| 00:24.01 | louipc | hm build failed |
| 01:08.46 | Maloeran | brlcad might be at least a half-god of the software, but I still suspect he'll need more information |
| 01:14.05 | ``Erik | mal, just for you... http://cache.gizmodo.com/assets/resources/2007/02/pcmaclinux-thumb.jpg |
| 01:16.56 | Maloeran | Eheh, I saw it already, quite nice |
| 03:08.47 | *** join/#brlcad IriX64 (n=IriX64@bas3-sudbury98-1168056909.dsl.bell.ca) | |
| 03:49.57 | Maloeran | Linux threads really execute in very constant, predictable patterns |
| 03:50.20 | Maloeran | It's good for the cache of processors, but I wish I could easily insert some randomness to test things out |
| 03:53.24 | louipc | aarg I don't think I'm getting the latest versions of everything |
| 03:59.07 | brlcad | louipc: que? |
| 04:00.06 | louipc | I think it has something to do with the way cvs packages are prepared on my distro |
| 04:00.30 | louipc | because doing it manually works |
| 04:37.33 | louipc | yep |
| 05:25.47 | IriX64 | nytol :) |
| 06:03.41 | Maloeran | Ouch. Looks like a nice theater |
| 08:23.16 | *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch) | |
| 08:23.59 | *** join/#brlcad cad10 (n=c636cad2@bz.bzflag.bz) | |
| 14:29.00 | louipc | yar I got that same make install error as IriX64. making the directory just makes it give Error 126 instead |
| 15:38.49 | brlcad | the farther ports gets from 5.4, the more fault-prone it seems to get |
| 15:39.56 | brlcad | I usually run: cvsup -g -L2 /etc/su |
| 15:40.03 | brlcad | /etc/supfile |
| 15:41.18 | brlcad | along with, portsdb -Uu, pkgdb -F, portupgrade -arR -l logfile |
| 15:43.28 | brlcad | curious to know if the scan comes up with anything |
| 15:46.17 | ``Erik | <-- not upgrading anything, but installed a couple things |
| 15:46.44 | ``Erik | everything php5 has buffer overflow vulnerabilities it'd seem |
| 15:46.58 | ``Erik | I threw on 'portsnap' and 'portaudit' |
| 15:47.27 | ``Erik | and ports are supposed to work fine on RELENG_5, which is now 5.5... you're running 5.2.1 :) |
| 15:47.41 | brlcad | yep |
| 15:47.55 | ``Erik | single proc machine? |
| 15:48.13 | ``Erik | <-- keeps thinking it's a dual for some reason |
| 15:48.26 | brlcad | i was inclined to upgrade the kernel, but frankly it freaks me a bit to mess with a stable config |
| 15:48.29 | brlcad | it's a single |
| 15:48.38 | brlcad | duals are way too overpriced |
| 15:49.04 | brlcad | heck even their singles are overpriced now or I would have just bought new hardware and migrated |
| 15:50.04 | ``Erik | you're leasing their hw, right? would it be easier to put together a machien and coloc it? |
| 15:50.34 | brlcad | yep, and not really |
| 15:50.49 | brlcad | the hardware lease is one of the cheapest parts |
| 15:51.11 | brlcad | bandwidth is the killer, and I got a sweet deal |
| 15:54.17 | brlcad | kerash |
| 15:59.05 | ``Erik | heh, yeah |
| 15:59.14 | ``Erik | there was some initialization command that I forgot or something |
| 15:59.24 | ``Erik | and it's seeing damaged ports before it gets there |
| 16:02.27 | brlcad | http://fisheye1.cenqua.com/browse/brlcad/brlcad |
| 16:02.59 | brlcad | fun with stats, they hooked us up |
| 16:03.06 | brlcad | http://fisheye1.cenqua.com/browse/brlcad |
| 16:05.08 | ``Erik | I'll assume one of those big jumps in the line count graph was tcl/tk ? |
| 16:05.26 | ``Erik | not sure what the other was |
| 16:11.23 | brlcad | no, tcl doesn't cause that much |
| 16:12.02 | brlcad | i'm asking them what the second spike is.. we're not at 4.5 million lines |
| 16:12.39 | brlcad | the first is the open sourcing, licenses, headers, footers, ws, reorg |
| 16:13.29 | brlcad | thought it might have been some branch activity, but they show that on HEAD too |
| 16:13.49 | ``Erik | hrmmm |
| 16:13.54 | ``Erik | I wonder if they count the attic |
| 16:14.15 | ``Erik | mv lib* src/ could account for that, if they do? |
| 16:16.10 | brlcad | i doubt it, otherwise it would have been much more than double |
