:44 weirdo, I've upgraded the vim-ecl patch for 7.2, it's on http://groups.google.com/group/vim_dev/web/vim-patches 12:52:16 weirdo, if you have more info, please join #limp or send a private message. I'm afk for a couple of hours. Thanks! 12:52:39 tic, it was about a year ago, got the link from lisp gardeners 12:53:11 tic, you probably found out that one long time ago. didn't notice at first you were the author. small world, isn't it :-) 12:53:40 -!- yoshinori [n=read_eva@router1.gpy1.ms246.net] has quit [Remote closed the connection] 12:54:08 yoshinori [n=read_eva@router1.gpy1.ms246.net] has joined #lisp 12:54:18 -!- vcgomes[away] is now known as vcgomes 12:55:27 -!- brickhazel [n=brickhaz@63.144.132.78] has quit ["Leaving"] 12:56:24 jpcooper [n=justin@unaffiliated/jpcooper] has joined #lisp 12:59:23 araujo [n=araujo@gentoo/developer/araujo] has joined #lisp 12:59:38 later folks 12:59:39 -!- fusss [n=chatzill@pool-72-66-40-106.washdc.east.verizon.net] has quit ["ChatZilla 0.9.83 [Firefox 2.0.0.17/2008082909]"] 13:00:43 brickhazel [n=brickhaz@63.144.132.78] has joined #lisp 13:01:05 -!- brickhazel [n=brickhaz@63.144.132.78] has quit [SendQ exceeded] 13:01:29 brickhazel [n=brickhaz@63.144.132.78] has joined #lisp 13:09:10 http://uncov.com/thanks-for-fucking-my-future is one of the odder sbcl posts turned up by my daily blog search. 13:11:04 -!- yoshinori [n=read_eva@router1.gpy1.ms246.net] has left #lisp 13:11:40 yoshinori [n=read_eva@router1.gpy1.ms246.net] has joined #lisp 13:11:43 "I Would Gladly Take It Up The Ass From SBCL" 13:12:16 I'll be sure to put *that* endorsement right on the front of the web site. 13:12:17 does that blog post have any actual meaning or is it just a really good turing test program? 13:12:31 I can't discern any real meaning. 13:12:56 good, not only me then 13:13:43 *stassats`* started to read it three times and didn't understand anything 13:14:25 I tried twice and remembered that I have real work to do. 13:15:52 -!- StasNev [n=wizard@90.150.125.68] has quit [Read error: 110 (Connection timed out)] 13:16:45 Very motivational, I guess. 13:16:46 StasNev [n=wizard@94.50.175.122] has joined #lisp 13:17:01 well, at least the guy (entity) has a fuckable future! 13:17:52 duck1123 [i=duck@adsl-75-46-28-24.dsl.sfldmi.sbcglobal.net] has joined #lisp 13:18:21 chandler: that quote should be part of the startup message. our answer to the clisp menorah 13:21:36 -!- ace4016 [i=ace4016@cpe-76-168-248-118.socal.res.rr.com] has quit ["night"] 13:22:17 Jasko [n=tjasko@209.74.44.225] has joined #lisp 13:23:28 jgracin [n=jgracin@82.193.208.195] has joined #lisp 13:24:00 dlowe [n=dlowe@ita4fw1.itasoftware.com] has joined #lisp 13:24:29 Riastradh: Hah, real work, so you get paid for posting on #lisp? 13:25:20 "posting"? 13:25:50 chandler: Sorry, jeebuz, chatting. (Is that better?) 13:27:17 xan_, maybe an AI written in lisp wrote it 13:28:03 I read it and couldn't tell if the guy actually liked CL or was just doing some pathetic CS version of name dropping. 13:28:08 sellout [n=greg@c-71-232-4-167.hsd1.ma.comcast.net] has joined #lisp 13:28:31 Xach: uncov is pretty odd in general 13:28:38 -!- NuMaStresa [n=andreiha@unaffiliated/numastresa] has quit [Read error: 110 (Connection timed out)] 13:28:41 -!- Eno_ [n=anon@fl-67-76-215-49.sta.embarqhsd.net] has quit [Read error: 104 (Connection reset by peer)] 13:28:42 that post's unusually incomprehensible, though 13:28:47 think it's a guest contribution 13:29:44 sellout- [n=greg@guest-fw.dc4.itasoftware.com] has joined #lisp 13:29:56 jpcooper, that or a really confused/'under influence' human, yep 13:29:58 Eno_ [n=anon@fl-67-76-215-49.sta.embarqhsd.net] has joined #lisp 13:31:14 florist [i=shreyas@unaffiliated/florist] has joined #lisp 13:31:18 -!- florist [i=shreyas@unaffiliated/florist] has quit [Client Quit] 13:32:00 Anyway, given his(/her/its) stated desire, maybe we can task him(/her/it) with managing the SF File Release System of DOOM, which just ate the binary I put in the dropbox because I deleted another (misnamed) file release 13:32:10 xan_: or just a college-student. 13:32:14 -!- locklace [i=locklace@gateway/tor/x-c3476c73b2410c66] has quit [Remote closed the connection] 13:32:33 matimago: same thing. 13:33:03 I assume SBCL's just on sourceforge due to sheer inertia at this point? 13:33:08 -!- l_a_m [n=lam@194.51.71.190] has quit ["Lost terminal"] 13:33:08 it really is an awful system 13:33:13 especially that file release thing 13:33:30 -!- c|mell [n=cmell@p4186-ipbf4108marunouchi.tokyo.ocn.ne.jp] has quit [Read error: 60 (Operation timed out)] 13:33:31 locklace [i=locklace@gateway/tor/x-7d1527f8e725fe4c] has joined #lisp 13:34:29 tritchey [n=tritchey@c-98-226-113-211.hsd1.in.comcast.net] has joined #lisp 13:34:30 Heh, that's always a tough call. At what point is the cost of transitioning to something new outweigh the cost of having to repeatedly do something to get it to work? 13:35:12 sf.net has mirrors. 13:35:30 I'm also impressed with the fact that a download link to a file that doesn't actually exist doesn't return a 404, instead promising that your download will start soon 13:35:36 -!- segv [n=mb@p4FC1D967.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)] 13:35:41 Are there mirrors for common-lisp.net? 13:35:44 segv_ [n=mb@p4FC1E0FC.dip.t-dialin.net] has joined #lisp 13:36:23 matimago: who said anything about common-lisp.net? 13:36:35 The name's nice ;-) 13:36:36 tmh: what alternative would you suggest? 13:37:19 novaburst [i=nova@sourcemage/mage/novaburst] has joined #lisp 13:37:35 chandler: I wouldn't suggest one, I haven't spent the time looking. 13:39:28 http://en.wikipedia.org/wiki/Comparison_of_free_software_hosting_facilities 13:39:58 My point was that it shouldn't be assumed that we're comparing SF and common-lisp.net. Although, yes, it does have a nice name. 13:40:09 haiwei [n=haiwei@61.149.75.2] has joined #lisp 13:42:26 tmh: I would like to move to a system where the SBCL web page could be hosted as an application (rather than a generated blob of HTML) so that the links on the download table are always up-to-date 13:42:42 -!- fe[nl]ix [n=algidus@88-149-210-249.dynamic.ngi.it] has quit [Read error: 60 (Operation timed out)] 13:43:05 common-lisp.net at least would work for that; I don't think anything based on the sourceforge/gforge system would. 13:44:29 another alternative would be to maintain our own server, but i don't think anyone wants the extra work that would generate 13:50:08 ths: the linux/ppc machine I was using is the old ubuntu LTS (6.04), so it's quite out of date. but as much as I'd like to believe that this is a kernel/glibc/gcc version issue, it seems more like randomness to me 13:50:22 cosmic rays, as Xof would say :-) 13:50:53 Beket [n=stathis@athedsl-4415404.home.otenet.gr] has joined #lisp 13:51:10 ths: also, binary release naming convention is sbcl----binary.tar.bz2 - seeing yours in the list made me realize I screwed up the sparc-linux binary name too 13:51:25 -!- sellout [n=greg@c-71-232-4-167.hsd1.ma.comcast.net] has quit [Connection timed out] 13:51:26 -!- mindCrime [n=chatzill@cpe-075-177-141-190.nc.res.rr.com] has quit ["ChatZilla 0.9.82.1-rdmsoft [XULRunner 1.9b5/2008043010]"] 13:52:51 Hun [n=Hun@p50993726.dip0.t-ipconnect.de] has joined #lisp 13:53:31 lichtblau [n=user@wallstrasse.knowledgetools.de] has joined #lisp 13:53:39 -!- ephcon [n=ephcon@student165-162.hampshire.edu] has quit [Read error: 104 (Connection reset by peer)] 13:54:10 ths: I downloaded, untarred, retarred, and uploaded it. thanks for building this! 13:54:12 -!- xreyes_ [n=xreyes@8.84-48-175.nextgentel.com] has quit [Read error: 60 (Operation timed out)] 13:56:10 -!- ManateeLazyCat [n=Andy@222.212.136.7] has quit [Remote closed the connection] 13:56:16 -!- b4|hraban [n=b4@0brg.xs4all.nl] has quit ["Leaving"] 13:56:37 I just mistakeningly typed a float in the REPL with a comma for the decimal separator, which is valid in some countries. Would it be that difficult to modify the reader to accept comma for the decimal separator? 13:56:55 tmh: yes. 13:57:03 tmh: the problem is that there's no standard hook to do that. 13:57:20 -!- sellout- is now known as sellout 13:57:22 rdd [n=user@c83-250-142-219.bredband.comhem.se] has joined #lisp 13:57:28 Okay, I don't need it, the mistake just peeked my curiousity. 13:57:33 tmh: if you want the patch to apply to existing reading/loading functions, you'll have to hook heavily the reader macros. 13:58:24 besiria [n=student@195.251.210.108] has joined #lisp 13:58:46 tmh: otherwise, you could use a portable reader such as http://darcs.informatimago.com/lisp/common-lisp/reader.lisp which has the needed hook. 13:58:47 fe[nl]ix [n=algidus@88-149-209-64.dynamic.ngi.it] has joined #lisp 13:59:19 c|mell [n=cmell@p3209-ipbf1802marunouchi.tokyo.ocn.ne.jp] has joined #lisp 14:00:40 chandler: Thing is, I've never seen such a problem on ppc, except for a short time when it was caused by a sourceful bug, and eventually fixed. 14:00:51 -!- lemonodor [n=lemonodo@adsl-76-214-15-172.dsl.lsan03.sbcglobal.net] has quit [] 14:01:26 OTOH, a number of people have seen something like that, so I don't trust the cosmic rays theory. 14:01:46 -!- tritchey [n=tritchey@c-98-226-113-211.hsd1.in.comcast.net] has quit [] 14:02:12 I don't think it's cosmic rays either, just certain combinations of kernel/glibc/gcc triggering a latent SBCL bug. 14:02:13 -!- myrkraverk [n=johann@unaffiliated/myrkraverk] has quit [Connection timed out] 14:02:53 xreyes_ [n=xreyes@8.84-48-175.nextgentel.com] has joined #lisp 14:03:01 tripwyre [n=tripwyre@117.193.166.102] has joined #lisp 14:03:19 LiamH [n=none@common-lisp.net] has joined #lisp 14:03:47 dthomp [n=dat@nmd.sbx08736.mcminor.wayport.net] has joined #lisp 14:05:47 i'm pretty sure it's a bug in the PPC port of GENCGC 14:06:42 malumalu [n=malu@hnvr-4dbbae52.pool.einsundeins.de] has joined #lisp 14:06:50 *attila_lendvai* is restarting nodes one by one due wrappers with invalid layout... :( 14:07:15 ... 14:07:17 for the darcs fans: I'm apparently on darcs 2 now. Should I worry about the scp errors when pulling from cl.net, which has darcs 1, or are those harmless? 14:07:26 i hate that they wait until the deadline and then squash the servers... 14:07:47 attila_lendvai: i'm getting to the PCL deadlock next week 14:07:49 lichtblau: harmless, gone in more recent darcs 14:08:23 attila_lendvai: thanks 14:08:24 nikodemus: i'll consider updating to sbcl head tonight. maybe that'll help some. 14:08:39 which version are you running? 14:09:47 nikodemus: 1.0.12.2 14:10:41 pixel5 [n=pixel@copei.de] has joined #lisp 14:10:52 but i have an AVER patch that prevents the endless loop at least... (attached in mail "PCL infinite recursion") 14:12:28 -!- aerique [n=euqirea@xs2.xs4all.nl] has quit ["..."] 14:12:40 -!- wlr [n=walt@c-65-96-92-150.hsd1.ma.comcast.net] has quit [Remote closed the connection] 14:16:30 -!- c|mell [n=cmell@p3209-ipbf1802marunouchi.tokyo.ocn.ne.jp] has quit [Connection timed out] 14:18:50 c|mell [n=cmell@p3209-ipbf1802marunouchi.tokyo.ocn.ne.jp] has joined #lisp 14:20:44 attila_lendvai: right. that's good -- i'll commit the AVER 14:20:51 If I'm debuging from the repl, and want to break when a certan function/method has been called 10 times. Would that be easy to do without modifying the code? 14:22:11 for your own functions, which are not inline -- not heard 14:22:14 hard, even 14:22:31 _mathrick [n=mathrick@wireless.its.sdu.dk] has joined #lisp 14:22:52 hmm. Actualy it is not so dificult (if it is not inline), no. 14:22:57 hugo [n=hugo@unaffiliated/hugo] has joined #lisp 14:23:09 Couldn't he shadow DEFUN and re-compile related function? 14:24:34 Actually, I don't know what would be the implications of shadowing CL:DEFUN. 14:24:36 -!- mathrick [n=mathrick@wireless.its.sdu.dk] has quit [Read error: 110 (Connection timed out)] 14:26:49 vy: probably best avoided :) 14:27:11 cemerick [n=la_mer@75.147.38.122] has joined #lisp 14:27:34 antgreen [n=green@localhost.wbb.net.cable.rogers.com] has joined #lisp 14:29:08 md3 [n=user@85-135-131-179.adsl.slovanet.sk] has joined #lisp 14:29:21 neurogeek_ [n=neurogee@201.208.73.206] has joined #lisp 14:30:07 nikodemus pasted "you could do it like this" at http://paste.lisp.org/display/69631 14:31:38 ThomasI [n=thomas@unaffiliated/thomasi] has joined #lisp 14:33:47 -!- besiria [n=student@195.251.210.108] has quit ["leaving"] 14:34:24 -!- xreyes_ [n=xreyes@8.84-48-175.nextgentel.com] has quit [Read error: 60 (Operation timed out)] 14:35:26 vy pasted "shadowing CL:DEFUN" at http://paste.lisp.org/display/69632 14:38:55 nikodemus: Fantastic! Why do we need (declare (dynamic-extent args)) declaration? 14:39:49 -!- yoshinori [n=read_eva@router1.gpy1.ms246.net] has left #lisp 14:40:43 you don't 14:40:57 it doesn't change the semantics 14:41:11 So why do you prefer to use it? 14:42:03 after a while of programming in lisp you start to hate unnecessary consing 14:42:23 waste is rarely a virtue 14:42:52 -!- Piranha__ [i=jabber-i@number-41.thoughtcrime.us] has left #lisp 14:43:50 tritchey [n=tritchey@c-98-226-113-211.hsd1.in.comcast.net] has joined #lisp 14:45:30 nikodemus: To give the hint that ARGS should be stack allocated in the scope of the caller, some sort of chance for optimization? 14:45:35 c|mell: "unneccessary"? to me, saving time and gaining safety is often more important than short-term memory efficiency. this is even more true as garbage is often much cheaper than one would expect. 14:45:43 xreyes_ [n=xreyes@8.84-48-175.nextgentel.com] has joined #lisp 14:46:18 c|mell: so no. after a while of programming in lisp, you start to appreciate how incredibly well garbage collection works and how little you have to concern yourself about consing. 14:47:11 I value programmer time over computer time. 14:47:43 H4ns, when i first started i thought that gc was great 14:47:52 but with a little experience 14:48:06 i now love the stack 14:48:29 Don't waste vertical IRC space. 14:48:30 knobo annotated #69631 with "using the tracer" at http://paste.lisp.org/display/69631#1 14:49:14 Xach, you need to have your IRC session on a monitor in portrait mode. 14:50:09 Piranha__ [i=jabber-i@number-41.thoughtcrime.us] has joined #lisp 14:50:10 -!- ejs [n=eugen@nat.ironport.com] has quit [Read error: 110 (Connection timed out)] 14:50:30 ejs [n=eugen@80.91.178.218] has joined #lisp 14:50:32 But it is implementation specific. 14:50:40 agnel [n=joel@115.99.2.68] has joined #lisp 14:50:47 Isn't it? 14:51:08 tic: That's tricky to do on a laptop. 14:51:38 nyef, how about a clever setup with mirrors? 14:51:52 knobo: AFAIK, TRACE is just TRACE. Syntax: trace function-name* => trace-result (At least, CLHS says so.) 14:52:01 daniel_ [i=daniel@unaffiliated/daniel] has joined #lisp 14:52:09 nikodemus: re check-wrapper-validity, shouldn't anything that depends on force-cache-flushes/check-wrapper-validity ``succeeding'' hold the pcl lock? 14:53:26 -!- tripwyre [n=tripwyre@117.193.166.102] has quit [Remote closed the connection] 14:53:29 H4ns: My first major lisp optimization problem turned out to be caused by excessive consing. The GC was cheap, the allocation time was killing me. I think anyone who was burned in the same fashion would be a bit leery of consing too much. 14:54:12 (Okay, I was probably also paying the price for excessive unboxing of integers, but still...) 14:54:16 pkhuong: good question 14:54:31 nyef, but the point is probably that you should focus on your problem first. getting a good algorithm. then profile. 14:54:35 nyef: i am not saying that garbage collection solves all problem. it is just that it solves more problems than it creates, and declaring consing as a general enemy is just silly. 14:54:45 H4ns: Fair enough. 14:55:14 we're pretty far away from the topic of declaring &rest arguments DX (and what that truly means :) 14:55:26 *H4ns* shuts up 14:55:41 *nyef* goes back to figuring out ARM assembly. 14:56:14 matley [n=matley@matley.imati.cnr.it] has joined #lisp 14:57:38 hah. that symbol-macro, type declaration, defmethod body fix in recent sbcl just triggered a ZPB-TTF declaration bug 14:57:52 i was wondering why i never saw it until now 14:58:42 knobo annotated #69631 with "As a macro" at http://paste.lisp.org/display/69631#2 15:00:20 rme [n=rme@pool-70-104-125-111.chi.dsl-w.verizon.net] has joined #lisp 15:00:48 Nshag [n=shagoune@Mix-Orleans-105-4-98.w193-250.abo.wanadoo.fr] has joined #lisp 15:01:15 pkhuong: possibly we could use class-lock there instead of pcl-lock, as well 15:03:33 -!- michaelw [i=michaelw@88.198.49.16] has quit [Read error: 110 (Connection timed out)] 15:03:35 Except it should restore previus behavor of traceing for that function when done. If it was traced before, it should be traced aftrer. 15:06:16 willb [n=wibenton@128.105.48.219] has joined #lisp 15:06:23 -!- jdz [n=jdz@85.254.248.11] has quit ["Somebody rebooted me"] 15:07:33 yoshinori [n=read_eva@router1.gpy1.ms246.net] has joined #lisp 15:07:59 Adrinael_ [n=adrinael@barrel.rolli.org] has joined #lisp 15:08:00 -!- daniel [i=daniel@unaffiliated/daniel] has quit [Read error: 110 (Connection timed out)] 15:08:58 -!- matimago [n=user@cac94-10-88-170-236-224.fbx.proxad.net] has quit [Remote closed the connection] 15:09:06 hosiawak [n=user@host81-137-9-70.in-addr.btopenworld.com] has joined #lisp 15:09:37 ineiros_ [n=ineiros@kosh.hut.fi] has joined #lisp 15:09:59 -!- Adrinael [i=adrinael@rid7.kyla.fi] has quit [Read error: 104 (Connection reset by peer)] 15:10:07 -!- willb [n=wibenton@128.105.48.219] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- rme [n=rme@pool-70-104-125-111.chi.dsl-w.verizon.net] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- ejs [n=eugen@80.91.178.218] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- ThomasI [n=thomas@unaffiliated/thomasi] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- cemerick [n=la_mer@75.147.38.122] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- c|mell [n=cmell@p3209-ipbf1802marunouchi.tokyo.ocn.ne.jp] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- sellout [n=greg@guest-fw.dc4.itasoftware.com] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- attila_lendvai [n=ati@business-89-132-61-222.business.broadband.hu] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- seelenquell [n=seelenqu@pD9E46D9C.dip.t-dialin.net] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- ebzzry [n=rmm@124.217.85.39] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- adeht [n=death@nessers.org] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- yahooooo3 [n=yahooooo@c-24-19-6-19.hsd1.wa.comcast.net] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- manic12_ [n=manic12@c-98-227-126-106.hsd1.il.comcast.net] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- ia [n=ia@89.169.165.188] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- mikezor [n=mikael@c-a388e253.04-404-7570701.cust.bredbandsbolaget.se] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- ryepup1 [n=ryepup@216.155.97.1] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- olejorgenb [i=bronner@gaupe.stud.ntnu.no] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- pitui [n=pitui@doh.research.att.com] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- ineiros [n=ineiros@kosh.hut.fi] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- ZabaQ [n=ZabaQ@194-105-174-193.ifb.co.uk] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- jkantz [n=jkantz@ita4fw1.itasoftware.com] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- delYsid [n=user@debian/developer/mlang] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- tic [n=tic@c83-249-197-248.bredband.comhem.se] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- SUNWjoejaxx [i=joejaxx@fluxbuntu/founder/joejaxx] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- p8m [n=dmm@mattlimech.com] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- ``Erik_ [i=erik@c-68-54-174-162.hsd1.md.comcast.net] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- sabetts [n=sabetts@S0106000a95700fd8.gv.shawcable.net] has quit [lem.freenode.net irc.freenode.net] 15:10:07 -!- alexsuraci [n=Alex@pool-71-188-133-67.aubnin.fios.verizon.net] has quit [lem.freenode.net irc.freenode.net] 15:11:02 tic [n=tic@c83-249-197-248.bredband.comhem.se] has joined #lisp 15:11:12 I'd like to optimize this function: http://pastie.org/306426 15:11:44 willb [n=wibenton@128.105.48.219] has joined #lisp 15:11:44 rme [n=rme@pool-70-104-125-111.chi.dsl-w.verizon.net] has joined #lisp 15:11:44 ejs [n=eugen@80.91.178.218] has joined #lisp 15:11:44 cemerick [n=la_mer@75.147.38.122] has joined #lisp 15:11:44 c|mell [n=cmell@p3209-ipbf1802marunouchi.tokyo.ocn.ne.jp] has joined #lisp 15:11:44 sellout [n=greg@guest-fw.dc4.itasoftware.com] has joined #lisp 15:11:44 attila_lendvai [n=ati@business-89-132-61-222.business.broadband.hu] has joined #lisp 15:11:44 seelenquell [n=seelenqu@pD9E46D9C.dip.t-dialin.net] has joined #lisp 15:11:44 ebzzry [n=rmm@124.217.85.39] has joined #lisp 15:11:44 adeht [n=death@nessers.org] has joined #lisp 15:11:44 yahooooo3 [n=yahooooo@c-24-19-6-19.hsd1.wa.comcast.net] has joined #lisp 15:11:44 manic12_ [n=manic12@c-98-227-126-106.hsd1.il.comcast.net] has joined #lisp 15:11:44 ``Erik_ [i=erik@c-68-54-174-162.hsd1.md.comcast.net] has joined #lisp 15:11:44 ryepup1 [n=ryepup@216.155.97.1] has joined #lisp 15:11:44 alexsuraci [n=Alex@pool-71-188-133-67.aubnin.fios.verizon.net] has joined #lisp 15:11:44 sabetts [n=sabetts@S0106000a95700fd8.gv.shawcable.net] has joined #lisp 15:11:44 pitui [n=pitui@doh.research.att.com] has joined #lisp 15:11:44 mikezor [n=mikael@c-a388e253.04-404-7570701.cust.bredbandsbolaget.se] has joined #lisp 15:11:44 ZabaQ [n=ZabaQ@194-105-174-193.ifb.co.uk] has joined #lisp 15:11:44 p8m [n=dmm@mattlimech.com] has joined #lisp 15:11:44 ineiros [n=ineiros@kosh.hut.fi] has joined #lisp 15:11:44 olejorgenb [i=bronner@gaupe.stud.ntnu.no] has joined #lisp 15:11:44 delYsid [n=user@debian/developer/mlang] has joined #lisp 15:11:44 jkantz [n=jkantz@ita4fw1.itasoftware.com] has joined #lisp 15:11:44 SUNWjoejaxx [i=joejaxx@fluxbuntu/founder/joejaxx] has joined #lisp 15:11:46 -!- ebzzry [n=rmm@124.217.85.39] has quit [Remote closed the connection] 15:11:46 ThomasI [n=thomas@91-64-144-252-dynip.superkabel.de] has joined #lisp 15:11:47 ebzzry [n=rmm@124.217.85.39] has joined #lisp 15:11:47 -!- SUNWjoejaxx [i=joejaxx@fluxbuntu/founder/joejaxx] has quit [Success] 15:11:47 how can I create an array in sbcl and fill it with instances of an object but each instance must be different ? 15:11:58 any suggestions ? 15:12:01 ia [n=ia@89.169.165.188] has joined #lisp 15:12:03 -!- ineiros [n=ineiros@kosh.hut.fi] has quit [Connection reset by peer] 15:12:17 hosiawak: make the array, loop and fill it. 15:12:22 hosiawak: (map-into (make-array max :element-type 'fake) (lambda () (make-instance 'fake)) ; note that that's equivalene to :element-type T. 15:13:05 -!- yoshinori [n=read_eva@router1.gpy1.ms246.net] has left #lisp 15:13:16 matimago [n=user@cac94-10-88-170-236-224.fbx.proxad.net] has joined #lisp 15:14:01 is the element type when upgraded === t even checked for default safety? 15:14:03 -!- adeht [n=death@nessers.org] has quit ["changing servers"] 15:14:03 make-instance is likely what costs the most here anyway. 15:14:08 -!- Jacob_H [n=jacob@92.3.170.158] has quit [Read error: 110 (Connection timed out)] 15:14:14 -!- H4ns [n=Hans@p57BB9CBA.dip0.t-ipconnect.de] has quit ["Leaving."] 15:14:19 adeht [n=death@nessers.org] has joined #lisp 15:15:13 Jacob_H [n=jacob@92.3.136.178] has joined #lisp 15:15:20 -!- daniel_ is now known as daniel 15:15:47 weirdo: what check would you propose? 15:15:56 type-error 15:16:04 all values are of type T. 15:16:22 The originaly given type is forgotten. 15:16:29 oh 15:16:41 weirdo: type declarations are mostly about performance in CL. SBCL doesn't currently check that you're not doing anything obviously wrong wrt the expressed element type, not does it take advantage of it (instead of the actual element type) to enable more optimisations. 15:17:12 -!- ecraven [n=nex@140.78.42.103] has quit ["bbl"] 15:18:22 pkhuong: you forget (declare (type (simple-array (integer 0 3) (*)) foo)), which means that (aref foo ...) can be taken as (the (integer 0 3) (aref foo ...)) -- and IIRC we do do that 15:18:24 matimago, does the same hold true for clos? :type appears to be ignored 'cept for introspection reasons, but aside from (unimplemented) class locking i couldn't think of a possible use for optimization 15:18:25 manic12 [n=manic12@c-98-227-126-106.hsd1.il.comcast.net] has joined #lisp 15:18:53 nikodemus: oh right. Krystof's patch; Neat! 15:19:17 weirdo: implementation dependant. sbcl does some type checking for slots. (There's no "upgrading" there). 15:19:51 -!- jlf` [n=user@netblock-68-183-235-19.dslextreme.com] has quit [No route to host] 15:20:12 weirdo pasted "no type checking for clos" at http://paste.lisp.org/display/69633 15:20:19 weirdo: only in safe code 15:20:26 oh! that changes a lot 15:20:26 thx 15:20:57 I've got (declaim (sb-ext:muffle-conditions (or style-warning compiler-note)) (optimize (speed 0) (space 0) (debug 3) (safety 3))) ; in my ~/.sbclrc 15:22:30 milanj [n=milan@77.46.250.113] has joined #lisp 15:22:44 -!- dcrawford [n=dcrawfor@ita4fw1.itasoftware.com] has quit [Connection reset by peer] 15:23:17 nikodemus: and we also do some checks on the expressed element type, since the same commit, I suppose. 15:24:56 i don't remember. possible 15:24:57 dcrawford [n=dcrawfor@ita4fw1.itasoftware.com] has joined #lisp 15:24:57 -!- dcrawford [n=dcrawfor@ita4fw1.itasoftware.com] has quit [Client Quit] 15:24:57 dcrawford_ [n=dcrawfor@ita4fw1.itasoftware.com] has joined #lisp 15:25:04 weirdo: never mind, SBCL actually uses declared element types both for safety and performance. (I always forget that for some reason...) 15:25:06 -!- dcrawford_ is now known as dcrawford 15:25:09 mindCrime [n=chatzill@216.27.62.2] has joined #lisp 15:25:23 but the declare is not tied to the original make-array form in any way 15:26:49 brmiller [n=brmiller@vip2.corp.dotster.net] has joined #lisp 15:27:11 -!- brmiller [n=brmiller@vip2.corp.dotster.net] has left #lisp 15:29:08 -!- mikesch [n=axel@static-87-79-66-80.netcologne.de] has quit [Read error: 113 (No route to host)] 15:34:11 sohail [n=Sohail@unaffiliated/sohail] has joined #lisp 15:34:27 -!- mindCrime [n=chatzill@216.27.62.2] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- ebzzry [n=rmm@124.217.85.39] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- ThomasI [n=thomas@91-64-144-252-dynip.superkabel.de] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- mikezor [n=mikael@c-a388e253.04-404-7570701.cust.bredbandsbolaget.se] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- ryepup1 [n=ryepup@216.155.97.1] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- c|mell [n=cmell@p3209-ipbf1802marunouchi.tokyo.ocn.ne.jp] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- ejs [n=eugen@80.91.178.218] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- manic12_ [n=manic12@c-98-227-126-106.hsd1.il.comcast.net] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- olejorgenb [i=bronner@gaupe.stud.ntnu.no] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- pitui [n=pitui@doh.research.att.com] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- yahooooo3 [n=yahooooo@c-24-19-6-19.hsd1.wa.comcast.net] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- ZabaQ [n=ZabaQ@194-105-174-193.ifb.co.uk] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- rme [n=rme@pool-70-104-125-111.chi.dsl-w.verizon.net] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- sellout [n=greg@guest-fw.dc4.itasoftware.com] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- jkantz [n=jkantz@ita4fw1.itasoftware.com] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- delYsid [n=user@debian/developer/mlang] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- attila_lendvai [n=ati@business-89-132-61-222.business.broadband.hu] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- p8m [n=dmm@mattlimech.com] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- ``Erik_ [i=erik@c-68-54-174-162.hsd1.md.comcast.net] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- sabetts [n=sabetts@S0106000a95700fd8.gv.shawcable.net] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- alexsuraci [n=Alex@pool-71-188-133-67.aubnin.fios.verizon.net] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- willb [n=wibenton@128.105.48.219] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- cemerick [n=la_mer@75.147.38.122] has quit [lem.freenode.net irc.freenode.net] 15:34:27 -!- seelenquell [n=seelenqu@pD9E46D9C.dip.t-dialin.net] has quit [lem.freenode.net irc.freenode.net] 15:35:13 mindCrime [n=chatzill@216.27.62.2] has joined #lisp 15:35:13 ebzzry [n=rmm@124.217.85.39] has joined #lisp 15:35:13 ThomasI [n=thomas@91-64-144-252-dynip.superkabel.de] has joined #lisp 15:35:13 willb [n=wibenton@128.105.48.219] has joined #lisp 15:35:13 rme [n=rme@pool-70-104-125-111.chi.dsl-w.verizon.net] has joined #lisp 15:35:13 ejs [n=eugen@80.91.178.218] has joined #lisp 15:35:13 cemerick [n=la_mer@75.147.38.122] has joined #lisp 15:35:13 c|mell [n=cmell@p3209-ipbf1802marunouchi.tokyo.ocn.ne.jp] has joined #lisp 15:35:13 sellout [n=greg@guest-fw.dc4.itasoftware.com] has joined #lisp 15:35:13 attila_lendvai [n=ati@business-89-132-61-222.business.broadband.hu] has joined #lisp 15:35:13 seelenquell [n=seelenqu@pD9E46D9C.dip.t-dialin.net] has joined #lisp 15:35:13 yahooooo3 [n=yahooooo@c-24-19-6-19.hsd1.wa.comcast.net] has joined #lisp 15:35:13 manic12_ [n=manic12@c-98-227-126-106.hsd1.il.comcast.net] has joined #lisp 15:35:13 ``Erik_ [i=erik@c-68-54-174-162.hsd1.md.comcast.net] has joined #lisp 15:35:13 ryepup1 [n=ryepup@216.155.97.1] has joined #lisp 15:35:13 alexsuraci [n=Alex@pool-71-188-133-67.aubnin.fios.verizon.net] has joined #lisp 15:35:13 sabetts [n=sabetts@S0106000a95700fd8.gv.shawcable.net] has joined #lisp 15:35:13 pitui [n=pitui@doh.research.att.com] has joined #lisp 15:35:13 mikezor [n=mikael@c-a388e253.04-404-7570701.cust.bredbandsbolaget.se] has joined #lisp 15:35:13 ZabaQ [n=ZabaQ@194-105-174-193.ifb.co.uk] has joined #lisp 15:35:13 p8m [n=dmm@mattlimech.com] has joined #lisp 15:35:13 olejorgenb [i=bronner@gaupe.stud.ntnu.no] has joined #lisp 15:35:13 delYsid [n=user@debian/developer/mlang] has joined #lisp 15:35:13 jkantz [n=jkantz@ita4fw1.itasoftware.com] has joined #lisp 15:35:41 [ot] are 'colonel' and 'kernel' homophones? 15:36:24 depends on dialect 15:36:27 -!- Jacob_H [n=jacob@92.3.136.178] has quit ["Leaving"] 15:36:29 but generally yes, more or less 15:36:40 -!- ebzzry [n=rmm@124.217.85.39] has quit [Remote closed the connection] 15:36:47 nikodemus` [n=nikodemu@cs27013130.pp.htv.fi] has joined #lisp 15:37:10 -!- manic12_ [n=manic12@c-98-227-126-106.hsd1.il.comcast.net] has quit [Connection timed out] 15:37:32 ebzzry [n=rmm@124.217.85.39] has joined #lisp 15:37:41 weirdo: as far as I'm concerned, yes. 15:38:14 thx. wonder where did the extra 'r' in the former come from 15:38:38 same place massachusetts gets it I guess 15:38:45 perhaps borrowed from Featherstonehaugh 15:38:52 weirdo: some dialects of English have no `r' sound in "kernel". 15:39:28 so colonel borrowed it from kernel? 15:39:45 massachusetts gets an extra 'r'? wow, i need some english lessons, seriously... 15:39:54 not the word, the dialect 15:40:03 dcrawford: I'm pretty certain colonel is a non-English import 15:41:01 it's from Italian "colonnello" 15:41:21 *splittist* reads along with Wikipedia 15:41:51 probably came to English with the Normands. 15:42:46 For those who aren't following along: "In modern English, the word colonel is pronounced similarly to kernel (of grain) as a result of entering the language from Middle French in two competing forms, dissimilated coronel and colonel." 15:42:53 beach: there's various non-rhotic english dialects, both in England and the US 15:43:12 I don't know if there's non-rhotic dialects in Australia and Canadia, I should check... 15:44:15 certainly in Australia and New Zealand there are. 15:44:18 *rme* looks up rhotic in the dictionary 15:45:21 -!- nikodemus [n=nikodemu@cs27013130.pp.htv.fi] has quit [Read error: 110 (Connection timed out)] 15:45:29 see, and we think COMPUTER language standards have problems :) 15:45:41 -!- cracki [n=cracki@sglty.kawo2.RWTH-Aachen.DE] has quit ["The funniest things in my life are truth and absurdity."] 