| 16:17.28 | brlcad | i vaguely recall accounting for a second spike in the statcvs report, maybe a ws thing, but then their report should account for that as it looks like its supposed to be tracking + and - lines |
| 16:19.01 | brlcad | ./topic BRL-CAD Open Source Solid Modeling || http://brlcad.org || http://sf.net/projects/brlcad || http://fisheye1.cenqua.com/browse/brlcad/brlcad || 7.10 release preparations coming soon to a tarball near you... |
| 16:19.05 | *** topic/#brlcad by brlcad -> BRL-CAD Open Source Solid Modeling || http://brlcad.org || http://sf.net/projects/brlcad || http://fisheye1.cenqua.com/browse/brlcad/brlcad || 7.10 release preparations coming soon to a tarball near you... | |
| 16:19.19 | brlcad | chmod 775 topic |
| 16:19.25 | louipc | woohoo |
| 16:25.18 | brlcad | ``Erik: ew, that didn't exist for a reason :P (but I don't really care either) |
| 16:26.16 | ``Erik | what didn't? the pkg shtuff? |
| 16:26.23 | ``Erik | or /usr/tmp ? |
| 16:26.34 | brlcad | tmp |
| 16:26.42 | ``Erik | I'll remove it when I'm done *shrug* |
| 16:26.45 | louipc | it looks like MKDIR_P has no value in the Makefile which is causing the make install error |
| 16:29.09 | brlcad | ``Erik: as I said, i don't really care .. it's just been entertaining/informative to see what tools presumed it existed |
| 16:29.23 | ``Erik | hehe |
| 16:29.42 | ``Erik | you'd be amazed what goes crazy if you have /usr/Xorg instead of /usr/X11R6 ;) |
| 16:29.51 | brlcad | heh |
| 16:30.18 | brlcad | louipc: hm.. that's odd -- we even provide mkdirp.m4 just in case |
| 16:31.02 | ``Erik | louipc: is that straight out of a distributed tarball? or are you running aclocal and automake? or autogen.sh? |
| 16:31.27 | louipc | that's from CVS & autogen.sh |
| 16:31.44 | brlcad | my guess is your version of automake is doing something .. different |
| 16:31.54 | brlcad | what are your versions? |
| 16:32.05 | brlcad | ./autogen.sh --version |
| 16:33.01 | louipc | Autoconf 2.61, automake 1.10, libtool 1.5.22 |
| 16:34.20 | brlcad | it could be a case sensitivity thing.. automake used to use mkdir_p not MKDIR_P in the Makefiles |
| 16:34.46 | brlcad | louipc: try this, mv m4/mkdirp.m4 misc/ |
| 16:34.48 | louipc | oi |
| 16:34.59 | brlcad | then redo autogen.sh, configure, make, etc |
| 16:35.23 | louipc | ok |
| 16:38.08 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/m4/mkdirp.m4: automake 1.10+ seems to use uppercase instead, subst both upper and lower |
| 16:48.54 | *** join/#brlcad IriX64 (n=IriX64@bas3-sudbury98-1168056909.dsl.bell.ca) | |
| 16:50.27 | ``Erik | uhhh, how do you mean? |
| 16:50.50 | IriX64 | that configure.ac on cvs is *not the full blown one. |
| 16:52.16 | ``Erik | how is it not? what's missing? |
| 16:52.45 | IriX64 | let me look again, i downloaded and got 485 bytes. |
| 16:54.24 | IriX64 | slap me with a trout. |
| 16:55.20 | ``Erik | I'm sorry, but I don't like you in that tway |
| 16:55.30 | ``Erik | s/ t/ / |
| 16:55.41 | louipc | I only have halibut |
| 16:55.59 | IriX64 | heh |
| 16:56.25 | IriX64 | man replace 105kb with 450kb (whatd you do?) |
| 16:56.26 | brlcad | when you run autogen.sh, it generates configure from configure.ac |
| 16:56.36 | IriX64 | i know that. |
| 16:56.51 | brlcad | noir:~ morrison$ wc brlcad/configure.ac 3954 12243 123166 brlcad/configure.ac |
| 16:57.00 | brlcad | it should be about 4000 lines |
| 16:59.58 | IriX64 | http://pastebin.com/884544 |
| 17:00.34 | IriX64 | thats from browse cvs man. |
| 17:02.07 | brlcad | you saved the web page |
| 17:02.25 | IriX64 | man i clecked on configure.ac |
| 17:02.27 | brlcad | presumably you pulled that from the cvs web interface |
| 17:02.38 | IriX64 | ie again? |
| 17:02.44 | brlcad | nope, your error ;) |
| 17:02.58 | IriX64 | heh all right, ill keep trying. |
| 17:03.06 | brlcad | if you right-click the download link, it'd work |