15:46:07 of all the linguistics fields I could've gotten into, it just -had- to be phonology :-\ 15:46:13 Only non-rhotic speakers know the true meaning of D'yer Ma'ker 15:46:15 *splittist* thinks he is a non-rhotic speaker 15:46:40 splittist: how do you pronounce 'park my car'? :) 15:47:27 ths_ [n=ths@p549AEEB3.dip.t-dialin.net] has joined #lisp 15:47:27 sykopomp: not being a phonologist I think I pronounce it like it is spelled. 15:47:42 Heh 15:47:49 -!- jtoy [n=jtoy@58.63.171.97] has quit [] 15:48:00 sykopomp: 'pahk my cah' ? 15:48:08 not like 'pahk my cahhh'? (yes, that's my poor attempt at phonetic spelling of a bahstan accent) 15:48:29 splittist: that's probably non-rhotic, then :P 15:48:57 If I'm putting on an 'American' accent I have to make an effort to pronounce the r's. 15:49:12 *sykopomp* should probably stop the OT 15:49:12 -!- ths [n=ths@p549AC6B4.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)] 15:49:13 one of the worst England/New England pronunication horrors: worcester 15:49:49 basically: wuhsta 15:50:03 Is that like cholmondeley? 15:50:08 dcrawford: haha. Woostah! 15:51:07 Assume everyone has seen http://en.wikipedia.org/wiki/Ghoti 15:51:22 dcrawford: i prefer haverhill as an example. 15:51:24 then we can talk about 'macrolet'/'chevrolet'... 15:51:47 haha. 15:51:51 I say wor-ster (rhotically) ... I love watching people encounter "worcestershire sauce" for the first time, though ... "wor-chest-er-shi-er" 15:52:46 so you say wor-ster-shi-er ? 15:52:51 gloucester in massachusetts and new gloucester in maine are pronounced with significant differences, too. 15:53:19 tic: wor-ster-sheer 15:53:19 Xach: what's haverhill? Heavill? 15:53:22 (not just the "new") 15:53:30 Neat. ARM does have a "Count Leading Zeros" instruction. 15:53:34 sykopomp: approximately hay-vrill. 15:53:56 tic: others would say wo-sta-sha 15:54:21 *tic* has yet lot to learn. 15:54:31 My dad likes to call it "what's this here sauce" 15:54:33 Uhm. Still. Like I said. 15:54:37 sellout, hehe 15:54:53 in my house, that's approximately "wooster-sheer". 15:55:11 oh yeah, what sellout said. 15:55:15 -!- vy [n=user@213.139.194.86] has quit [Remote closed the connection] 15:55:31 jlf` [n=user@209.204.171.109] has joined #lisp 15:55:45 I'm not an MA native, but after 5 years of living here, I've settled for saying "Wooster" 15:56:34 -!- _mathrick [n=mathrick@wireless.its.sdu.dk] has quit [Read error: 110 (Connection timed out)] 15:57:00 elurin [n=user@81.213.203.109] has joined #lisp 15:58:38 H4ns [n=hans@p57A0D51F.dip.t-dialin.net] has joined #lisp 16:01:07 hjpark` [n=user@61.109.28.117] has joined #lisp 16:01:35 -!- eevar2 [n=jalla@106.80-203-27.nextgentel.com] has quit ["This computer has gone to sleep"] 16:07:14 -!- ramkrsna [n=ramkrsna@unaffiliated/ramkrsna] has quit [Remote closed the connection] 16:07:29 -!- srcerer [n=chatzill@dns2.klsairexpress.com] has quit [Remote closed the connection] 16:08:45 gonzojive [n=red@fun.Stanford.EDU] has joined #lisp 16:10:16 -!- ia [n=ia@89.169.165.188] has quit [Read error: 110 (Connection timed out)] 16:10:23 -!- x6j8x [n=x6j8x@tmo-100-35.customers.d1-online.com] has quit [Read error: 110 (Connection timed out)] 16:10:24 -!- gonzojive [n=red@fun.Stanford.EDU] has quit [Remote closed the connection] 16:10:40 -!- CrazyEddy [n=CrazyEdd@wrongplanet/CrazyEddy] has quit [Connection timed out] 16:10:56 gonzojive [n=red@fun.Stanford.EDU] has joined #lisp 16:11:01 Wombat1 [n=willy@216-31-242-4.static-ip.telepacific.net] has joined #lisp 16:11:23 ia [n=ia@89.169.165.188] has joined #lisp 16:12:13 CrazyEddy [n=CrazyEdd@220-253-113-170.VIC.netspace.net.au] has joined #lisp 16:16:03 manuel__ [n=manuel@HSI-KBW-085-216-061-130.hsi.kabelbw.de] has joined #lisp 16:17:29 -!- hjpark [n=user@61.109.28.117] has quit [Read error: 110 (Connection timed out)] 16:18:26 -!- dthomp [n=dat@nmd.sbx08736.mcminor.wayport.net] has quit [Read error: 60 (Operation timed out)] 16:18:37 _mathrick [n=mathrick@users177.kollegienet.dk] has joined #lisp 16:19:53 -!- matley [n=matley@matley.imati.cnr.it] has quit [Remote closed the connection] 16:20:02 matley [n=matley@matley.imati.cnr.it] has joined #lisp 16:24:21 -!- _mathrick is now known as mathrick 16:24:49 -!- manuel__ [n=manuel@HSI-KBW-085-216-061-130.hsi.kabelbw.de] has quit [] 16:25:02 -!- agnel [n=joel@115.99.2.68] has quit [Read error: 104 (Connection reset by peer)] 16:25:07 acrid [n=mckay@reverse.control4.com] has joined #lisp 16:26:23 Yuuhi [n=user@p5483D4FE.dip.t-dialin.net] has joined #lisp 16:26:40 UtopiahGHML [i=utopiah@gateway/shell/blinkenshell.org/x-499bc50f17992b3e] has joined #lisp 16:27:31 LeCamarade [n=LeCamara@41.222.7.35] has joined #lisp 16:28:05 lemonodor [n=lemonodo@adsl-76-214-15-172.dsl.lsan03.sbcglobal.net] has joined #lisp 16:28:09 -!- UtopiahGHML [i=utopiah@gateway/shell/blinkenshell.org/x-499bc50f17992b3e] has left #lisp 16:28:46 -!- milanj [n=milan@77.46.250.113] has quit ["This computer has gone to sleep"] 16:31:00 -!- oudeis [n=oudeis@bzq-79-176-18-217.red.bezeqint.net] has quit ["This computer has gone to sleep"] 16:31:11 -!- reaver__ [n=reaver@212.88.117.162] has quit [Read error: 104 (Connection reset by peer)] 16:31:21 alley_cat [i=AlleyCat@sourcemage/elder/alleycat] has joined #lisp 16:31:56 -!- nikodemus` [n=nikodemu@cs27013130.pp.htv.fi] has quit [Read error: 60 (Operation timed out)] 16:32:08 -!- haiwei [n=haiwei@61.149.75.2] has quit ["Leaving."] 16:33:57 mikesch [n=axel@xdsl-78-34-217-150.netcologne.de] has joined #lisp 16:34:53 -!- kiuma [n=kiuma@83-103-59-156.ip.fastwebnet.it] has quit ["Bye bye ppl"] 16:35:14 rindolf [n=shlomi@bzq-219-139-216.static.bezeqint.net] has joined #lisp 16:38:37 -!- epoch [n=K-I-S-S@p3m/member/epoch] has quit ["  "] 16:39:07 -!- brickhazel [n=brickhaz@63.144.132.78] has quit [Remote closed the connection] 16:39:23 grkz [n=martin@c-7bf9e255.010-54-6f72652.cust.bredbandsbolaget.se] has joined #lisp 16:39:59 srcerer [n=chatzill@dns2.klsairexpress.com] has joined #lisp 16:42:26 -!- lichtblau [n=user@wallstrasse.knowledgetools.de] has quit [Remote closed the connection] 16:42:47 lichtblau [n=user@wallstrasse.knowledgetools.de] has joined #lisp 16:42:54 -!- ejs [n=eugen@80.91.178.218] has quit [Remote closed the connection] 16:43:13 nikodemus [n=nikodemu@cs27013130.pp.htv.fi] has joined #lisp 16:43:19 -!- matley [n=matley@matley.imati.cnr.it] has quit [Read error: 110 (Connection timed out)] 16:46:32 dthomp [n=dat@dyn-188-dynamic.linfield.edu] has joined #lisp 16:46:54 H4ns1 [n=hans@p57A0D51F.dip.t-dialin.net] has joined #lisp 16:47:26 lieven_ [n=lieven@ip-213-49-240-62.dsl.scarlet.be] has joined #lisp 16:47:50 tcr [n=tcr@host145.natpool.mwn.de] has joined #lisp 16:48:36 So, I've been trying to generate unit testing for all of my code. The testing I write is not comprehensive, but I at least try to have some obvious tests for each function. One thing I'm running into, though, is that I'd like tests to have dependency. So if test X fails, don't bother with test Y because it depends on X. This isn't possible with lisp-unit because it is dead simple. 16:49:10 seems pretty straightforward. 16:49:18 -!- H4ns [n=hans@p57A0D51F.dip.t-dialin.net] has quit [Read error: 104 (Connection reset by peer)] 16:49:20 Is dependency a reasonable requirement? 16:50:02 davazp [n=user@72.Red-88-6-205.staticIP.rima-tde.net] has joined #lisp 16:50:05 5am has that 16:51:01 fe[nl]ix: okay, I'll go look at that again. I've been working with lisp-unit because it is so simple, I didn't want to get bogged down with reading until it was warranted. I may be getting there. 16:51:05 *dlowe* always runs every test and ignores dependencies 16:51:33 Are you asking to not run tests that you know will fail as some kind of optimization, or is there another reason to deflate your failing test count? 16:52:03 Only one failing test (that everything depends on). Near success! 16:52:05 dlowe: Yeah, that is my current approach. I'm not well versed on unit testing, so I wasn't sure if having dependencies violated some goal of unit testing. 16:52:26 that isn't something that most people would care to have in their test framework. 16:52:35 tmh: it will become less maintainable. You never know when those dependencies are going to change. 16:52:44 depends on how long it takes for their tests to run 16:52:46 "oh no, I've regressed, the failures went from only 1 to 500" 16:53:22 "oops, that one test caused the other 500 not to be attempted." 16:53:33 nyef: My thinking was that if a core dependency fails, lets focus on it and not wade through a lot of other failures. 16:53:45 maybe what you actually want is a test suite that does halt-on-fail so you don't have to watch 500 fails scroll by, then dig your way back up to the first one 16:53:56 -!- envi^home [n=envi@220.121.234.156] has quit ["Leaving"] 16:53:59 dlowe: Good point re: maintainability. 16:54:01 -!- gdmfsob [n=gdmfsob@dm.sonopia.com] has quit [Read error: 110 (Connection timed out)] 16:54:01 tmh: That sounds like a case for test case severity, not test case elision. 16:54:03 generally dependencies in a test framework would be used for optional features that your environment may or may not provide 16:54:04 I guess if your reporting was real time then it would matter less, you could stop the test run and go investigate the top level failure 16:54:09 stefil signals a condition and lets you pick what you want to do 16:54:43 e.g. If I don't have ssl support, don't run ssl tests. 16:55:03 -!- hjpark` [n=user@61.109.28.117] has quit [Connection timed out] 16:56:07 I'm finding this discussion somewhat appropriate right now, because I'm knee deep in writing some code that I'd love to have test cases for, but I can't even compile it yet and I am aware of no hooks in the build process to stop partway and run a test suite. 16:56:41 foom: not necessarily, e.g. if the connect-to-DB test fails it makes no sense to test transactions 16:57:24 (eval-when (:compile-toplevel) (run-early-tests)) ; not quite joking 16:57:27 fe[nl]ix: True, but it makes sense for those tests to fail with a reason of "unable to connect to DB". 16:57:28 fe[nl]ix: but since all your tests always pass, internal dependencies don't actually matter much. 16:58:21 *dlowe* recommends The Art of Unit Testing by Roy Osherove 16:58:58 At the root of my question is whether I should continue with lisp-unit and just modify it to fit my usage or invest the time in learning another framework. I think I can still get by with lisp-unit. I have started maintaining my own copy of it with my little additions. 16:59:18 dlowe: thanks for the reference. 16:59:23 *tmh* notes it down 16:59:31 IMO if you have to worry much about so many tests failing that you can't figure out what you broke, you're doing something wrong. 16:59:32 perhaps it would be more useful to have deps and some graphical tool for visualizing test results, so that you be able see that a certain node(failed test) is blocking some other tests 16:59:35 tmh: you can buy it as an ebook, too 16:59:36 tmh: I'd rather go with something with more community support. 17:00:13 The last couple times I did anything with unit testing in Lisp, I used SB-RT. 17:00:18 tcr: Any recommendations? 17:00:43 5am or rt/sb-rt, maybe 17:00:56 tmh: Personally I use sb-rt because it comes with SBCL. It's pretty simple. If I needed more, I'd probably go with 5am or stefil. 17:01:00 And I've found that I tend not to write stuff that's really "dependent". 17:01:27 Ah, I didn't know sbcl had one included. 17:01:36 When writing a parser, for example, I'd rather see -all- of the failing test cases, because that way I have a better chance at figuring out the pattern to the failures. 17:01:38 michaelw [i=michaelw@lambda.foldr.org] has joined #lisp 17:01:46 SB-RT is a contrib, actually. 17:02:19 tmh: What's the software you're writing tests for? 17:02:40 what's the easiest way to check if a symbol has (lexical) function definition associated? 17:02:45 ie. at macroexpansion time 17:03:19 tcr: Personal stuff. Linear algebra code supporting my PhD. In the past month, consulting code to generate input for an finite element analysis program based on a few parameters. 17:03:27 mathrick: fboundp 17:03:31 I'm trying to get my unit testing up-to-date. 17:03:41 though that will not tell you about flet/label function 17:03:47 photon2 [n=photon@unaffiliated/photon] has joined #lisp 17:04:51 segv_: Hence, not lexical. 17:05:03 we've written huge unit tests using stefil but never needed dependencies. a tree of test suites is useful though... 17:05:25 segv_: yeah, that's why I asked about lexical 17:05:43 yeah, my brain skipped over that word 17:05:44 mathrick: I'm going to go out on a limb here and say "there isn't one". 17:05:49 we just start the tests, if there are a zillion errors/fails then fix the bug in the basic infrastructure, then restart tests 17:06:00 Hmmm, I'll have to dig into RT a little more before I switch because I want to be able to run my unit tests on other implementations. 17:06:04 nyef: hrmpf 17:06:08 s/RT/SB-RT/ 17:06:18 Finding out if a given symbol has a macro definition, even lexically, is easy enough, but a function definition is trickier. 17:06:30 mathrick: The information is not necessarily there at that time. 17:06:45 -!- LeCamarade [n=LeCamara@41.222.7.35] has quit [Connection timed out] 17:06:57 tcr: Well, if it's a lexical-definition-only thing, wouldn't it have to be? 17:07:11 tcr: howso? It needs to be somewhere for the compiler to compile 17:07:49 mathrick: The specific requirement is to determine if a given symbol is named in an enclosing flet or labels, right? 17:08:07 oudeis [n=oudeis@89-138-123-60.bb.netvision.net.il] has joined #lisp 17:08:38 nyef: yes 17:08:39 x6j8x [n=x6j8x@77.21.43.186] has joined #lisp 17:09:16 mathrick: You can try your luck with sb-cltl2. 17:09:23 *mathrick* thinks again about what he wants to do again 17:09:38 tcr: what do I want to look at in that case? 17:09:39 Or have an enclosing macro which code walks 17:09:50 Yeah, the environment access stuff is your best bet here, and as it was stripped from the standard after CLtL2 it's not always available. 17:10:00 yeah, that sucks 17:10:11 mathrick: sb-cltl2:function-information 17:10:14 env access is one of those things CL+ needs to address 17:10:29 But, really, you might find it easier to reconsider your use-case. 17:10:33 Yeah, I have a portability layer for that ... only does SBCL and CCL, though. 17:10:42 should be trivial to make it work in ACL, too 17:10:53 sellout: segv does so too in arnesi. 17:11:06 tcr: Ah damn. I'll throw mine away first. 17:11:11 s/first/now/ 17:11:20 Should have checked there first, anyhow 17:12:38 nyef: I'm basically writing a sugar macro for flattening heavily nested calls, and wanted to make it decide between expanding into (SYM ...) vs. (funcall SYM ...) automatically 17:12:41 matley [n=matley@matley.imati.cnr.it] has joined #lisp 17:12:58 And, yeah, I guess the information must be there somewhere to make (macrolet ((foo ...)) (flet ((foo ...)) (foo 42)) work. 17:13:58 sellout: fyi, i've factored out the walker from arnesi into a slim cl-walker. the env api was also worked on since then... 17:14:33 yCrazyEdd [n=CrazyEdd@220.253-199-221.VIC.netspace.net.au] has joined #lisp 17:15:03 hm, although, i think the env accessor implementations need some care anywhere else than sbcl and maybe clisp 17:15:11 -!- lemonodor [n=lemonodo@adsl-76-214-15-172.dsl.lsan03.sbcglobal.net] has quit [] 17:15:26 But it can be tricky. For example if you have (defmacro m (&environment env) `(with-foo (...) ,(frob (function-information 'quux env))), the ENV of M does not contain any information about locally bound functions in WITH-FOO. 17:16:10 -!- antgreen [n=green@localhost.wbb.net.cable.rogers.com] has left #lisp 17:16:40 tcr, do you think it should? 17:18:09 Adamant [n=Adamant@c-98-244-152-196.hsd1.ga.comcast.net] has joined #lisp 17:19:18 tic: there are cases for both "yes" and "no" 17:20:21 I'd love to hear the case for "yes" 17:20:23 seelenquell_ [n=seelenqu@pD9E46863.dip.t-dialin.net] has joined #lisp 17:20:24 tic: I can't think of how it could 17:20:35 it would no doubt be very confused about what a backquote is 17:20:52 in case of `(with-foo ... ,(env-stuff)) where is not way env-stuff can know about things bound by with-foo 17:20:53 schoppenhauer [n=css@unaffiliated/schoppenhauer] has joined #lisp 17:21:17 but `(with-foo ... (env-stuff)) can 17:21:27 unless that was (with-my-code-walker `(with-foo ... ,(env-stuff))) :-) 17:22:47 brickhazel [n=brickhaz@ool-45717ef5.dyn.optonline.net] has joined #lisp 17:22:48 Nah, chandler. Lisp just has to be telepathic; it has all the other AI already, right? 