| 17:03.07 | louipc | clicking the file will give you a changelog or something |
| 17:03.09 | brlcad | but only that link |
| 17:03.24 | brlcad | i.e. it actually says (download) |
| 17:03.37 | IriX64 | as i said slap me with a halibut. :) |
| 17:03.40 | brlcad | not that it matters |
| 17:03.46 | ``Erik | irix, have you looked at http://www.wincvs.org/ |
| 17:03.47 | brlcad | that file is not going to work by itself |
| 17:04.14 | brlcad | you can't just take configure.ac and expect/hope it'll work -- it's tied to changes throughout the source tree |
| 17:04.37 | brlcad | autogen.sh is stand-alone which is why that one can be updated by itself |
| 17:05.16 | brlcad | i can pretty much guarantee that using a configure.ac from CVS head will fail on a 7.8 source tree |
| 17:05.27 | IriX64 | ty |
| 17:05.44 | IriX64 | but still i click download and it views the bloody thing. |
| 17:05.56 | brlcad | right-click, save as? |
| 17:06.04 | IriX64 | saves as html. |
| 17:06.26 | brlcad | not if you right click the (download) link |
| 17:06.53 | brlcad | if you just hit save on the menu, sure, or right-click one of the other non-download links, perhaps |
| 17:06.56 | IriX64 | ty learned something. |
| 17:08.02 | IriX64 | saves as configure.ac.txt |
| 17:08.59 | IriX64 | my fault the old one was there to system renamed it automagically :) |
| 17:09.05 | brlcad | that's IE for you |
| 17:09.11 | IriX64 | heh |
| 17:09.14 | brlcad | really should just use 'cvs' |
| 17:09.29 | brlcad | either wincvs or cvs in cygwin |
| 17:10.29 | Maloeran | IriX64, you should consider installing Firefox |
| 17:10.40 | brlcad | or tortoise cvs, another free windows gui cvs |
| 17:10.48 | louipc | or linux :P |
| 17:10.50 | Maloeran | Or just Linux, that'll solve many other problems too |
| 17:10.53 | IriX64 | had it before the great crash of 2006 |
| 17:10.59 | ``Erik | or freebsd, to solve the problems linux creates |
| 17:11.05 | ``Erik | or openbsd, or ... :D |
| 17:11.05 | louipc | hahaha |
| 17:11.14 | IriX64 | or bsd to solve those ones erik :P |
| 17:11.37 | ``Erik | heh |
| 17:11.40 | ``Erik | on a vax? |
| 17:11.44 | ``Erik | or a pdp? :D |
| 17:11.49 | IriX64 | actually vms blows them all away. |
| 17:12.01 | brlcad | or just buy a Mac when you get tired of configuring and/or fighting your OS? :) |
| 17:12.19 | IriX64 | its the apps, the os is fine :) |
| 17:12.31 | ``Erik | except after configuring/fighting osX to get macports all groovy and applisms extracted, ... *cough* :D |
| 17:13.50 | IriX64 | 5'll get you ten i get this to work :P |
| 17:14.00 | brlcad | so.. if you import the configuration environments from other systems.. you end up needing to configure/fight.. who would have thought! |
| 17:14.28 | IriX64 | cooperative operating systems what a thought :) |
| 17:14.35 | ``Erik | hehehe, but it's supposed to be unix! it's mostly fbsd under the hood! :D |
| 17:14.40 | IriX64 | the unios. |
| 17:14.59 | ``Erik | a twisted mutated derivative of fbsd4 with other stuff hammered on in different places, but ... |
| 17:15.08 | brlcad | yes, but it's a sheep, shaven clean, wearing a leather jacket |
| 17:15.30 | IriX64 | that cow is pisses :P |
| 17:15.36 | IriX64 | pissed too. |
| 17:18.43 | IriX64 | he found him in mombassa, in a bar room drinking Gin,... |
| 17:23.36 | IriX64 | just what is this AM_PATH_SDL not found in library anyway. |
| 17:24.40 | ``Erik | you need libsdl installed (with it's m4 fu) to run autogen |
| 17:24.46 | ``Erik | it's used in adrt |
| 17:25.07 | IriX64 | thats not in the source tarball? |
| 17:25.31 | ``Erik | hm, looks like there's a copy in m4/, n/m :) |
| 17:27.53 | IriX64 | http://pastebin.com/884559 |
| 17:28.18 | IriX64 | result of aclocal followed by automake autogen has already run. |
| 17:28.29 | ``Erik | aclocal needs to have -I m4 |
| 17:28.33 | ``Erik | autogen.sh does that for you |
| 17:28.47 | IriX64 | thank ill try it. |
| 17:29.11 | IriX64 | ill autogen verbose if you like and pastebin the output. |