17:23:17 I think my favorite name for a WITH-* macro has to be WITH-MALICE-AFORETHOUGHT, but I haven't found a suitable macro to use with it... 17:23:22 -!- photon2 [n=photon@unaffiliated/photon] has quit ["Leaving"] 17:23:33 A simple matter of implementing Schroedinger environments 17:23:47 splittist: That would be a catastrophe. 17:24:01 I wasn't actually thinking about the case with backquote, but I just realised that no-backquote case could be handled by a handoff to another macro 17:24:08 so no, there aren't any cases where it should 17:24:51 nyef: nah, it'd be a blessing 17:24:52 tcr, modify all with-foo:ers. But that'd be weird to enforce. I'm just curious, I have no answer to that. 17:25:01 nyef pasted "I don't know if I should be pleased with these instruction definitions or not" at http://paste.lisp.org/display/69642 17:25:13 -!- schoppenhauer [n=css@unaffiliated/schoppenhauer] has quit [Remote closed the connection] 17:25:13 you just make it compile all possible sources, and choose the ones that compile and work 17:25:16 epoch [n=K-I-S-S@p3m/member/epoch] has joined #lisp 17:25:41 -!- epoch [n=K-I-S-S@p3m/member/epoch] has quit [Read error: 104 (Connection reset by peer)] 17:26:10 nyef: well, I'd have a better chance of telling what goes on if FROB were named something useful, and there was a comment before the start of the FROB calls that gave an example of what the args are 17:26:18 but maybe that's not what you're asking about 17:26:23 -!- CrazyEddy [n=CrazyEdd@wrongplanet/CrazyEddy] has quit [Read error: 110 (Connection timed out)] 17:26:25 -!- yCrazyEdd is now known as CrazyEddy 17:27:31 amnesiac [n=amnesiac@p3m/member/Amnesiac] has joined #lisp 17:27:44 chandler: Okay, so maybe if I made it a defmacro instead of a macrolet and added a bit more supporting documentation? 17:28:08 epoch [n=K-I-S-S@p3m/member/epoch] has joined #lisp 17:28:14 -!- epoch [n=K-I-S-S@p3m/member/epoch] has quit [Read error: 104 (Connection reset by peer)] 17:28:37 nyef: Are there any bitfield-emitters with holes? (that is, do you need to specify both elements of the byte-spec or just the positions?) 17:28:54 s/positions/sizes/ 17:29:17 s/or just the positions// (: 17:29:21 splittist: There's a check in define-bitfield-emitter to make sure that there are no holes and no overlaps. And the fields can be in any order. 17:29:47 -!- jao [n=user@74.Red-80-24-4.staticIP.rima-tde.net] has quit [Read error: 110 (Connection timed out)] 17:30:15 chandler, I'm firing off a new build of SBCL, unprivileged and without noatime. 17:30:16 *Riastradh* vanishes. 17:30:25 Riastradh: OK, great. 17:34:14 -!- seelenquell [n=seelenqu@pD9E46D9C.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)] 17:36:56 rpg [n=rpg@216.243.156.16.real-time.com] has joined #lisp 17:36:56 -!- matley [n=matley@matley.imati.cnr.it] has quit [Read error: 110 (Connection timed out)] 17:38:23 -!- hosiawak [n=user@host81-137-9-70.in-addr.btopenworld.com] has quit ["ERC Version 5.3 (IRC client for Emacs)"] 17:38:32 nikodemus` [n=nikodemu@cs27013130.pp.htv.fi] has joined #lisp 17:39:07 sykopomp` [n=user@unaffiliated/sykopomp] has joined #lisp 17:40:35 |Soulman| [n=kvirc@87.80-202-238.nextgentel.com] has joined #lisp 17:41:03 -!- Yuuhi [n=user@p5483D4FE.dip.t-dialin.net] has quit ["ERC Version 5.3 (IRC client for Emacs)"] 17:41:32 -!- Odin- [n=sbkhh@s121-302.gardur.hi.is] has quit ["Am I missing an eyebrow?"] 17:41:47 -!- rindolf [n=shlomi@bzq-219-139-216.static.bezeqint.net] has quit ["Yay! I'm a Llama again!"] 17:42:09 -!- mikesch [n=axel@xdsl-78-34-217-150.netcologne.de] has quit [Read error: 113 (No route to host)] 17:43:02 -!- sykopomp [n=user@unaffiliated/sykopomp] has quit [Nick collision from services.] 17:43:07 -!- sykopomp` is now known as sykopomp 17:45:42 schoppenhauer [n=css@unaffiliated/schoppenhauer] has joined #lisp 17:46:10 -!- nikodemus [n=nikodemu@cs27013130.pp.htv.fi] has quit [Read error: 110 (Connection timed out)] 17:47:54 chris2 [n=chris@ppp-88-217-94-251.dynamic.mnet-online.de] has joined #lisp 17:48:48 milanj [n=milan@79.101.169.218] has joined #lisp 17:51:10 Odin- [n=sbkhh@s121-302.gardur.hi.is] has joined #lisp 17:51:25 -!- x6j8x [n=x6j8x@77.21.43.186] has quit [] 17:52:05 Okay, synchronized , so I think I'm done for the day. 17:52:08 Or at least for now. 17:53:42 -!- Adamant [n=Adamant@unaffiliated/adamant] has quit [] 17:54:43 SBCL on mobile phones? Yay! 17:55:07 -!- mega1 [n=mega@pool-016d1.externet.hu] has quit [Read error: 110 (Connection timed out)] 17:58:00 ooh, I'm getting a Pandora (openpandora.org) and it has an ARM Cortex-A8 17:58:17 is there something that checks if a form destructures to a given spec? 17:59:00 destructuring-bind 17:59:09 -!- Soulman [n=kvirc@132.80-202-239.nextgentel.com] has quit [Read error: 110 (Connection timed out)] 17:59:23 -!- milanj [n=milan@79.101.169.218] has quit ["This computer has gone to sleep"] 17:59:58 Xach: right, but I was thinking of DESTRUCTURES-P 18:00:10 guess I'll have to wrap it with a handler-case myself 18:00:18 vasa [n=vasa@mm-72-92-84-93.dynamic.pppoe.mgts.by] has joined #lisp 18:00:28 oudeis_ [n=oudeis@93-173-165-236.bb.netvision.net.il] has joined #lisp 18:04:23 nyef: kudos! 18:05:07 Now I just need to figure out how to do the rest of the instructions... 18:05:30 And I can worry about non-ARMv5TE systems later. 18:05:37 tomoyuki28jp [n=tomoyuki@w221062.ppp.asahi-net.or.jp] has joined #lisp 18:05:45 (Saves a lot of instruction definitions.) 18:06:25 I was wondering --- can anyone point me to some nice examples of use of the sbcl sockets library? I'm not entirely sure how to make a socket to listen... 18:07:05 UnwashedMeme [n=nathan@216.155.97.1] has joined #lisp 18:07:09 -!- nullwork_ [n=nullwork@c-24-245-23-122.hsd1.mn.comcast.net] has quit ["leaving"] 18:07:14 rpg: let me show you something that does that 18:07:19 nyef: you should be able to (with reasonably light kludging) get vectors of unsigned-byte-8 out of the assembler, which you can then copy & disassemble on the target using GDB, if you want to check that your definitions are right 18:07:27 Question regarding elephant. Does anyone know a good way to use auto-incremented integer type in elephant? 18:07:36 hah. oops. i guess my example doesn't use sb-bsd-sockets, it uses zsockets. sorry. 18:07:55 zsockets! 18:07:57 rpg: you could use usocket or iolib 18:08:03 -!- nullwork [n=kyle@c-24-245-23-122.hsd1.mn.comcast.net] has quit [Remote closed the connection] 18:08:24 rpg: it's basically make-inet-socket, socket-bind, socket-listen, socket-accept 18:08:34 rpg: if you're somewhat familiar with how to do it in C on Unix, it's pretty similar. 18:08:42 cracki [n=cracki@41-077.eduroam.RWTH-Aachen.DE] has joined #lisp 18:08:44 <_3b> nyef: x86-64 has 32 registers? 18:08:46 Xach: My C is about 20 years rusty now... 18:09:06 Xach: Just trying to figure out some type signatures. E.g., what 18:09:15 rpg: Here is something I wrote when I was first learning Lisp: http://blog.pettomato.com/index.html?p=17 18:09:22 's a host (for ipv4) 32-bit int? 18:09:39 nikodemus`: I'd actually be happy enough validating on single-instruction segments with sb-rt and known encodings. 18:09:49 Xach: does that mean you're working on a "zbp-sockets"? 18:09:53 rpg: that's covered in the manual decently 18:09:57 nullwork [n=kyle@c-24-245-23-122.hsd1.mn.comcast.net] has joined #lisp 18:09:57 lichtblau: worked. it works fine for me. 18:10:19 lichtblau: sadly in an unreleasable state, and i don't really want to listen to idiots say "WHY DIDN'T YOU JUST USE USOCKET" 18:10:44 weird, that would have been my next question (except in lowercase) :-) 18:11:00 Xach: really? I'm looking at the bind docs, and I see "For the inet family, pass address and port as two arguments;" but no indication of the type.s. 18:11:08 rpg: back up. 18:11:22 _3b: I seem to recall that it does. Two extra bits of register identification with the prefix byte. 18:11:37 -!- xreyes_ [n=xreyes@8.84-48-175.nextgentel.com] has quit [Read error: 60 (Operation timed out)] 18:11:53 lichtblau: clearly that's the distinguishing characteristic 18:12:04 Xach: "# IP addresses and ports are represented in slightly friendlier fashion than "network-endian integers". " 18:12:07 (defvar *qword-register-names* (make-array 32 :initial-element nil)) 18:12:09 rpg: specifically, http://www.sbcl.org/manual/INET-Domain-Sockets.html#INET-Domain-Sockets 18:12:12 seems to be so 18:12:37 rpg: so, "back up and go over" i guess :) 18:13:10 Xach: Any real objection to my submitting a patch that replaces that earlier, vaguer spec, with the actual answer? 18:14:07 rpg: I have no control over it. I think documenting the format for ipv4 addresses in the "inet" section of the manual is not *too* confusing. 18:14:20 milanj [n=milan@212.200.193.217] has joined #lisp 18:14:20 Hrm... No multiply-by-constant instruction due to the presence of shift-and-add and shift-and-reverse-subtract instructions. 18:14:24 -!- gilberth [n=gilbert@c200186.adsl.hansenet.de] has quit [Read error: 110 (Connection timed out)] 18:14:27 epoch [n=K-I-S-S@p3m/member/epoch] has joined #lisp 18:14:39 Xach: That's not what's confusing --- what's confusing is having documentation elsewhere that doesn't specify clearly, and that doesn't point to the real documentation. 18:14:53 I guess that gets any power of two, plus or minus one. 18:15:12 -!- oudeis [n=oudeis@89-138-123-60.bb.netvision.net.il] has quit [Read error: 110 (Connection timed out)] 18:15:16 rpg: Well, I have no objection, and can't influence whether your patch is accepted, rejected, or ignored. If you think it's better, go for it! 18:15:18 Xach: Honestly, it would be better just removing the earlier, vaguer sentence, because it acts as a decoy. 18:15:28 vixey [n=vicky@amcant.demon.co.uk] has joined #lisp 18:15:37 sb-bsd-socket patches are welcome as always 18:15:48 make that "patches are welcome as always" 18:15:52 Man, I hate when papers don't have numbered pages *cough*Pascal*cough* 18:16:42 Which means single-instruction constant-multiply for 1-5, 7-9, 15-17, etc... 18:16:44 Yeah! Get on with the program, use LaTeX like everyone else. 18:17:21 mrsolo [n=mrsolo@nat/yahoo/x-020b8c5e8c2cfbcc] has joined #lisp 18:17:24 -!- splittist [n=splittis@213.235.9.122] has quit ["I'll just non-rhotic off, then"] 18:18:07 sellout: i tend to get a little confused between the number displayed on a PDF page and the physical page count number 18:18:17 sakana [n=itami@h219-110-182-179.catv02.itscom.jp] has joined #lisp 18:18:31 something new from Pascal? 18:18:35 Xach: I never see page count numbers. 18:19:15 -!- oudeis_ [n=oudeis@93-173-165-236.bb.netvision.net.il] has quit [Read error: 110 (Connection timed out)] 18:19:28 tic: Not that new ... Reflection for the Masses. 18:19:43 2 months ago. 18:19:45 Does the sb-bsd-sockets library allow for "you choose me a port"? 18:19:56 <_3b> nyef: which endianness are you targeting? (assuming i remember correctly that arm is switchable) 18:20:14 rpg: use 0 as port 18:21:05 fe[nl]ix: Thanks! 18:21:07 -!- lieven_ [n=lieven@ip-213-49-240-62.dsl.scarlet.be] has quit [Remote closed the connection] 18:21:13 sellout, published on my birthday, yet I haven't read it! Thanks. *reading* 18:23:21 xreyes_ [n=xreyes@8.84-48-175.nextgentel.com] has joined #lisp 18:24:41 -!- milanj [n=milan@212.200.193.217] has quit [Read error: 60 (Operation timed out)] 18:26:28 pchrist [n=spirit@gentoo/developer/pchrist] has joined #lisp 18:26:29 dialtone [n=dialtone@adsl-99-136-101-166.dsl.pltn13.sbcglobal.net] has joined #lisp 18:26:34 dash__ [n=dash@dslb-088-067-001-056.pools.arcor-ip.net] has joined #lisp 18:26:53 josemanuel [n=josemanu@67.1.222.87.dynamic.jazztel.es] has joined #lisp 18:27:30 <_3b> nyef: ah, nevermind, you specified BE in the doc 18:27:34 antgreen [n=green@74.210.24.192] has joined #lisp 18:29:55 milanj [n=milan@79.101.197.67] has joined #lisp 18:32:03 Jikes. Too difficult for me. 18:32:37 ivan4th [n=ivan4th@212.1.228.48] has joined #lisp 18:33:06 lhz [n=shrekz@c-3143e455.021-158-73746f34.cust.bredbandsbolaget.se] has joined #lisp 18:33:32 -!- Odin- [n=sbkhh@s121-302.gardur.hi.is] has quit [] 18:34:32 -!- cracki [n=cracki@41-077.eduroam.RWTH-Aachen.DE] has quit ["The funniest things in my life are truth and absurdity."] 18:37:12 -!- dash_ [n=dash@dslb-084-057-007-110.pools.arcor-ip.net] has quit [Read error: 110 (Connection timed out)] 18:37:14 -!- milanj [n=milan@79.101.197.67] has quit ["This computer has gone to sleep"] 18:37:31 -!- nikodemus` [n=nikodemu@cs27013130.pp.htv.fi] has quit [Read error: 60 (Operation timed out)] 18:39:05 Krystof [n=csr21@84-51-132-95.christ977.adsl.metronet.co.uk] has joined #lisp 18:39:35 nikodemus` [n=nikodemu@cs27013130.pp.htv.fi] has joined #lisp 18:39:41 -!- davazp [n=user@72.Red-88-6-205.staticIP.rima-tde.net] has quit [Remote closed the connection] 18:40:55 -!- sakana [n=itami@h219-110-182-179.catv02.itscom.jp] has left #lisp 18:42:17 Out of curiosity: Why does SBCL want /proc mounted on BSDoid systems? 18:42:55 maybe to find the executable path? 18:43:04 What does it want that for? 18:43:16 for (save-lisp-and-die :executable t) 18:44:29 -!- nikodemus` [n=nikodemu@cs27013130.pp.htv.fi] has quit [Client Quit] 18:44:30 There's now a sysctl() which is the preferred way of finding that now 18:45:19 (on FreeBSD, anyway) 18:45:24 _3b: I expect that most/all of the backend will actually be endian-neutral. 18:46:20 Merging argv[0] with the working directory at the beginning doesn't suffice? 18:47:28 lemonodor [n=lemonodo@66.43.112.62] has joined #lisp 18:47:35 FZ [n=chatzill@unaffiliated/fz] has joined #lisp 18:47:38 that's not reliable 18:48:36 and doing manual PATH handling would suck 18:48:46 ...oh, right. 18:49:19 -!- grkz [n=martin@c-7bf9e255.010-54-6f72652.cust.bredbandsbolaget.se] has quit [Remote closed the connection] 18:49:24 Odin- [n=sbkhh@s121-302.gardur.hi.is] has joined #lisp 18:50:59 oudeis [n=oudeis@bzq-84-108-12-78.cablep.bezeqint.net] has joined #lisp 18:51:02 rindolf [n=shlomi@bzq-219-139-216.static.bezeqint.net] has joined #lisp 18:51:48 matley [n=matley@83.225.100.192] has joined #lisp 18:51:51 -!- silenius [n=jl@yian-ho03.nir.cronon.net] has quit [Remote closed the connection] 18:56:06 milanj [n=milan@79.101.197.67] has joined #lisp 18:57:10 trebor_win [n=none_ask@mail.dki.tu-darmstadt.de] has joined #lisp 18:57:39 hello. 18:58:31 nikodemus [n=nikodemu@cs181229041.pp.htv.fi] has joined #lisp 18:58:34 x6j8x [n=x6j8x@77.21.43.186] has joined #lisp 18:59:01 trebor_win: Hello. 19:00:04 CMUCL ran on the IBM RT? 19:00:14 Too bad I tossed mine. 19:00:34 -!- rread [n=rread@c-76-21-116-77.hsd1.ca.comcast.net] has quit [] 19:00:37 -!- ThomasI [n=thomas@91-64-144-252-dynip.superkabel.de] has quit ["Bye Bye!"] 19:00:50 sellout: Learn something new every day, huh? 19:01:35 that was the original platform, wasn't it? 19:01:53 Note that CMUCL actually has -two- RT backends. 19:01:57 nyef: Nice log, btw. 19:02:01 sellout: Thanks. 19:03:01 I've been thinking about writing a little something-something to add a bit of HTML encoding and produce a feed, but it's mostly "WIBNI" at this point. 19:03:42 -!- robot_jesus [n=csanders@hoovers-241.hoovers.com] has quit [Remote closed the connection] 19:04:14 b4|hraban [n=b4@0brg.xs4all.nl] has joined #lisp 19:11:56 -!- oudeis [n=oudeis@bzq-84-108-12-78.cablep.bezeqint.net] has quit ["This computer has gone to sleep"] 19:12:11 milanj- [n=milan@79.101.197.67] has joined #lisp 19:12:13 Hmm. For some reason, when running the sb-simple-streams test CREATE-READ-MAPPED-FILE-1, SBCL is passing an offset of 0x14c to the mmap system call to map the test file into memory, even though the call in sb-simple-streams/file.lisp to SB-POSIX:MMAP passes zero there. 19:13:04 What platform? 19:13:14 NetBSD/macppc 19:13:27 tltstc` [n=nine@192.207.69.1] has joined #lisp 19:13:31 32-bit or 64-bit? 19:14:09 32-bit 19:14:21 -!- Eleanore [n=a@c-bf70e155.185-1-64736c10.cust.bredbandsbolaget.se] has quit ["aoeu"] 19:14:26 Hunh. 19:14:49 -!- neurogeek_ [n=neurogee@201.208.73.206] has quit [Read error: 110 (Connection timed out)] 19:15:37 If that's a boxed fixnum, it feels like an 87... 