| 17:29.25 | louipc | neato |
| 17:38.52 | IriX64 | autoconf -f takes a while. |
| 17:42.28 | IriX64 | brb |
| 17:43.28 | *** join/#brlcad IriX64 (n=who@bas3-sudbury98-1168056909.dsl.bell.ca) | |
| 18:06.48 | IriX64 | sigh.... i like a slow hand :) |
| 18:07.57 | IriX64 | honest officer, the last thing on my console is autoconf -f |
| 18:09.54 | IriX64 | finally, running automake. |
| 18:10.33 | IriX64 | please don't tell me you want me to repeat this with 1.10 :) |
| 18:12.29 | brlcad | heh, wow |
| 18:12.57 | brlcad | killall with no args on irix is apparently equivalent to shutdown |
| 18:13.50 | IriX64 | just picked this name to be kewl ;) |
| 18:14.26 | brlcad | ``Erik: i think i took down cad if you happen to be going by there.. |
| 18:15.13 | brlcad | it was failing on mounts to a particular bsd system |
| 18:17.58 | ``Erik | not today |
| 18:18.14 | IriX64 | missed that autoheader ran without even a warning. |
| 18:18.16 | ``Erik | it's a holiday, yo |
| 18:18.41 | IriX64 | same for autoconf. |
| 18:19.39 | ``Erik | heh, anyone with leenewx x86 asm fu wanna help me figure somethin' out? :D |
| 18:21.06 | IriX64 | x86? blame it on 18.2 ticks per second ``Erik ;) |
| 18:21.19 | ``Erik | heh |
| 18:21.33 | brlcad | ``Erik: it is? |
| 18:21.51 | ``Erik | I got a brainfuck compiler that outputs x86 assembly in both unix and dos/linux calling conventions... |
| 18:22.11 | IriX64 | heh |
| 18:22.12 | ``Erik | works dandy with the unix calling convention (fbsd), and the simpler test programs work fine on linux |
| 18:22.20 | ``Erik | but I have a more complicated one that fails on linux |
| 18:22.31 | brlcad | holy cow, it is! |
| 18:22.45 | ``Erik | it should almost immediately start printing output, but it immediately says "killed" |
| 18:22.55 | ``Erik | and even with -gstabs to as, I can't get a debugger to do anything with it |
| 18:23.28 | brlcad | i was getting dressed even |
| 18:23.46 | ``Erik | <-- remembered halfway to the office, turned around in a church parking lot and left some rbber on the road |
| 18:23.54 | ``Erik | at ferakin' seven something this morning |
| 18:24.00 | brlcad | hah |
| 18:24.02 | brlcad | sucker |
| 18:24.31 | brlcad | i mean, er, nice catch there right-o lad |
| 18:24.34 | ``Erik | and as I was driving, I was bitching at myself for not doing anything productive yesterday (thinking monday)... |
| 18:25.07 | ``Erik | 4 day weekend, w00 |
| 18:27.44 | brlcad | not having to decide between heroes and work, w00 |
| 18:29.29 | brlcad | i think ive asked you before, but what actually generates the assembler in you compiler? is it outputting binary in scheme or C or something else? |
| 18:29.55 | ``Erik | it's a C program, reads a bf file and outputs x86 asm |
| 18:30.27 | ``Erik | ./bf2as hi.bf > hi.s && as -o hi.o hi.s && ld -o hi hi.o && ./hi |
| 18:30.45 | brlcad | ive toyed with writing a shell script compiler several times |
| 18:31.00 | brlcad | ah, so you still call as |
| 18:31.22 | ``Erik | yeah, I have a shell to generate straight machien code, but I haven't started filling it in |
| 18:31.54 | ``Erik | and I have an interpreter that generates bytecode and executes it, as well as a version that outputs ansi C code |
| 18:34.16 | ``Erik | (the mandelbrot program takes about 50 seconds in the interpreter, 12 seconds as C compiled with -O0, 6 seconds as asm, and 3.5 seconds as C gcc -O1... a hair more with O2 or O3) |
| 18:52.33 | IriX64 | guys, my beautiful packard bell 3070a 17" monitor kaffed last night, lost the pincushion, whats retail for a good 17 inch in your country, i'm debating going new instead of taking it to joes tv repair shop. |
| 19:02.12 | ``Erik | http://froogle.google.com/froogle?q=17+monitor&hl=en&btnG=Search |
| 19:02.20 | IriX64 | ``Erik, interesting project, I went the other way, take binary and decode it to pal neumonics, relativizing the addresses and index cells and all that, 72 opcodes to work with, but you can take a crash dump, run it thru this thing and invalid opcodes stick out like a sore thumb. |