19:15:41 neurogeek_ [n=neurogee@201.208.73.206] has joined #lisp 19:15:54 bakkdoor [n=bakkdoor@xdslr056.osnanet.de] has joined #lisp 19:16:30 mmap(2) has a relatively large number of arguments, maybe something's wonky with stack argument passing? 19:16:30 Is it possible to run SBCL under gdb? I'd like to see what argument is being passed to the mmap stub in libc. 19:16:46 It is... sometimes possible. 19:17:04 I haven't had much luck with it the last few times I've tried. 19:19:09 edon [n=edon@albalug/edon] has joined #lisp 19:19:16 Let's see how spectacularly tracing SB-POSIX:MMAP will crash things. 19:19:40 *dlowe* saves his work. 19:19:48 Hmm. It didn't have any observable effect. 19:19:56 Probably inlined? 19:20:01 Could be. 19:20:02 -!- milanj [n=milan@79.101.197.67] has quit [Read error: 110 (Connection timed out)] 19:20:17 -!- S11001001 [n=sirian@pdpc/supporter/active/S11001001] has quit [Read error: 104 (Connection reset by peer)] 19:20:27 S11001001 [n=sirian@74-137-146-187.dhcp.insightbb.com] has joined #lisp 19:20:51 oudeis [n=oudeis@bzq-84-108-12-78.cablep.bezeqint.net] has joined #lisp 19:23:39 vy [n=user@88.231.235.151] has joined #lisp 19:24:45 -!- bakkdoor [n=bakkdoor@xdslr056.osnanet.de] has left #lisp 19:25:04 -!- jgracin [n=jgracin@82.193.208.195] has quit [Remote closed the connection] 19:25:36 -!- pixel5 [n=pixel@copei.de] has left #lisp 19:25:42 I've seen this before 19:26:24 -!- pok [i=pok@tarrant.klingenberg.no] has quit [Read error: 104 (Connection reset by peer)] 19:26:25 it shows up with 64-bit arguments on ppc, let me see if I can dig up more details from when I investigated it 19:26:36 -!- johs [n=johs@hawk.netfonds.no] has quit [Read error: 113 (No route to host)] 19:26:48 actually, I think I posted to the list about it a while ago 19:27:04 -!- FZ [n=chatzill@unaffiliated/fz] has quit [Read error: 110 (Connection timed out)] 19:27:09 photon2 [n=photon@unaffiliated/photon] has joined #lisp 19:28:20 birdsbite [n=user@75.110.164.248] has joined #lisp 19:30:08 -!- Adrinael_ is now known as Adrinael 19:30:11 -!- dthomp [n=dat@dyn-188-dynamic.linfield.edu] has quit [Read error: 110 (Connection timed out)] 19:30:37 http://marc.info/?l=sbcl-devel&m=120967647930096&w=2 19:32:24 johs [n=johs@hawk.netfonds.no] has joined #lisp 19:32:24 Oh: the file in question is eighty-three bytes long. 19:32:33 pok [i=pok@tarrant.klingenberg.no] has joined #lisp 19:32:42 ... Did I say 87? I meant 83. 19:33:20 nobody replied so I sort of shoved it onto the back shelf, but maybe you'll have better luck figuring it out 19:33:58 Alien-types are, frankly, frightening. 19:34:29 But the core bit is probably that which surrounds %alien-funcall, either as a VOP or as a DEFTRANSFORM. 19:34:34 -!- duck1123 [i=duck@adsl-75-46-28-24.dsl.sfldmi.sbcglobal.net] has quit [Read error: 104 (Connection reset by peer)] 19:34:42 it's certainly confusing to figure out how they were compiled 19:34:43 deflate [n=me@79.173.195.58] has joined #lisp 19:34:45 dkcl [n=dkcl@115.66.218.87.dynamic.jazztel.es] has joined #lisp 19:34:46 ... I vaguely remember this for some reason. 19:35:00 but that was almost a year ago, and I don't have time to track it down now 19:35:07 FZ [n=chatzill@78.184.235.213] has joined #lisp 19:35:12 Curiously, the *size* argument passed to the mmap system call was 0x53, which is far from the size of the file. 19:35:15 perhaps you remember my sbcl-devel post from march 19:35:41 -!- photon2 is now known as photon 19:35:41 ...um. 19:35:44 Please excuse me. 19:36:02 I see. 19:36:20 The size is correct, and it looks like mmap is getting an extra argument from the stack somehow. 19:36:29 Possibly SBCL is confused about the size of an off_t? 19:36:36 -!- deflate [n=me@79.173.195.58] has left #lisp 19:36:39 plausible 19:36:43 Oh, that's highly pluasible. 19:36:47 er 19:37:02 that url I pasted was in relation to your problem Riastradh 19:37:02 Is off_t 32 or 64 bits? 19:37:19 I believe off_t has been 64-bit for decades in BSD. 19:37:37 Are 64-bit values required to be 64-bit aligned on 32-bit platforms? 19:37:40 yes, and I also saw this problem in linux with a 64-bit off_t iirc 19:37:51 but not on darwin 19:38:17 Darwin's off_t is not 64-bits?? 19:38:19 Ooh. Or is any alignment requirement host-OS dependent? 19:38:19 off-t is groveled, though 19:38:22 ...or you didn't see the problem? 19:38:26 (in Darwin) 19:38:30 no, I think the darwin ppc stack works differently 19:38:35 njsg [n=njsg@unaffiliated/njsg] has joined #lisp 19:38:45 booyaa [n=booyaa@adsl-67-121-157-251.dsl.pltn13.pacbell.net] has joined #lisp 19:38:50 I think it is 64-bit, but I didn't see the problem 19:38:58 (Sorry, I don't have a web browser handy, so I didn't look at the message.) 19:39:00 Darwin stack alignment is different from the ELF ABI. 19:39:23 Riastradh: ah 19:40:21 well from what I remember, when an alien function is defined with an (unsigned 64) argument following a system-area-pointer then the (unsigned-64) gets shifted forward or back by 32-bits from what it should be 19:40:28 Darwin's stack has to be 16-byte aligned at the point of a procedure call, or it can get cranky. 19:41:12 Darwin has a separate deftransform for %alien-funcall. 19:41:45 Riastradh: can you check what the correct argument sizes for mmap are there, and compare to what sb-posix thinks they are? 19:42:13 -!- oudeis [n=oudeis@bzq-84-108-12-78.cablep.bezeqint.net] has quit [Read error: 113 (No route to host)] 19:42:27 How do I find what sb-posix thinks the argument sizes are? 19:42:59 Look at the grovelled function definition? 19:43:04 oh, it looks like at the end of that somewhat incoherent email I have a regression test patch that demonstrates the problem 19:43:10 -!- hugod_ [n=hugo@bas1-montreal50-1279442362.dsl.bell.ca] has quit [] 19:43:21 macroexpand the (define-call ("mmap" ...) ...) in contrib/sb-posix/interface.lisp 19:43:43 off-t is in constants.lisp-temp in the same directory 19:44:10 -!- jeremiah [n=jeremiah@31.Red-213-98-123.staticIP.rima-tde.net] has quit [Read error: 104 (Connection reset by peer)] 19:44:10 constants.lisp-temp reports that OFF-T has sixty-four bits. 19:44:16 I don't know if it still applies, but you could try applying this and check of foreign.test.sh passes: http://marc.info/?l=sbcl-devel&m=120967647930096&q=p5 19:44:35 oudeis [n=oudeis@p11811120.orange.net.il] has joined #lisp 19:45:00 nyef or nikodemus: any chance I could get a victim to test the new procedure for updating the SBCL web site? 19:45:09 -!- Bzek [n=SK_sj@mcc-dyn-31-225.kosnet.ru] has quit ["L o v e."] 19:45:17 chandler pasted "instructions for updating the SBCL web site" at http://paste.lisp.org/display/69649 19:45:41 just testing the first part of the procedure is good enough 19:46:36 -!- oudeis [n=oudeis@p11811120.orange.net.il] has quit [Client Quit] 19:46:41 are file-descriptors 32 bits there? 19:46:41 ... In src/compiler/ppc/c-call.lisp, in deftransform %alien-funcall, there is a comment ";; Need to pad for alignment." 19:46:54 chandler: tomorrow 19:47:03 nikodemus: OK 19:47:13 -!- antgreen [n=green@74.210.24.192] has quit [No route to host] 19:47:30 Is the logic, both above and below (the WHEN and the IF) correct? 19:47:58 I remember looking at that code before, but I can't say now 19:48:00 just about to log off 19:48:04 myrkraverk [n=johann@unaffiliated/myrkraverk] has joined #lisp 19:48:19 ok, later -> 19:48:23 -!- nikodemus [n=nikodemu@cs181229041.pp.htv.fi] has quit ["Leaving"] 19:48:37 ecraven [n=nex@plc31-103.linzag.net] has joined #lisp 19:49:29 Ah, args are passed in regs with spill space reserved on the stack? 19:49:54 there is a good summary PDF of the ABI from IBM that I found somewhere 19:49:58 let me see if I can dig that up... 19:49:59 yea, I think they spill after 6 or 8 arguments 19:50:22 chandler, so, do you want this binary -- just as broken in this respect as the existing one, I'm sure --, or shall we wait until SBCL's argument passing is fixed? 19:50:35 Riastradh: does it have SB-POSIX? 19:50:40 It has SB-POSIX. 19:50:56 sb-posix with broken mmap 19:51:00 Right. 19:52:10 ehu [i=52aa21ad@gateway/web/ajax/mibbit.com/x-9cbebc72113684ed] has joined #lisp 19:52:44 -!- rme [n=rme@pool-70-104-125-111.chi.dsl-w.verizon.net] has left #lisp 19:53:02 If you do want it, it's in the same place as before. 19:53:57 actually, this is what I was remembering: http://www.cloudcaptech.com/MPC555%20Resources/Programming%20Environment/SVR4abippc.pdf 19:54:10 I'm pretty sure NetBSD follows this ABI too. 19:54:28 (it has to, or our alien calls wouldn't be working at all!) 19:54:37 milan [n=milan@212.200.192.182] has joined #lisp 19:54:58 Linux and OpenBSD both seem to use the same ABI, so NetBSD probably does too 19:56:17 -!- rdd [n=user@c83-250-142-219.bredband.comhem.se] has quit [Read error: 104 (Connection reset by peer)] 19:56:18 oudeis [n=oudeis@p11811120.orange.net.il] has joined #lisp 19:57:30 -!- tcr [n=tcr@host145.natpool.mwn.de] has quit ["Leaving."] 19:57:56 ... why is there separate logic here for long longs and doubles? 19:58:02 gilberth [n=gilbert@d124035.adsl.hansenet.de] has joined #lisp 19:58:17 regarding stack alignment, that is 19:58:19 -!- cmm [n=cmm@bzq-79-182-117-83.red.bezeqint.net] has quit [Remote closed the connection] 19:58:31 cmm [n=cmm@bzq-79-182-117-83.red.bezeqint.net] has joined #lisp 20:00:07 well, this explains why I'm not seeing a problem on linux/ppc: off_t appears to be 32-bit on my system 20:00:23 tcr [n=tcr@host145.natpool.mwn.de] has joined #lisp 20:00:34 ...yes, Linux is schizophrenic about off_t. 20:00:44 score one for NetBSD 20:00:46 It's pretty annoying. 20:01:01 For all BSDs -- as I said, off_t has been 64-bit for decades, as far as I know. 20:01:19 chandler: Separate logic because doubles are in fprs and long longs in gprs? 20:01:33 nyef: yeah. I realized what I said was stupid as soon as I said it. 20:01:39 jeremiah [n=jeremiah@31.Red-213-98-123.staticIP.rima-tde.net] has joined #lisp 20:01:43 Nonetheless, the stack alignment should be identical. 20:02:28 REPLeffect [n=REPLeffe@69.54.115.254] has joined #lisp 20:02:35 as far as runing ansi-tests ABCL seems as fast as SBCL.. what is it that ABCL is much slower on? 20:03:09 is SBCL slower than say CLISP? 20:03:20 -!- milanj- [n=milan@79.101.197.67] has quit [Read error: 110 (Connection timed out)] 20:03:53 running ansi tests is not really representative of many things you would do normally 20:04:08 sbcl is not slower than CLISP for many tasks 20:04:09 SBCL seems better with heap/memory than CLISP .. like clisp at least the version that ships with "yum install clisp" 20:04:29 I'm surprised about that. 20:04:44 -!- elurin [n=user@81.213.203.109] has quit [Read error: 110 (Connection timed out)] 20:04:53 supprised about the ansi-tests? 20:04:55 There are a lot of situations in running the ANSI tests where compiling things is a performance loss, because things are only called once. 20:04:58 No, the memory thing. 20:05:27 I think SBCL runs the ANSI tests faster if you use the interpreter, too. Try (setf sb-ext:*evaluator-mode* :interpret) 20:05:29 oh well the memory thing has to do with the ansi tests causing heap overflow in clisp in ansi-tests 20:06:09 Umm... 20:06:30 ok going to try SBCL with (setf sb-ext:*evaluator-mode* :interpret) .. perhaps SBCL was slowing down becasue it was trying to compile everything 20:06:52 chandler: This document seems to say that if there's a long long argument followed by a 32-bit argument and there's only 1 GPR left, the long long ends up on the stack and the 32-bit arg ends up in a register. 20:06:55 I'm not sure what to say about that. The default CLISP heap could be too small, or there could be a bug in CLISP that's causing an infinite allocation loop when running the tests. 20:07:26 rdd [n=user@c83-250-142-219.bredband.comhem.se] has joined #lisp 20:07:35 nyef: which page are you on? 20:07:47 3-19 and 3-20. 20:07:58 yeah i tried to pass the commandline argument to increase he heap for clisp but still getting the eroro 20:08:29 -!- benny [n=benny@i577A0519.versanet.de] has quit [Read error: 104 (Connection reset by peer)] 20:09:10 pass me a URL to page 3-19 if you can 20:09:12 bertskert [n=mor_och_@c83-252-190-193.bredband.comhem.se] has joined #lisp 20:09:24 dmiles_afk: http://www.cloudcaptech.com/MPC555%20Resources/Programming%20Environment/SVR4abippc.pdf 20:10:28 Sinden [n=Sinden@nickweinhold.demon.co.uk] has joined #lisp 20:10:57 If I write a function that manipulates some list, and I want to write a test function that tests the first one with test-cases in a list returning the test-cases that went wrong.. how would I write that test function? Wouldn't I have to use the same method/algorithms to maipulate the data inside that one as in the real function? 20:11:38 How would the test function know if the "answer" is correct? 20:11:46 bertskert: Use known data. 20:11:48 gdmfsob [n=gdmfsob@92.112.107.168] has joined #lisp 20:11:53 true, thx! 20:12:12 Either you have a separate algorithm for your oracle, or you are comparing against known data. 20:13:14 disumu [n=disumu@p57A25119.dip.t-dialin.net] has joined #lisp 20:14:49 hrrm i realize the stack dumps of CLISP i can get past them for 409 out of 21693 total tests failed: 20:15:00 The next possibility is that the stack args aren't getting spilled properly. This might not get noticed very much, as 8 arg passing regs is sufficient for a lot of APIs. 20:15:13 that number should be more like 50 right? 20:15:24 -!- ivanst [i=ivans@93-138-56-95.adsl.net.t-com.hr] has quit ["Leaving"] 20:15:42 -!- dialtone [n=dialtone@adsl-99-136-101-166.dsl.pltn13.sbcglobal.net] has quit ["leaving"] 20:16:00 dmiles_afk: It ought to be zero. 20:16:17 nyef: the problem I saw occurred with only one pointer and one 64-bit int arg 20:16:29 bertskert: If you can write the inverse function, you can test for f o f^{1} == id. (Having the inverse function may be to strong, but perhaps you have other functions which together with your function hold some properties.) 20:16:41 oops that url chandler, thats not a CLISP doc right? 20:16:59 dmiles_afk: No, it had nothing to do with your conversation. I wasn't sure why you asked me for it. 20:17:08 :) 20:17:14 it seemed the same regardless of where the 64-bit arg was, as long as it followed a s-a-p 20:17:16 Hrm... 20:17:33 but only after a sap? Odd. 20:17:36 ivanst [i=ivans@93-138-56-95.adsl.net.t-com.hr] has joined #lisp 20:17:37 No, even. 20:17:42 Grr. 20:17:46 Is this a 64-bit userland? 20:18:05 Shouldn't have any effect, as SBCL/PPC is 32-bit only. 20:18:07 (I'm sure I asked this earlier.) 20:18:08 no, 32-bit on 32-bit hardware 20:18:17 -!- Sinden [n=Sinden@nickweinhold.demon.co.uk] has quit ["Leaving"] 20:18:32 milanj- [n=milan@212.200.192.182] has joined #lisp 20:18:52 joshe: so a uint32_t followed by a uint64_t works just fine, but a SAP followed by uint64_t bombs? 20:18:52 chandler: Could affect the size expected of a machine pointer at the alien interface. 20:18:59 -!- ivanst [i=ivans@93-138-56-95.adsl.net.t-com.hr] has quit [Client Quit] 20:19:03 chandler: from what I remember, yes 20:19:08 nyef: I'm not seeing how it could. 20:19:41 ivans [i=ivans@93-138-56-95.adsl.net.t-com.hr] has joined #lisp 20:20:08 -!- ivans [i=ivans@93-138-56-95.adsl.net.t-com.hr] has quit [Client Quit] 20:20:35 -!- brickhazel [n=brickhaz@ool-45717ef5.dyn.optonline.net] has quit [Read error: 104 (Connection reset by peer)] 20:21:19 dialtone [n=dialtone@adsl-99-136-101-166.dsl.pltn13.sbcglobal.net] has joined #lisp 20:21:25 Let me know if there's anything you'd like me to test on this machine. 20:22:12 Hm. There might be in a minute. 20:22:46 -!- matley [n=matley@83.225.100.192] has quit [Read error: 110 (Connection timed out)] 20:22:48 hkBst [n=hkBst@gentoo/developer/hkbst] has joined #lisp 20:22:53 also if anyone has time.. if they can run ansi-tests on clisp and tell me if it gets through them 20:23:30 jao [n=user@72.Red-79-155-244.staticIP.rima-tde.net] has joined #lisp 20:23:56 hrrm (setf sb-ext:*evaluator-mode* :interpret) slowed the ansi tests in SBCL 20:24:26 151secs -> 309secs 20:24:40 well, nevermind that then! 20:25:11 when i am comparing SBCL to ABCL its int eh newest java 1.7 20:25:33 I think I found it. 20:25:36 ABCL is 141secs 20:26:18 this isnt a great benchmark.. defiantely TAK or something else is better 20:26:20 In the COND that does the padding and whatnot based on arg type. 20:26:47 Have a look at the last case, and what it doesn't do for gprs / fprs / stack. 