| 19:05.19 | IriX64 | http://pastebin.com/884630 |
| 19:05.32 | IriX64 | if you want a little look at how i did it. |
| 19:05.51 | IriX64 | btw automake is still running :) |
| 19:06.01 | IriX64 | no warnings yet. |
| 19:06.48 | ``Erik | to do what now? O.o |
| 19:07.02 | IriX64 | wait on automake ;) |
| 19:08.47 | IriX64 | the idea is to take it and build a crash dump analyzer, problem is you must have something to produce the crash dump to be read. |
| 19:09.16 | IriX64 | tempted to adapt it to windows. |
| 19:09.17 | ``Erik | yeah... it won't produce any kinda dump or anyhting |
| 19:09.34 | IriX64 | windows does a memory dump. |
| 19:09.34 | ``Erik | it immediately dies with 'killed', even if I break _start |
| 19:10.09 | ``Erik | if it crashed normal, it'd coredump... if it hit an illegal opcode, it'd die with sigill... |
| 19:10.19 | IriX64 | need something to dump memory before the restart thats all. |
| 19:13.34 | IriX64 | if the machine kaffs, set a flag errm no trap illegal and dump core to disk? |
| 19:14.55 | IriX64 | meantime back at the cpu, automake waves and keeps on trucking:) |
| 19:15.34 | brlcad | you used to run a bbs? |
| 19:15.51 | IriX64 | many years ago yes. |
| 19:16.08 | IriX64 | now where'd that come from? |
| 19:16.17 | IriX64 | whatd i paste. |
| 19:16.56 | IriX64 | The DogStar bbs, ===> finest kinnd. :) |
| 19:17.38 | brlcad | just wondering, used to sysop years ago myself |
| 19:17.45 | IriX64 | as? |
| 19:18.13 | IriX64 | meaning bbs name? |
| 19:19.55 | IriX64 | remeber Juge, he's famous :) |
| 19:20.17 | brlcad | ah hell, I don't remember .. it was 20 years ago |
| 19:20.33 | IriX64 | heh does Juge still have his up? |
| 19:20.37 | brlcad | do remember we changed the name a couple times |
| 19:20.55 | IriX64 | I toyed with that. |
| 19:21.02 | IriX64 | had a lot of fun. |
| 19:21.45 | brlcad | played with pcboard, but mostly used wwiv |
| 19:22.07 | brlcad | at least if memory serves.. |
| 19:22.13 | IriX64 | both good systems, i like wildcat too. |
| 19:22.18 | brlcad | yeah |
| 19:22.33 | IriX64 | mine was based on adept, os2 pm app , sweet. |
| 19:26.34 | brlcad | zone 4 fidonet seems to ring a bell |
| 19:31.29 | Maloeran | A nice gesture of sanity, it seems Maryland is going to ban capital punishment... though you guys surely know that already |
| 19:34.52 | brlcad | not much of an issue for maryland, there have only been 5 total in the past 30 years |
| 19:35.53 | brlcad | pretty clear-cut cases of guilt too iirc, regardless of whether you agree with the punishment |
| 19:37.40 | Maloeran | I see. With 5 cases in 30 years, it has to be rather exceptional |
| 19:39.07 | brlcad | pretty much |
| 19:41.36 | brlcad | I think one of the first ones actually dropped his plea and requested execution |
| 19:45.51 | IriX64 | this goes beyond patience tonic, but i'll have another swig. |
| 19:48.25 | IriX64 | those days bitch slap ``Erik for saying that :) |
| 19:49.38 | ``Erik | (heh, yeah, maryland is a little less killhappy than, say, texas....) |
| 19:49.41 | ``Erik | morning, mal |
| 19:52.19 | Maloeran | Hey Erik, feel free to tell me where RF crashes on big endian if you try |
| 19:52.39 | ``Erik | tomorrie :) and iirc, it wasn't crashing, just showing the wrong colors |
| 19:52.46 | Maloeran | And don't try little<->big endian distributed processing yet, I'm not swapping bytes for user messages |
| 19:53.10 | Maloeran | *nods* I need to test the geometry group file format too |
| 19:53.46 | ``Erik | care to entertain a linux/x86/asm problem, mal? I'm as lost as irix64 at a kbd :D |
| 19:53.55 | IriX64 | Maloeran? You two were discussing multi-cpu clusters the other day. |
| 19:54.16 | Maloeran | I read above, Erik, but... I'm not too used to syscalls, I rather call libc functions |
| 19:54.26 | ``Erik | well, I d'no if that's the issue here |