20:27:16 (Oh yeah, and this entire approach is demented. WTF?) 20:27:50 Urgh. I don't think I want to look at that code! 20:28:13 -!- milan [n=milan@212.200.192.182] has quit [Read error: 110 (Connection timed out)] 20:28:39 nyef: Yes, I was just arriving at that same conclusion myself. 20:29:48 and of course this is invisible most of the time because it *only* fires when there's a long long argument in the function's arglist 20:29:53 silenius [n=jl@dslb-088-073-082-208.pools.arcor-ip.net] has joined #lisp 20:30:22 No, it only fires when there's a non-integer-or-float argument before a long long. 20:31:02 -!- Lycurgus [n=juan@cpe-72-228-150-44.buffalo.res.rr.com] has quit [Remote closed the connection] 20:31:08 -!- oudeis [n=oudeis@p11811120.orange.net.il] has quit ["This computer has gone to sleep"] 20:31:14 nyef: look at the IF surrounding the body of the deftransform 20:31:53 chandler: Yes, I'm not entirely discounting your criteria, I'm supplying tighter criteria. 20:32:28 I think it's buggy no matter where the SAP appears in the arglist, as long as there's a long long argument somewhere. 20:33:27 Umm... No, because two SAPs will throw it back on. 20:33:58 And if what follows a single SAP is 32-bit arguments, it won't break either. 20:34:39 does anyone know of a scalar vector machine (SVM) implementation for Lisp? 20:34:50 -!- lichtblau [n=user@wallstrasse.knowledgetools.de] has quit [Read error: 60 (Operation timed out)] 20:35:00 And for more fun, there may be ways to tickle this problem with types other than SAPs. 20:35:17 gonzojive: i don't, but i saw your earlier question about ZS3, and the answer is: http://xach.com/lisp/zs3/ 20:35:40 nyef: since we don't support passing structs by value atm, I thought everything was int, float, or SAP 20:35:44 At least we don't support passing struc... yeah. 20:35:58 Xach: cool, thanks for letting me try it early. it came just in time 20:36:01 much to vixey's dismay 20:36:16 Xach, what do you want? 20:36:24 gonzojive: the version you got is somewhat different from the final version, interface-wise. 20:36:38 Of course, if we -did- support structs by value, I expect that this entire bit of demented logic would go away anyway, as we'd need to support multiple arg-tns for a single argument anyway, and we only do that for up to two values for a return arg. 20:36:59 I implemnted structs as values for SBCL on ppc 20:37:08 chandler pasted "bugged, even when there's an int32 between the sap and the int64" at http://paste.lisp.org/display/69651 20:37:42 nyef: or *cough* they could be implemented using this demented logic :-) 20:38:33 Xach, happy? 20:38:59 chandler: Of course it's still bugged then. But is it still bugged with two SAPs? 20:39:28 Xach, what do you want? 20:40:40 And structure pass-by-value would be implemented as :arg-tn alien-type-methods, not this demented junk. 20:41:11 Yes, I was joking. 20:41:32 -!- tmh [n=thomas@pdpc/supporter/sustaining/tmh] has left #lisp 20:42:19 vixey: Out of curiousity, how did you end up implementing it? 20:42:24 I don't know 20:42:32 I just hacked until it worked 20:42:41 *chandler* raises an eyebrow 20:42:52 *vixey* is super interested in your eyebrow 20:42:56 Let me guess... This didn't actually end up in mainline SBCL, did it? 20:43:05 no my patch got ignored in the mailing list 20:43:12 given how the existing logic doesn't work, I'm not sure how you can claim that your code worked either 20:43:49 nyef: OK. I didn't expect it to work with two SAPs. What am I missing. 20:43:51 it's presumably still up there somewhere, but I'm sure there's nothing valuable in that code 20:43:57 -!- alley_cat [i=AlleyCat@sourcemage/elder/alleycat] has quit [Remote closed the connection] 20:44:22 Oh, I see. It makes it "aligned" again. 20:44:26 chandler: It's a state machine that is supposed to add padding when there's an odd number of prior arguments. 20:44:28 I only wanted my library to work,.. to implement this properly you have to rip out the whole crap FFI system in SBCL and reimplement it 20:44:29 So, yeah. 20:45:19 vixey: Fair enough. Thanks. We're just taking a look at a bug in part of the crap FFI system in SBCL now, and discussing implications. 20:46:30 is SBCL as fast as CMUCL ? 20:46:38 dmiles_afk: yes 20:46:42 (an odd number of registers used for prior arguments) 20:47:16 Xach, ah good.. and Allegro? 20:47:27 Riastradh: No, supposed to be 8-byte aligned on stack as well. 20:48:04 they use Allegro for some code i am getting the next few days and i am hoping to use SBCL on it if not ABCL 20:48:06 Sorry: an odd number of machine words, either in registers or on the stack, used for prior arguments. 20:48:26 dmiles_afk: to get a useful answer, you generally have to answer "at what?" 20:48:38 code tends to be fastest on the implementation it was optimized for 20:49:10 (I cant afford to purchase Allegro) .. at what.. probly implemeting a OODB with speical inerrprer 20:49:15 every time I write something in good faith and try to run it on non-sbcl/cmucl implementations, it's painfully slow 20:49:35 and I have no idea of what kind of hoops to jump through to make it fast 20:49:50 Xach, hosting an expoert system with a OODB 20:50:04 it really depends on what the code's doing 20:50:21 and the same thing can be observed with code originally written for gcl and originally written for acl 20:50:38 so in general .. probly runing tight looped unification like prolog 20:51:12 So, according to this ABI spec, all structs passed by value are actually passed as a pointer to a copy of the struct. 20:51:41 which ABI is that? 20:52:03 -!- epoch [n=K-I-S-S@p3m/member/epoch] has quit ["  "] 20:52:14 nyef, can the existing demented logic be made to work easily, before thinking about much more elaborate changes such as structs? 20:52:16 vixey: the SVR4 API for PowerpC 20:54:44 Riastradh: Yes. Find "(let ((new-result-type". Five lines above should be the start of the T clause of a COND, with a body of (new-args arg) (new-arg-types type). Add (if (< gprs 8) (incf gprs) (incf stack)) to that clause and try it. 20:55:03 -!- fe[nl]ix [n=algidus@88-149-209-64.dynamic.ngi.it] has quit ["Valete!"] 20:56:58 Dr. Demento pasted "SAP alignment of DOOOM" at http://paste.lisp.org/display/69653 20:57:06 in easy patch form, if you like 20:57:30 fe[nl]ix [n=algidus@88-149-209-64.dynamic.ngi.it] has joined #lisp 20:57:38 dthomp [n=dat@dyn-188-dynamic.linfield.edu] has joined #lisp 20:57:38 Hmm, I guess `sh make.sh' from the beginning after making that change without deleting any FASLs won't do the right thing. 20:57:41 Oh well. 20:57:50 ... it ought to. 20:58:07 It just won't be any faster than doing it in a clean tree. 20:58:14 brickhazel [n=brickhaz@63.144.132.78] has joined #lisp 20:58:15 *Riastradh* coughs. 20:58:24 If it's not, somethings broken. 20:58:35 The `right thing' here is not rebuilding more than need be rebuilt. 20:58:44 hugod [n=hugo@modemcable187.149-200-24.mc.videotron.ca] has joined #lisp 20:58:59 (Of course, I suppose this is hard with a bootstrapping build to do right.) 20:59:01 dmiles_afk, is your keyboard broken? 20:59:09 Well, you can always try your luck with slam.sh. 20:59:15 Too late! 20:59:17 patmatch [n=wmbot@dsl093-216-036.aus1.dsl.speakeasy.net] has joined #lisp 21:00:04 patmatch: should I be worried that your IRC username is "wmbot"? 21:00:06 jpcooper, my nick means i am only here for when i want service and not when others want service 21:00:45 dmiles_afk, well I can't counter your wit 21:00:46 dmiles_afk: I wonder sometimes if your keyboard is broken because you rarely put together a sentence without multiple typing errors. 21:01:07 Xach, maybe he switched to dvorak 21:01:11 actually though whenever my irc client disconnects and uses auto-connect it changes my nick to _afk 21:01:19 chandler: FYI, mips-linux is also on 1.0.22 now, just in case you want to do another webpage update. :-) 21:01:31 Younder complained about the unfairness of it all. 21:01:39 guys move to #meta-lisp where you can talk about people being annoyingly meta and low SNR 21:01:40 ths_: you know, I've been looking for a victim to try http://paste.lisp.org/display/69649 ... 21:01:50 -!- vy [n=user@88.231.235.151] has quit [Remote closed the connection] 21:02:17 -!- gdmfsob [n=gdmfsob@92.112.107.168] has quit [Connection timed out] 21:02:21 -!- rindolf [n=shlomi@bzq-219-139-216.static.bezeqint.net] has quit ["Yay! I'm a Llama again!"] 21:02:26 -!- tst___ [n=Tim@p4FD2C829.dip.t-dialin.net] has quit ["leaving"] 21:02:30 patmatch: You're not exactly passing the turing test. 21:02:47 -!- ChanServ has set mode +o chandler 21:02:56 -!- chandler has set mode +b *!n=wmbot@* 21:03:03 -!- chandler [n=chandler@opendarwin/developer/chandler] has been kicked from #lisp 21:03:05 -!- chandler has set mode -o chandler 21:03:45 ths_: Feel like being a guinea pig, or shall I update the page? 21:04:52 I have little time right now, and am on vacation the next two weeks. 21:05:06 So please go ahead. 21:05:10 `Vacation' means `Lisp-hacking time', right? 21:05:14 Sure, no problem. This is just the big-endian binary, right? 21:05:32 Riastradh: Qemu hacking time, if any. 21:05:33 H4ns [n=hans@p57A0C101.dip.t-dialin.net] has joined #lisp 21:05:35 -!- gonzojive [n=red@fun.Stanford.EDU] has quit [] 21:05:37 chandler: Yes. 21:06:00 Riastradh: oh, that reminds me: I'm very amused by that GXEmul bug. 21:06:02 Hmmm, Qemu. Some day I ought to send in the patches I needed to apply to make Qemu run on NetBSD/macppc. 21:06:03 manuel_ [n=manuel@HSI-KBW-085-216-061-130.hsi.kabelbw.de] has joined #lisp 21:06:24 -!- tomoyuki28jp [n=tomoyuki@w221062.ppp.asahi-net.or.jp] has left #lisp 21:07:39 Riastradh: You don't happen to have a SPARC/Solaris system still, do you? 21:07:51 I remember you actually giving me an account on one at some point. 21:08:15 I have one. It's capable of building T; SBCL would be a trifle more taxing. I completely gave up trying to build MIT Scheme on it. 21:10:52 ccl-logbot [n=ccl-logb@master.clozure.com] has joined #lisp 21:10:52 21:10:52 -!- names: ccl-logbot V-ille NoorDextor Soulman__ herbieB z0d manuel_ H4ns hugod brickhazel dthomp fe[nl]ix silenius jao hkBst dialtone milanj- disumu bertskert rdd REPLeffect jeremiah tcr cmm gilberth ehu ecraven myrkraverk booyaa njsg FZ dkcl pok johs birdsbite photon S11001001 edon neurogeek_ tltstc` b4|hraban x6j8x trebor_win Odin- lemonodor Krystof lhz ivan4th josemanuel dash__ pchrist xreyes_ mrsolo vixey nullwork UnwashedMeme vasa chris2 schoppenhauer 21:10:52 -!- names: |Soulman| sykopomp rpg amnesiac seelenquell_ CrazyEddy michaelw H4ns1 srcerer acrid mathrick ia Wombat1 jlf` ebzzry jkantz delYsid olejorgenb p8m ZabaQ mikezor pitui sabetts alexsuraci ryepup1 ``Erik_ yahooooo3 attila_lendvai sellout c|mell cemerick willb mindCrime sohail dcrawford manic12 adeht matimago tic ineiros_ Adrinael Nshag daniel Piranha__ tritchey md3 hugo malumalu LiamH Hun Beket novaburst segv_ locklace Eno_ dlowe Jasko StasNev 21:10:52 -!- names: araujo jpcooper weirdo fdr- antoszka pchrist|univ intrados1 kuwabara xan_ pierre_thierry froog mejja kib2 xjrn bob_f VityokOrgUa arbscht knobo spacebat abeaumont ushdf stassats` fihi09 timchen1` replor_ kidd mogunus beach hsaliak mtd bpt dto _8david joshe gigamonkey` slyrus_ fisxoj mgr kpreid pragma__ faheem_ technik_ specbot minion lisppaste mdxi egn andrewy schme_ lemoinem hefner Modius__ zenbalrog spiderbyte jsnell Buganini bohanlon djinni` 21:10:52 -!- names: maxote `nipra pinterface Aisling lyte Maghnus dfox_ kuhzoo Cel meingbg matthew` billstclair ``Erik lnostdal sad0ur xinming zbigniew BrianRice brandelune eno abend_ gigamonkey housel prip ahaas Chrononaut enn persi jajcloz cods boyscared dmiles_afk PissedNumlock eevar cmeme scode Bucciarati mornfall thijso kmkaplan jrockway sbok Tristam esden Patzy djkthx eirik puddingpimp yango larstobi ivarref e271 jmcphers lucca Cryovat tarbo isomer proq 21:10:52 -!- names: defn joga jollygood_______ guenther__ Guest53748 aking wgl merlincorey tessier Xof spiaggia thedonvaughn felipe Partyzant rlpowell mcxx Khisanth Zhivago jsimonss vcgomes nyef ianmcorvidae r0bby turbo24prg bascule m4thrick Ifur phadthai bfein foom clog mvilleneuve bunz esden`away mqt albino fnordus _3b esden_ maskd andrerav Xach dublpaws rumbleca jamesjb dostoyevsky Fade chii jolby chandler Balooga rsynnott emma Ash qebab rey_ pkhuong antifuchs 21:10:52 -!- names: gz bdowning froydnj keithr _CitizenKane_ retupmoca Draggor Thas Riastradh erg agemo xristos authentic luis DrForr tltstc sjbach bougyman slyrus azuk drewc 21:10:53 I do, however, have access to other SPARC/Solaris machines, which I don't own, but which I can build SBCL on. 21:11:15 (...other machines which are just a trifle beefier.) 21:11:22 rme_ [n=rme@pool-70-106-128-43.chi01.dsl-w.verizon.net] has joined #lisp 21:11:48 -!- rme_ [n=rme@pool-70-106-128-43.chi01.dsl-w.verizon.net] has left #lisp 21:14:25 borism [n=boris@195-50-199-64-dsl.krw.estpak.ee] has joined #lisp 21:16:27 vmCodes [n=vmCodes@92.41.191.226.sub.mbb.three.co.uk] has joined #lisp 21:16:56 -!- malumalu [n=malu@hnvr-4dbbae52.pool.einsundeins.de] has quit ["Verlassend"] 21:17:09 Riastradh: Just getting an updated version on the site would be worth it, if you'd be willing. 21:17:37 matley [n=matley@83.225.53.145] has joined #lisp 21:17:46 I would still like to find a method to build it myself ultimately, but for now I am sick of seeing >2 year old builds. 21:19:15 -!- fisxoj [n=fisxoj@149.43.108.7] has quit ["Ex-Chat"] 21:19:35 I'll do it tonight, if I remember, when no one is using those machines. 21:19:50 (and if SBCL 0.9.11 is capable of building 1.0.22...!) 21:20:45 I built sbcl on solaris recently and I think the binary on the download page worked fine 21:20:51 Riastradh, otherwise you'll have to build each successively! 21:21:27 gonzojive [n=red@DNab422456.Stanford.EDU] has joined #lisp 21:21:38 nonsense. otherwise, you'll cross-build from something that has a more up to date binary 21:21:51 cbrannon [n=user@ip68-12-115-220.ok.ok.cox.net] has joined #lisp 21:23:25 ths [n=ths@X4d78.x.pppool.de] has joined #lisp 21:23:42 -!- H4ns1 [n=hans@p57A0D51F.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)] 21:25:11 (I don't think I have access to any Solaris/x86 hardware, however.) 21:25:16 -!- x6j8x [n=x6j8x@77.21.43.186] has quit [] 21:27:09 Riastradh: don't worry about that: I'm fighting with it in Fusion right now. 21:29:42 x6j8x [n=x6j8x@77.21.43.186] has joined #lisp 21:30:46 -!- ecraven [n=nex@plc31-103.linzag.net] has quit ["bbl"] 21:31:12 oudeis [n=oudeis@p11811112.orange.net.il] has joined #lisp 21:31:21 chandler, all tests passed on NetBSD/macppc. 21:31:35 Hey, fantastic. 21:32:19 I can make another new binary distribution with the change to src/compiler/ppc/c-call.lisp, if you'd like -- but it wouldn't match the distributed 1.0.22 sources. 21:32:36 No, that's fine. It's not like the next release is more than a few weeks away. 21:32:42 OK. 21:32:43 -!- Beket [n=stathis@athedsl-4415404.home.otenet.gr] has quit ["Leaving"] 21:37:28 ace4016 [i=ace4016@cpe-76-168-248-118.socal.res.rr.com] has joined #lisp 21:38:59 -!- md3 [n=user@85-135-131-179.adsl.slovanet.sk] has quit [Read error: 60 (Operation timed out)] 21:39:17 md3 [n=user@85-135-174-17.adsl.slovanet.sk] has joined #lisp 21:39:40 fisxoj [n=fisxoj@149.43.108.7] has joined #lisp 21:40:55 -!- lhz [n=shrekz@c-3143e455.021-158-73746f34.cust.bredbandsbolaget.se] has quit ["Leaving"] 21:44:58 nyef, Riastradh, joshe: 1.0.22.14 should be a little less bogus (though no less demented, I'm afraid) 21:46:06 thanks, I just saw the commit 21:46:18 -!- kpreid [n=kpreid@cpe-67-242-12-135.twcny.res.rr.com] has quit [Remote closed the connection] 21:46:43 ah, and a build of my old openbsd ppc branch with that c-call.lisp change just finished successfully 21:46:44 -!- vmCodes [n=vmCodes@92.41.191.226.sub.mbb.three.co.uk] has left #lisp 21:47:15 kpreid [n=kpreid@cpe-67-242-12-135.twcny.res.rr.com] has joined #lisp 21:47:20 -!- fisxoj [n=fisxoj@149.43.108.7] has quit [Remote closed the connection] 21:47:59 -!- Jasko [n=tjasko@209.74.44.225] has quit [Read error: 110 (Connection timed out)] 21:48:07 fisxoj [n=fisxoj@149.43.108.7] has joined #lisp 21:48:51 milanj [n=milan@212.200.192.182] has joined #lisp 21:49:08 rvirding [n=chatzill@h235n3c1o1034.bredband.skanova.com] has joined #lisp 21:49:31 Jasko [n=tjasko@c-98-235-21-22.hsd1.pa.comcast.net] has joined #lisp 21:49:53 Sinden [n=Sinden@nickweinhold.demon.co.uk] has joined #lisp 21:49:55 silenius_ [n=jl@dslb-088-073-082-208.pools.arcor-ip.net] has joined #lisp 21:50:33 -!- sellout [n=greg@guest-fw.dc4.itasoftware.com] has quit [] 21:51:01 -!- silenius [n=jl@dslb-088-073-082-208.pools.arcor-ip.net] has quit [Read error: 104 (Connection reset by peer)] 21:53:22 -!- rlpowell [n=rlpowell@chain.digitalkingdom.org] has quit ["leaving"] 21:55:12 -!- LiamH [n=none@common-lisp.net] has quit ["Leaving."] 21:55:16 -!- milanj- [n=milan@212.200.192.182] has quit [Read error: 110 (Connection timed out)] 21:56:34 -!- rdd [n=user@c83-250-142-219.bredband.comhem.se] has quit [Read error: 104 (Connection reset by peer)] 21:56:57 rlpowell [n=rlpowell@chain.digitalkingdom.org] has joined #lisp 21:58:44 -!- dlowe [n=dlowe@ita4fw1.itasoftware.com] has quit ["*poof*"] 21:59:49 davazp [n=user@72.Red-88-6-205.staticIP.rima-tde.net] has joined #lisp 22:00:17 lichtblau [n=user@port-83-236-3-70.dynamic.qsc.de] has joined #lisp 22:00:33 -!- mejja [n=user@c-7d2472d5.023-82-73746f38.cust.bredbandsbolaget.se] has quit [Remote closed the connection] 22:01:41 -!- edon [n=edon@albalug/edon] has quit [Connection timed out] 22:02:15 -!- _8david [n=user@port-83-236-3-70.dynamic.qsc.de] has quit [Read error: 113 (No route to host)] 22:03:04 -!- novaburst [i=nova@sourcemage/mage/novaburst] has quit ["leaving"] 22:05:16 -!- FZ [n=chatzill@unaffiliated/fz] has quit ["ChatZilla 0.9.83 [Firefox 2.0.0.17/2008082909]"] 22:05:24 -!- josemanuel [n=josemanu@67.1.222.87.dynamic.jazztel.es] has quit ["Saliendo"] 22:07:20 -!- gonzojive [n=red@DNab422456.Stanford.EDU] has quit [] 22:07:41 -!- cemerick [n=la_mer@75.147.38.122] has quit [] 22:11:01 -!- ehu [i=52aa21ad@gateway/web/ajax/mibbit.com/x-9cbebc72113684ed] has quit ["http://www.mibbit.com ajax IRC Client"] 22:15:20 elurin [n=user@81.213.203.109] has joined #lisp 22:18:28 rdd [n=user@c83-250-142-219.bredband.comhem.se] has joined #lisp 22:19:48 Does anyone remember whether the unifier built in PAIP supports backtracking? 22:20:30 topo_ [n=topo@200.37.161.41] has joined #lisp 22:20:35 hello 22:22:05 tritchey_ [n=tritchey@c-98-226-113-211.hsd1.in.comcast.net] has joined #lisp 22:22:06 topo pasted "create structures" at http://paste.lisp.org/display/69660 22:22:27 -!- mindCrime [n=chatzill@216.27.62.2] has quit [Success] 22:22:48 one question, do anybody have any idea of whats wrong in this code? 22:22:50 http://paste.lisp.org/display/69660 22:23:09 it has bad indention 22:23:17 im trying to create randomly different opengl structures or a list full of random opengl statements 22:23:26 for example: (gl:scale 1 9 2 ) (gl:translate -2 -3 -1.5) (glut:solid-sphere 3.529d0 13 3) 22:23:34 or (gl:rotate 119 997 27 ) (glut:solid-cube 9.d0) 22:23:34 etc 22:24:09 topo_ use (list (glut:...) ...) instead of '( (glut:...) ...) 22:24:10 the problem is that i just get one transformation as result , always 22:24:14 S11001001: I have found a bug in w-r-i's design; I just forwarded a mail to you with a summary of the revised design that's currently in my head. 22:24:20 -!- elurin [n=user@81.213.203.109] has quit [Remote closed the connection] 22:24:39 when i got tranformation it should select one of my predefined transformations and then call again so the result should be a list most of the time 22:24:49 unless I'm misunderstanding what this is supposed to do (likely) 22:24:50 ummmmh 22:25:08 gonzojive [n=red@fun.Stanford.EDU] has joined #lisp 22:25:30 -!- Sinden [n=Sinden@nickweinhold.demon.co.uk] has quit ["Leaving"] 22:25:39 -!- vasa [n=vasa@mm-72-92-84-93.dynamic.pppoe.mgts.by] has quit [Read error: 104 (Connection reset by peer)] 22:26:16 topo annotated #69660 with "anot" at http://paste.lisp.org/display/69660#1 22:26:17 I don't know how to solve this: A function takes two lists of intervals (an interval beeing two integers) and should return a list of the intervals that are represented in both lists taken. My current method is to have (let ...) with start and ending variables for the first intervals in both lists and then iteratively go through them sending a result parameter with it. And when one list is empty I return the result. The problem for me 22:26:26 -!- borism [n=boris@195-50-199-64-dsl.krw.estpak.ee] has quit [Client Quit] 22:27:12 damn, gtg 22:27:17 bertskert: you were cut off after `for me' here. 22:28:04 topo_: what does make-expression do? 22:28:11 -!- ryepup1 is now known as ryepup 22:28:29 oops sorry, its should be (make-structure.) 22:28:40 The problem for me seems to be how I should handle all the interval-"collision checking".. Right now I'm having a 2D-cond structure. Do you have any tips? 22:28:42 its the call to the function that create the opengl structure 22:28:44 bertskert: naively, you could just iterate through all the pairs and take the union of the n^2 intersections 22:28:48 cracki [n=cracki@sglty.kawo2.RWTH-Aachen.DE] has joined #lisp 22:28:50 -!- gonzojive [n=red@fun.Stanford.EDU] has quit [Client Quit] 22:29:01 topo_: and it should also probably be inside the list 22:29:09 yes 22:29:12 im trying that 22:29:45 shouldn't topo be asking this in #cl-gardeners? 22:29:48 ryepup has chosen to play #lisp's Sisyphus today, I see. 22:29:52 (or any other channel that I'm not currently in) 22:30:02 (assuming you have closed intervals) more efficiently, you can just sort all the boundaries you have, sort them into [b_0, b_1, ... b_n] and work on the set of intervals [b_i, b_i+1] 22:30:08 chandler: I'm done with silly email forms and wanting to feel better about myself 22:30:08 gonzojive [n=red@fun.Stanford.EDU] has joined #lisp 22:30:17 helping folks usually solves that 22:30:26 pkhuong, I don't really get it? 22:30:29 -!- brickhazel [n=brickhaz@63.144.132.78] has quit [Remote closed the connection] 22:30:47 ryepup: The implication was that trying to help topo_ is futile. 22:31:01 sellout [n=greg@c-24-128-50-176.hsd1.ma.comcast.net] has joined #lisp 22:31:08 and you can represent that set with a (bit) vector, which gives you easy union, intersection and normalisation to a list of disjoint intervals. 22:31:26 I got the implication, but wasn't personally aware of the history, so gave the benefit of the doubt 22:31:27 ryepup i get this as result: 22:31:28 ((CL-OPENGL:TRANSLATE -2 -3 -1.5)) 22:31:36 just a list with one element 22:31:53 it is not supposed to get a list with more than one element? 22:31:58 : \ 22:32:00 But I'm working on some preset abstract datatypes, they're not pairs but abstract datatypes working as an interval 22:32:21 bertskert: and? 22:32:32 fusss [n=chatzill@pool-72-66-40-106.washdc.east.verizon.net] has joined #lisp 22:32:34 bah, . is treated specially by the reader? 22:32:43 mathrick, Of course 22:33:11 vixey: I thought it was only special for the QUOTE reader 22:33:22 topo annotated #69660 with "structures" at http://paste.lisp.org/display/69660#2 22:33:28 mathrick: Why would that be? 22:33:36 I think it's used to represent 'fake lists 22:33:59 chandler: because that's where it's useful to have it treated specially 22:34:02 (or better, to represent a cons cell which isn't a proper list) 22:34:13 hmm, right, you can use . in lambda lists too 22:34:15 topo_: is #2 the latest code? 22:34:15 mathrick: So, (read-from-string "(1 . 2)") should not work? 22:34:34 yes ryepup 22:34:47 topo_: what does make-expression do? 22:34:48 chandler: hmm 22:35:12 make-expression nothin 22:35:15 topo_: and it should also probably be inside the list 22:35:34 it was an error i should put make-sctructure 22:35:48 i should did 22:35:50 hehe 22:36:00 topo_: is #2 the latest code? 22:36:05 yes 22:36:07 mikesch [n=axel@xdsl-87-79-63-18.netcologne.de] has joined #lisp 22:36:11 topo_: what does make-expression do? 22:36:18 mathrick: It's treated specially the token reading part, and in the ( reader-macro. 22:36:23 +in 22:36:23 pkhuong, I get it now! Really thx! 22:36:28 ? 22:36:42 ryepup i told you , it was a mistake 22:36:46 OK. I am sick of this conversation now. 22:36:54 -!- ChanServ has set mode +o chandler 22:36:54 -!- tritchey [n=tritchey@c-98-226-113-211.hsd1.in.comcast.net] has quit [Read error: 110 (Connection timed out)] 22:36:55 its make-structure not make-expressio 22:36:57 tcr: yeah, and it's the token part that's tripping me up 22:37:57 mathrick: What exactly? 22:37:59 topo_: not according to the paste, I'm moving on to other tasks now, I suggest you look up nconc once you have make-structure returning nested lists 22:38:36 ryepup: after looking at the paste, I'm fairly sure that topo is trying to do something that he shouldn't be doing in the first place. 22:38:40 -!- silenius_ [n=jl@dslb-088-073-082-208.pools.arcor-ip.net] has quit [Remote closed the connection] 22:38:52 tcr: I wanted to have . be a separator in forms I feed into my macro, but it doesn't like more than one element following the . 22:38:54 why not chandler? 22:39:00 so I guess I'll have to use $ 22:39:14 bertskert: in the worst case, it's theoretically not really better than the naive intersect/union double loop, but in the real world it should, if only because the operations are simpler. 22:39:17 topo_: are these forms eventually going to wind up going to EVAL? 22:39:36 yes 22:39:46 mathrick: Example? 22:39:50 Do you have a really good argument for using EVAL? 22:39:51 isn't toppo the same guy who attempts to code lisp by throwing forms at lisp until it compiles? 22:39:56 mathrick: Yes, he is. 22:40:03 the idea is redefine function at run time with the generated structure from make-structure 22:40:06 lesson learned 22:40:37 i can't find anything in the sbcl manual on how it treats numerics, specially floating-point computation. I know ANSI guarantees that single-float and short-float are IEEE floats, implementation details? optimization? preferred forms of coercion/conversion/truncation? 22:40:51 tcr: ($ a b c . d e f . x y z) => (a b c (d e f (x y z))), where $ is the macro in question 22:41:15 I have a use-case where I'm having a lot of calls of this structure, so I want to flatten them 22:41:26 mathrick: Perhaps you should make a custom reader macro. 22:41:41 but I wonder what ($ a b c .1 d e f) should do 22:41:42 chandler: I'll just use $ 22:41:43 arguments? the idea is to explore the potential of common lisp for automatic generation of complex opengl structures using self generation of nested iterations and recursion 22:41:45 mathrick: and you don't want to go with, say. ($ (a b c) (d e f) (x y z)) ? 22:41:56 mathrick, don't use . 22:42:02 -!- ryepup is now known as ryepup_away 22:42:04 vixey: I figured that out already :) 22:42:11 mathrick, Well done..... 22:42:17 mathrick: That seems more to be a case of ('s dot parsing than of the token reader 22:42:20 pkhuong: that's a good question, I haven't decided yet 22:42:26 mathrick, so what are you actually asking, or just mumbling? 22:42:32 vixey: You're annoying. 22:42:46 tcr, You're kind of dumb 22:43:00 My fingers are getting itchy. 22:43:05 tcr, I don't think everyone in #lisp wants' to hear my opinion though, or yours 22:43:14 -!- chandler has set mode +b %topo*!*@* 22:43:17 vixey: You're half right. 22:43:21 back 22:43:30 That's exactly the reason why you came across being annoying. 22:43:35 -!- chandler has set mode -b RickedRoger!*@* 22:43:37 -!- chandler has set mode +b %vixey!*@* 22:43:47 -!- chandler has set mode -o chandler 22:44:29 pkhuong: I guess I'll just see how it feels, it's not like it's a huge amount of work to change it from one syntax to the other once it works 22:44:55 *fusss* yes, i would like to order a side dish of last hour's chat logs! 22:44:59 tcr: you're already way more into readtables than I ever intended to go (otoh, I now casually hack the readtable when I want something weird for a file ;) 22:45:10 fusss: less meta, more code 22:45:11 minion: chant 22:45:11 but I would suggest one change: expose the generator. 22:45:11 MORE CODE 22:45:41 does it always chant that? 22:45:52 S11001001: The problem is that W-R-I's current design does not allow writing a merge-readtables that is completely right 22:45:53 No, there are more options for it to chose from. 22:45:58 more chants 22:46:00 minion: chant 22:46:00 MORE CHANTS 22:46:00 minion: Chant! 22:46:00 MORE CHANTS 22:46:04 and then rename the spec to "more readtable accessors" or something 22:46:07 more of the same 22:46:09 minion: chant! 22:46:09 MORE OF THE SAME 22:46:23 oh, so it just takes the last "more of" 22:46:26 Did we narrowly miss "more readtable"? 22:46:30 bertskert: there's a better (in the algorithmic sense) solution if you normalise your inputs to sorted sequences of disjoint intervals first. 22:46:33 nyef: looks so 22:46:38 nyef: Yes, we did. 22:46:42 minion: chant 22:46:42 MORE READTABLE 22:46:43 *seems so 22:46:45 there ya go :-) 22:46:50 topo annotated #69660 with "fixed" at http://paste.lisp.org/display/69660#3 22:46:55 haha 22:46:57 mathrick: (list 'more 'of *) 22:47:27 It chants the last "more foo", which acts specially if foo is a prepositional phrase 22:47:29 S11001001: You mean to split up the proposals? 22:47:47 no, keep it one 22:47:55 but call it "more readtable accessors" or something like that 22:48:16 I almost said "extended" but that feels too formal to me 22:48:28 topo annotated #69660 with "fixed now" at http://paste.lisp.org/display/69660#4 22:48:36 because, after all, the point is that the existing readtable accessors aren't good enough 22:48:37 S11001001: Ah, yes, of course, as the revised CDR's scope will be broader. 22:48:57 -!- neurogeek_ [n=neurogee@201.208.73.206] has quit [Read error: 110 (Connection timed out)] 22:49:04 -!- tritchey_ [n=tritchey@c-98-226-113-211.hsd1.in.comcast.net] has quit [] 22:49:15 anyway, at least 2 of 3 implementations already have an exposable generator, it's just a matter of making it public 22:50:16 -!- disumu [n=disumu@p57A25119.dip.t-dialin.net] has quit [Connection timed out] 22:50:41 S11001001: It seems unreasonable to expect implementations to incorporate it when it'll change significantly again after a few months. 22:51:41 -!- dkcl [n=dkcl@unaffiliated/dkcl] has quit [Remote closed the connection] 22:52:13 Anyway, I just wanted to inform you that the current proposal is going to be revised. I'm currently hacking up an implementation into my sbcl branch. I won't have time for the actual CDR specification for the time being. 22:53:07 -!- gigamonkey` [n=user@adsl-99-184-207-175.dsl.pltn13.sbcglobal.net] has quit [Remote closed the connection] 22:54:57 is there a way to do a "safe" (read), one that avoids cases like (read-from-string "#.(setq *foo* \"gotcha\")") ? 22:55:00 -!- oudeis [n=oudeis@p11811112.orange.net.il] has quit ["This computer has gone to sleep"] 22:55:20 clhs *read-eval* 22:55:20 http://www.lispworks.com/reference/HyperSpec/Body/v_rd_eva.htm 22:55:49 -!- matley [n=matley@83.225.53.145] has quit ["ERC Version 5.3 (IRC client for Emacs)"] 22:56:06 locklace: Notice however that the implicit interning of symbols can't be turned off. 22:56:20 well, you can turn on *read-suppress* 22:56:23 tcr: great. any cases other than #. to worry about? 22:56:27 but that doesn't do you a whole lot of good 22:56:37 sure, an evil user can consume all of your memory with the array reader 22:56:41 ok, good point 22:56:42 -!- rdd [n=user@c83-250-142-219.bredband.comhem.se] has quit [Read error: 104 (Connection reset by peer)] 22:56:42 what are you trying to read? 22:59:19 chandler: just evaluating read as a way to do parsing of simple values in a reasonably controlled way. most of the cases i'm concerned with are not paranoid, like reading config files from well-meaning users, but as a matter of interest i'm also curious how well it can be sandboxed for the paranoid case 22:59:48 clhs 2.2 22:59:49 http://www.lispworks.com/reference/HyperSpec/Body/02_b.htm 23:00:14 If you really want to sandbox it, you'll need to effectively rewrite the reader. beach is actually working on an implementation of that. 23:00:31 The behavior in section 2.2 is hardwired in; anything else can be changed by turning off reader macros. 