| 19:54.39 | ``Erik | the simple ones that use syscalls work, the complex one doesn't... it gives me this |
| 19:54.40 | ``Erik | $ uname -mo ; as -o mandelbrot-as.o mandelbrot-as.s && ld -o mandelbrot-as mandelbrot-as.o && ./mandelbrot-as ; echo $? |
| 19:54.40 | ``Erik | i686 GNU/Linux |
| 19:54.40 | ``Erik | Killed |
| 19:54.40 | ``Erik | 137 |
| 19:54.41 | Maloeran | IriX64, yes we were |
| 19:55.04 | Maloeran | What does gdb say? |
| 19:55.18 | IriX64 | if you have a 40 cpu cluster what if you make it such that there is a master cpu to dole out tasks that need running to first idle cpu and what if you had a memory pool with memory doled out on a per task guestimate basis? |
| 19:56.03 | ``Erik | mal: http://paste.lisp.org/submit |
| 19:56.28 | Maloeran | IriX64, most such clusters use NUMA, you have multiple memory banks and you want each processor to work on its own local data |
| 19:56.43 | Maloeran | Erik, what do you want me to paste there? |
| 19:56.47 | brlcad | er, most commercial at least |
| 19:56.56 | IriX64 | i envision a pool instead of banks maybe a little more overhead but doable. |
| 19:57.44 | Maloeran | You are going to choke your memory bus if the 40 processors all go to the same memory bank |
| 19:57.45 | ``Erik | oh, hah, woops |
| 19:57.46 | ``Erik | sorry |
| 19:57.51 | ``Erik | http://paste.lisp.org/display/37073 |
| 19:57.57 | ``Erik | pasted the wrong piece |
| 19:58.02 | IriX64 | a pool of banks Malorean. |
| 19:58.37 | Maloeran | The banks are local to each processor for a good reason, it's faster to access them due to the reduced routing circuitry |
| 19:59.02 | Maloeran | Woohoo, that's neat Erik. What's the code? |
| 19:59.42 | ``Erik | http://math.missouristate.edu/~erik/mandelbrot-as.s |
| 20:00.51 | brlcad | typo on line 562 |
| 20:00.57 | brlcad | *ahem* :) |
| 20:01.03 | ``Erik | it's generated |
| 20:01.05 | ``Erik | so it's fugly |
| 20:01.55 | ``Erik | (but it's generated, so if there's a typo, it's gonna appear many places |
| 20:01.58 | brlcad | sure you're not blowing the heap or something |
| 20:02.02 | Maloeran | It runs for me after a quick fix, outputting nice letters |
| 20:02.14 | ``Erik | what quick fix? |
| 20:02.29 | ``Erik | brlcad: it works dandy on fbsd, and the C varients runs dandy with teh same sized heap |
| 20:02.38 | Maloeran | Renaming _start to main and gcc foo.s -o foo |
| 20:02.47 | ``Erik | hmmmmmmmm |
| 20:03.11 | ``Erik | that, uh |
| 20:03.16 | ``Erik | <-- scratches his chin |
| 20:03.35 | brlcad | if renaming _start does the trick.. that would likely be some frame setup / heap problem .. |
| 20:03.40 | ``Erik | yeahhhh |
| 20:04.04 | ``Erik | maaaybe I need to actually walk the heap and set everything to 0 |
| 20:04.12 | Maloeran | How did you create your faulty executable? |
| 20:04.28 | ``Erik | hrm? I pasted the line up there a bit |
| 20:04.53 | ``Erik | heh, it works with teh same sized heap in other places, dude... I bet it's an init thing |
| 20:05.19 | ``Erik | perhaps fbsd sets memory in the fork or exec syscall and linux depends on the program startup code in libc to do it? *shrug* |
| 20:05.20 | Maloeran | Erik, what commands did you use to assemble and link? |
| 20:05.30 | ``Erik | <``Erik> $ uname -mo ; as -o mandelbrot-as.o mandelbrot-as.s && ld -o mandelbrot-as mandelbrot-as.o && ./mandelbrot-as ; echo $? |
| 20:05.39 | Maloeran | Ah right, sorry |
| 20:06.33 | brlcad | heh |
| 20:06.35 | brlcad | "What's going to happen to the 3D market once Google licenses and further develops BRL-CAD? I can see them doing this but releasing it on one of their appliances." - Paul Bowers, USA |
| 20:07.41 | ``Erik | http://www.upfrontezine.com/2005/upf-431.htm ? |
| 20:08.04 | ``Erik | hehehe |
| 20:11.31 | IriX64 | wheres Mr. Tromey? why is there no progress indicator on this thing ;) |
| 20:12.05 | louipc | huh? Google is taking over BRL-CAD? |
| 20:12.40 | Maloeran | By the way Erik, your .s file works for me with _start too, I just had the reflex of letting gcc do the linking |