23:00:43 If you set # to be constituent, most of the standard reader macros go away. 23:02:16 rread [n=rread@c-76-21-116-77.hsd1.ca.comcast.net] has joined #lisp 23:02:28 minion: logs? 23:02:28 logs: #lisp logs are available at http://ccl.clozure.com/irc-logs/lisp/ ; older logs may be available at http://tunes.org/~nef/logs/lisp/ (down as of 2008/09/24) 23:03:05 chandler: thanks, will have an, er, read. ;) it seems to me that it would be very useful to have a fully-sanitisable read, so that one need not reimplement the lisp parser in cases where it's sufficient (even if those cases involve untrusted data) 23:03:09 -!- slyrus_ [n=slyrus@adsl-68-121-172-169.dsl.pltn13.pacbell.net] has quit [Read error: 104 (Connection reset by peer)] 23:03:48 topo pasted "self generation" at http://paste.lisp.org/display/69661 23:04:38 YA simple socket question --- what sort of thing is returned by sb-bsd-sockets:socket-make-stream? In particular, is this a stream that is compatible with being a gray stream? 23:04:48 *ushdf* throws up 23:04:55 what sort of thing??? 23:05:04 well, you've come to the right place, buddy! 23:05:36 i'm guessing it would be a ref erence to the stream 23:05:44 at the expected point of input 23:06:03 -!- willb [n=wibenton@128.105.48.219] has quit ["Leaving"] 23:06:13 -!- vixey [n=vicky@amcant.demon.co.uk] has quit [Client Quit] 23:06:42 the great eth0 23:06:50 rpg: What do you mean with compatible? 23:07:08 and its manifold children 23:07:13 slyrus_ [n=slyrus@adsl-68-121-172-169.dsl.pltn13.pacbell.net] has joined #lisp 23:07:19 rpg: It's a stream. In SBCL, it's an fd-stream, just as you would get from, say, OPEN. 23:07:22 ushdf: Got anything meaningful to say? 23:07:26 ushdf: You are rambling. 23:07:37 who said i was trying to be coherent 23:07:38 For other implementations of sb-bsd-sockets, it might be some other sort of stream. 23:07:49 "Jayne, your mouth is talking. You might want to see to that." 23:08:05 are you chandler from FRIENDS 23:08:06 chandler: Firefly? 23:08:11 nyef: Of course. 23:08:17 I was reading the Allegro discussion of gray streams, and they complain about the fact that you have classes for the directions and for the element type, making it cumbersome to work with things like the fd-stream, which seems to be bivalent and input/output. 23:08:23 ushdf: My fingers are getting itchy again. 23:08:34 maybe you have the mange 23:08:59 rpg: Ah. Wait, -and- the element type? 23:09:00 My question arises because I'm trying to emulate ACL's sockets, which actually are streams themselves. So wrapping up a socket together with a stream in one object. 23:09:02 ushdf: Maybe you should stop talking. It might help. 23:09:22 what's that, a threat? 23:09:46 or a "warning" as you might call it 23:09:51 nyef: This is what the Franz manualage says: "# Gray streams distinguish input and output directions per class, forcing combination and mixins in order to model the 3 different modes (input only, output only, and input/output) for various stream classes. 23:09:55 -!- ChanServ has set mode +o chandler 23:10:02 # Gray streams, in accordance with CL, distinguish streams by element-type. We have found this to be an unfortunate limitation, since it makes it hard to transfer varying-width elements in the same stream. Allegro CL 5.0.1 introduced bivalent streams, which allow both varying-width elements and character elements to be transferred. This was a step in the right direction, and enables web servers to be written more easily and efficiently. " 23:10:31 -!- chandler has set mode -b %qrck!*@* 23:10:35 nyef: will you be around in a couple hours? I'm preparing a presentation on peephole optimisation, and I think you'll like the paper too. 23:10:45 -!- chandler has set mode -b *!*4c5baa5e@* 23:10:48 rpg: Ah, so not the element type, just the directionality. And this is more a lack in the specification of streams in general, not just gray streams. 23:10:54 nyef: So I was trying to figure out, for my socket + stream class, whether I should be using simple-streams or gray-streams. 23:11:03 oudeis [n=oudeis@217.194.205.196] has joined #lisp 23:11:04 pkhuong: I could arrange to be for maybe the next four hours. 23:11:10 nyef: Seemed like gray-streams were more finished, hence preferable. 23:11:13 -!- chandler has set mode +b %ushdf!*@* 23:11:38 rpg: For SBCL? Yeah, either gray streams, or maybe something based on the underlying ansi-stream/fd-stream model. 23:11:41 nyef: sounds good, thanks. 23:11:51 -!- chandler has set mode -o chandler 23:11:55 it's a bantastic day 23:12:01 today must be fullmoon 23:12:29 *S11001001* fetches tcr's revs and set up gitweb on sbcl 23:12:42 epoch [n=K-I-S-S@p3m/member/epoch] has joined #lisp 23:12:45 S11001001: It seems to be! All I want is more on-topic coherent discussion. 23:12:54 pkhuong: I'd probably like to see the paper too, if you don't mind. 23:13:26 well that was easier than hgweb 23:13:26 chandler: i'm just presenting an old paper, "Automatic Generation of Peephole Optimizations", Davidson & Fraser, PLDI 1984 23:13:44 Ah. 23:13:44 nyef: Sorry, I think I'm confused --- it looks like for gray streams there are separate classes for binary versus character, as well as direction (and no i/o stream). 23:13:48 hgweb has the irritating property of using Python os.walk directly so it doesn't follow directory symlinks 23:13:52 minion: chant 23:13:53 MORE ON TOPIC 23:13:58 tcr: you were right, today is full moon 23:14:27 S11001001: I haven't yet pushed anything re. the proposal 23:14:40 So, should I try setting up a gitweb for SBCL/ARM and using it for my work-in-progress as well as posting my log? 23:14:47 no, it's not, nevermind 23:15:14 nyef: Well, I'd like to see your progress, but I'm not so sure about the git part of that proposition. 23:15:36 chandler: Have an alternative to suggest, then? 23:15:57 No, not really. 23:16:15 Other than making an actual SBCL CVS branch, but that might be more commitment than you were seeking from this relationship. 23:17:02 Using CVS? Yes, that's definately more commitment than I'm looking for. 23:17:10 nyef: Aha. Now I see that there's a socket-simple-stream in simple-streams library for SBCL. Should probably be the right thing for me to look at! 23:17:12 chandler: fork on repo.or.c? 23:17:29 nyef: but the "alpha quality" is a little nervous-making. 23:18:02 pkhuong: isn't that still some kind of git thingy?\ 23:18:13 rpg: And for more fun, sb-simple-streams doesn't (or didn't) pass its test suite on win32. 23:18:33 chandler: What's the problem with git? 23:18:54 rpg: do you know iolib? take a look if not. 23:18:57 -!- davazp [n=user@72.Red-88-6-205.staticIP.rima-tde.net] has quit [Read error: 104 (Connection reset by peer)] 23:19:21 nyef: Probably just that I don't understand it. 23:19:25 attila_lendvai: I will have a look again. Alas, I cannot just use iolib because my charge is to be ACL-compatible. 23:19:53 Are you working on acl-compat? 23:20:22 rpg: porting iolib to ACL (if it's needed) should be simple, unless windoze support is needed. fe[nl]ix is very helpful on #iolib... 23:21:01 chandler: Well, I'm more thinking of it as a web-browsable repository that can be used for people to view what I've done, not as anything I expect people to interact with to any real depth. 23:21:02 attila_lendvai: Sorry. Unclear. No, what I'm trying to do is build stuff in SBCL that's as close to following the ACL API as possible. 23:21:33 attila_lendvai: We have a large number of big systems written in ACL that we would like to see run in SBCL. Sockets are the remaining stumbling block. 23:21:40 nyef: Well, the git thing seems like a decent idea then. 23:21:43 chandler: I figure that once things are "complete enough", there will be a monster set of patches to apply to SBCL CVS and then it'll be done. 23:22:09 rpg: so "in sbcl" meaning something on sbcl that makes it easy to port your codebase to sbcl? 23:22:38 rpg: isn't this acl-compat's goal? or does it not do what you need? 23:22:46 drdo [n=psykon@167.111.54.77.rev.vodafone.pt] has joined #lisp 23:23:02 -!- sellout [n=greg@c-24-128-50-176.hsd1.ma.comcast.net] has quit [] 23:23:05 attila_lendvai: Yes. In practice, this seems to be writing a small number of classes to support a rewrite of acl's make-socket (this does the heavy lifting). 23:23:15 -!- fisxoj [n=fisxoj@149.43.108.7] has quit [Read error: 110 (Connection timed out)] 23:24:08 rpg: even then i suggest to at least talk to fe[nl]ix. in the very least he will probably have good ideas on how to implement it... but i suspect there's some layer in iolib that you could build on. 23:24:22 attila_lendvai: That would be very nice... 23:24:59 Actually, the main issue seems to be yoking together a socket and a stream and then having the resulting object implement the stream protocol by delegation. 23:25:44 rpg: well, iolib can certainly do that. supporting bivalent streams with external format 23:25:55 I assumed that gray-streams provided the means to do this, but the earlier-quoted Franz discussion seemed to shed doubt on that. 23:26:10 isn't flexi-streams implemented on top of gray streams? 23:26:23 attila_lendvai: Thank you, I will root around in iolib and see if I can find it. 23:27:11 topo annotated #69661 with "without lambda or compile" at http://paste.lisp.org/display/69661#1 23:27:16 rpg: the api is so straightforward that i, a complete socket noob, could use it. make-socket, accept, and you get a stream connected to a socket 23:27:46 chandler: Thanks, flexi-streams had completely slipped my mind. 23:28:28 attila_lendvai: iolib cannot work on ACL because of ACL's incomplete FFI 23:28:46 -!- vcgomes is now known as vcgomes[away] 23:28:47 no (unsigned) long long, no party 23:29:07 what's that used for in iolib? 23:29:32 off_t in stat() and a few other syscalls 23:29:32 fe[nl]ix: Is that for return value, or arg value? 23:29:51 'cause if it's just arg and struct field values, you can hack around it. 23:30:00 nyef: return, arg and struct member 23:30:03 Return values, though, you're SOL. 23:30:22 there's always the option of writing tiny C wrappers. 23:31:18 fe[nl]ix: rpg needs something on sbcl that has an ACL socket api 23:31:19 fe[nl]ix: Actually, running ON ACL isn't an issue for me --- I'm trying to run ON SBCL code that was written FOR ACL. So it doesn't matter if iolib doesn't work on ACL (although it's regrettable). 23:32:14 *rpg* probably just needs a better understanding of the way extensible streams are intended to work 23:32:28 -!- hkBst [n=hkBst@gentoo/developer/hkbst] has quit [Read error: 54 (Connection reset by peer)] 23:32:46 rpg: iolib's socket API is almost identical to ACL's. it should almost a no-op translating ACL socket code 23:32:53 except for socket options 23:33:34 fe[nl]ix: OK, thanks. I will see if I can get a copy of iolib working here to test. 23:34:43 fe[nl]ix: Has this ever been tested on MacOS? I ask because a good chunk of our team use Macs... 23:35:42 have you guys seen tTk? it's basically old Tk gui but themeable at run time. really purty. i wanna made CFFI bindings for libtk itself, instead of talking to wish via socket/pipe, a la LTK 23:35:43 Let me go back to what seems to be the core question, and please feel free to point me at an FM for me to R... 23:36:22 tcr: see, line 19 http://home.nocandysw.com/gitweb?p=cxml.git;a=blob;f=dom/unification.lisp;h=28947ada852550e5aa2c8e2d91d106e7c98041ae;hb=HEAD is all your fault ;) 23:36:28 gray streams are supposed to let us make new stream classes that act like normal streams, avoiding the closed nature of normal CL for this. Correct? 23:37:17 So they would solve my problem of wanting to handle the stream protocol by delegation, correct? 23:38:14 rpg: some time ago H4ns allowed me to test iolib on his dev machine, an openmcl/ppc and it ran the test suite fine. I'm not sure about the present situation 23:38:35 -!- milanj [n=milan@212.200.192.182] has quit ["Leaving"] 23:38:38 I think all I'm trying to do is to find a way to describe the fd-stream that sb-bsd-sockets gives me as some sort of gray stream, by delegation. 23:39:18 sellout [n=greg@c-24-128-50-176.hsd1.ma.comcast.net] has joined #lisp 23:39:28 "0.9.2.43: another slice of whitespace canonicalization (Anyone who ends up here with "cvs annotate" probably wants to look at the "tabby" tagged version.)" 23:39:41 How do I get past the canonicalization commit? 23:41:47 don't have an SBCL checkout handy right now, but I'd guess "cvs annotate -r tabby foo.lisp" 23:41:50 heh, we recently rejected whitespace canonicalization on weblocks because of annotate 23:42:08 -!- jpcooper [n=justin@unaffiliated/jpcooper] has quit [Remote closed the connection] 23:42:30 yangsx [n=yangsx@218.247.244.25] has joined #lisp 23:42:33 S11001001: I find that #U icky. Why is it (defgeneric unify, and not (defgenerc #Uunify btw? 23:43:17 lichtblau: I wonder whether I can get past the commit using gitweb. 23:43:54 icky but more pleasant than unify:: 23:44:14 I find that my eye now jumps over those when they're too present, and I can see the #U better 23:44:21 It's a chore to read, are you sure it's more pleasant? 23:44:31 cemerick [n=la_mer@c-71-232-219-241.hsd1.ma.comcast.net] has joined #lisp 23:44:35 -!- cemerick [n=la_mer@c-71-232-219-241.hsd1.ma.comcast.net] has quit [Remote closed the connection] 23:44:42 maybe I'll replace it when the branch is done, but not now 23:44:52 S11001001: I'd make it #U{foo} at least. 23:44:58 anyway, the reason is that my unify is side-effecty whereas unify:unify is side-effecty+valuey 23:45:16 wasn't the #U dispatch character used for Unicode strings? 23:45:27 cipher [n=cipher@inertia.acm.cs.rpi.edu] has joined #lisp 23:45:35 not in cxml-dom 23:45:54 hmmm 23:46:29 *fusss* really thinks of running swigs against /usr/include/tk .. what the hell is a grovel anyway? 23:47:30 LiamH [n=nobody@pool-138-88-126-7.res.east.verizon.net] has joined #lisp 23:49:52 -!- epoch [n=K-I-S-S@p3m/member/epoch] has quit ["  "] 23:51:06 Heh. And I'd have considered using #U for UUIDs. 23:51:56 -!- kib2 [n=chatzill@bd137-1-82-228-159-28.fbx.proxad.net] has quit ["ChatZilla 0.9.83 [Firefox 3.0.3/2008092417]"] 23:52:04 clhs equalp 23:52:04 http://www.lispworks.com/reference/HyperSpec/Body/f_equalp.htm 23:52:21 rpg: gray streams are CLOS instances, while fd-streams are structures. I'm not sure you can combine them 23:52:22 froog_ [n=david@87.192.28.247] has joined #lisp 23:52:46 nyef, my lisp uses #G 23:53:13 "guuid"s 23:53:30 `Globally universally unique identifiers'? 23:53:44 fe[nl]ix: You are right; I was not thinking clearly. What I should be doing was taking the gray-streams as a protocol, I believe, and just implementing the methods. It's somewhat confusing to me. If this was CLIM, there would be a notion of protocol, and I would implement that, and be done. 23:53:47 I thought it was GUID, not GUUID? 23:53:49 yeah.. since iuuids as well 23:53:56 O_4 [n=souchan@ip-118-90-65-113.xdsl.xnet.co.nz] has joined #lisp 23:53:58 oh my bad 23:54:37 fe[nl]ix: But there must be some notion of protocol underlying the gray-streams, right? Because they allow me to, e.g. write new methods for cl:close, which violates standard CL. Yes? 23:54:42 G(U)UID means its registered somewhere as well 23:54:52 rpg: AIUI, one of the complaints about the gray stream protocol is that it's not well defined which bits need to be implemented by someone building their own streams. 23:55:30 Another complaint is that it missed one or two standard stream functions... position or length, I think... 23:56:17 Riastradh, we starts with UUID .. then we "register them" .. or promise not to chage the UUID 23:56:52 whenever a UUID is made we go at and make sure its not in use ;P 23:57:03 go at/ go out 23:57:55 AIUI, it's not supposed to be possible to create the same UUID on machines with different MAC addresses on their network cards. 23:58:13 sometimes two machines try to coerce the same UUID to a GUID 23:58:50 If you have an adequate source of randomness, you will never, ever come up with the same UUID twice, even without resorting to obnoxious hacks like including the MAC address. 23:58:52 nyef, yeah.. oops sorry one machine running two instances of the applkication 23:58:59 I thought guids (at least the Microsoft ones) weren't based on mac adress any more 23:59:04 due to security concerns 23:59:32 -!- njsg [n=njsg@unaffiliated/njsg] has quit ["gone"] 23:59:50 our first public bug with this lisp was when a person instance of opencyc wouldnt start since they didnt have the internet 23:59:58 -!- sellout [n=greg@c-24-128-50-176.hsd1.ma.comcast.net] has quit []