| 20:17.51 | brlcad | IriX64: on what? |
| 20:18.22 | Maloeran | That's quite slow, Erik ;) |
| 20:18.41 | Maloeran | If you really are working on a Brainfuck optimizer that is |
| 20:18.45 | ``Erik | well, ok |
| 20:18.49 | ``Erik | what's a fast way to zero the heap |
| 20:19.39 | Maloeran | I don't suppose you want to hear about a SSE loop with prefetches, but at least use rep movsq or rep movsd |
| 20:21.48 | ``Erik | stosd looks better |
| 20:22.01 | Maloeran | Oops, indeed |
| 20:24.19 | Maloeran | instructions also |
| 20:38.42 | IriX64 | http://pastebin.com/884697 |
| 20:38.47 | IriX64 | it failed. |
| 20:44.00 | IriX64 | http://pastebin.com/884699 <this is what i get running them manually with no switches. |
| 20:47.02 | Maloeran | Ah, aren't autoconf, automake and libtool just great |
| 20:49.45 | IriX64 | http://pastebin.com/884705 <==== this is what the gebnerated configure does, no jokes about my system please. |
| 20:50.12 | IriX64 | remember i have successfully built from this source and ran the product. |
| 20:50.45 | louipc | I'm getting problems and I'm on linux :P |
| 20:51.02 | IriX64 | ran then in this order aclocal, autoconf, autoheader and automake. |
| 20:52.11 | IriX64 | running the script(when it succeeds, the syntax errors are not there) |
| 20:53.10 | IriX64 | confirmed, i just ran the script, automake fails but the generated configure does its job. |
| 20:57.21 | IriX64 | now it can't find x |
| 20:58.12 | IriX64 | and it tries to restart using config.site *doh* this is first time. |
| 20:59.22 | IriX64 | errands to run l8r all |
| 21:05.41 | ``Erik | heh, that'd put me over 20 makefiles, dude |
| 21:06.23 | ``Erik | aaaand, very interesting, if the first thing I do in this program is attempt to exit, it still fails... hmmm |
| 21:06.47 | brlcad | IriX64 isn't exactly the typical autoconf'ing user .. he seems to mindlessly edit files and then wonder why stuf breaks |
| 21:07.41 | ``Erik | heh, or extract an old tarball, copy one very new (and incompatable) file in and try that |
| 21:08.07 | ``Erik | at least he's trying :) |
| 21:09.15 | brlcad | problem is he starts on faulty premises *most* of the time.. tools not setup/installed correctly, stuff missing, stuff edited incorrectly, mismatches |
| 21:09.41 | brlcad | i mean, autoreconf/autogen.sh failed on him and yet he still tried to run configure... |
| 21:10.55 | brlcad | Maloeran: and as for his specific errors.. he is using a new configure.ac against old sources that I told him outright would not work, yet he insists on trying |
| 21:17.14 | dtidrow | brlcad: must need to use a bigger cluebat ;-) |
| 21:20.56 | louipc | mmm mv m4/mkdirp.m4 misc/, autogen, configure, make, make install fails with same error |
| 21:21.29 | louipc | how do you set that variable in the Makefiles? |
| 21:21.43 | louipc | (I'm kind of a noob) |
| 21:22.39 | louipc | oh no wait |
| 21:23.05 | louipc | ignore me |
| 21:24.40 | ``Erik | I just saved a bunch of money on my car insurance by threatening my agent with a golf club! |
| 21:49.25 | brlcad | louipc: hm, perhaps move mkdirp.m4 back and edit that file |
| 21:49.42 | brlcad | then add: |
| 21:49.52 | brlcad | MKDIR_P="$mkdir_p" |
| 21:50.11 | brlcad | AC_SUBST([MKDIR_P]) |
| 21:50.23 | brlcad | before the last AC_SUBST at the end of the file |
| 21:50.55 | louipc | you put that as the last two lines of the file yeah |
| 21:51.22 | brlcad | yeah, typo'd |
| 21:51.35 | brlcad | just noticed |
| 21:51.48 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/m4/mkdirp.m4: typo |
| 21:51.59 | brlcad | have to put them BEFORE the last line, else it's not inside the macro |
| 21:52.17 | louipc | oooh |
| 21:55.34 | louipc | I see now |
| 22:09.58 | ``Erik | http://xkcd.com/c180.html |
| 22:15.03 | brlcad | heh |
| 22:23.52 | ``Erik | I might print this one for my door... http://xkcd.com/c217.html |
| 22:53.12 | *** join/#brlcad IriX64 (n=who@bas3-sudbury98-1168056909.dsl.bell.ca) | |
| 23:03.25 | *** join/#brlcad docelic (n=docelic@212.91.112.46) | |
| 23:13.49 | Maloeran | I think I'm amazed that all these common chess openings and counter-moves have a name on wikipedia, every one of them |
| 23:14.43 | ``Erik | I'm amazed that there's a wikipedia page for "ray's urine" |
| 23:14.46 | ``Erik | :D |
| 23:14.53 | louipc | hey success |
| 23:15.28 | louipc | hah hah |
| 23:16.03 | brlcad | louipc: excellent |
| 23:22.11 | IriX64 | dropped back to distributed configure.ac, compiling now (thanks for the bleeding knuckles ;)) |
| 23:23.07 | louipc | ;) |
| 23:38.49 | brlcad | Twingy: frederick county has legislation going in to boost the county's solar incentive from 3k to 15k |
| 23:40.14 | Twingy | not too shabby |
| 23:40.46 | Twingy | a bit of a drive to work though |
| 23:41.39 | brlcad | :) |
| 23:42.08 | brlcad | that's got to be lobbying from Solar or whatever their name is out there |
| 23:42.22 | brlcad | that big solar panel company right off the interstate |
| 23:42.22 | Twingy | nano solar? |
| 23:42.29 | Twingy | oh, no clue |
| 23:42.52 | brlcad | you can't miss them driving down I70 .. massive factory covered in panels |
| 23:43.28 | Twingy | ah, haven't traveled down I70 since I went to buy my fireworks in 2003 I think |
| 23:43.46 | Twingy | maybe a different interstate I'm thinking of |
| 23:44.08 | Twingy | I'm not even sure if I've ever been on I70 |
| 23:44.24 | brlcad | I think BP Global might have bought them out a few years ago |
| 23:44.35 | brlcad | but the company's name was just "Solar" iirc |
| 23:44.58 | Twingy | too bad MD is only 20% still |
| 23:45.15 | Twingy | 15k means you'd have to spend 75k |
| 23:45.30 | Twingy | fed is 30%, but only 2k |
| 23:45.32 | brlcad | ahh, yep.. "BP Solar" now |
| 23:45.34 | brlcad | http://www.bp.com/sectiongenericarticle.do?categoryId=3050487&contentId=3060089 |
| 23:45.59 | brlcad | news claimed it would cover 50% |
| 23:46.25 | Twingy | I need to buy 2 more soon |
| 23:47.05 | brlcad | you could have had some interesting "neighbors" .. http://www.solarneighbors.com/ |
| 23:49.32 | Twingy | cnc'ing a jig right now |
| 23:51.08 | brlcad | ahh, that picture on bp is the frederick plant, apparently the main factory (and the company they took over was called Solarex (makers of solarex panels of course)) |
| 23:51.45 | Twingy | hopefully more people start putting panels on their roofs |
| 23:53.08 | Twingy | MRI? |
| 23:53.14 | Twingy | I did that a couple of times |
| 23:53.23 | brlcad | looks like it, but unspecified |
| 23:53.28 | Twingy | how many teslas was the scanner you were in? |
| 23:53.41 | brlcad | haven't done it yet |
| 23:53.44 | Twingy | oh |
| 23:54.10 | brlcad | if they're not going to give me the data, then they can go play with themselves in the scanner for all I care |
| 23:54.12 | Twingy | the resolution of the stuff they do for functional MRI case studies is pretty low |
| 23:54.31 | Twingy | only like 100-200 slices |
| 23:54.48 | brlcad | that was going to be my follow-up, but I almost expect it .. though nih does have the equipment for better.. just depends whom |
| 23:55.17 | Twingy | it depends on how long they are willing to keep you in the machine |
| 23:55.24 | brlcad | :) |
| 23:55.26 | Twingy | the really high detailed stuff takes several hours to collect |
| 23:55.33 | Twingy | some case studies are over night |
| 23:55.39 | Twingy | you just sleep in the machine |
| 23:56.53 | Twingy | chris cornell left audioslave |
| 23:57.18 | brlcad | so you're going to try and hire him |
| 23:57.32 | Twingy | not sure what use he would be to me :) |
| 23:57.46 | brlcad | a walking ipod |
| 23:57.59 | Twingy | right... |
| 23:58.15 | Twingy | you saw what happened to my last ipod... |
| 23:58.21 | brlcad | hah |
| 23:58.38 | brlcad | i can just picture it now |
| 23:59.24 | brlcad | that'd be a great family guy flashback |
