| 01:52.50 | *** join/#brlcad iraytrace (n=iraytrac@c-67-172-239-24.hsd1.co.comcast.net) |
| 02:06.57 | yukonbob | ?what's btclsh for |
| 02:07.59 | brlcad | it's a tclsh that has the brl-cad extensions preloaded |
| 02:08.53 | brlcad | e.g. run "bu_brlcad_data ." in btclsh |
| 02:09.00 | brlcad | it'll show you were the data path is |
| 02:09.00 | yukonbob | so that's why it's linked against libtclcad... is that the only way to get libtclcad functionality, or is it possible to load libtclcad into tclsh via package require? |
| 02:09.30 | brlcad | hm, I believe you could do a direct "load libtclcad.so" |
| 02:09.47 | brlcad | not sure if it'd need some pkgIndex.tcl or other Init() function, though |
| 02:09.50 | ``Erik | if a require could snarf up C code very easily, I'd imagine 'wish' would not be a standalone binary? *shrug* |
| 02:10.28 | yukonbob | "wish" is a convenience name... you can just as well run tclsh and package require tk. |
| 02:10.44 | brlcad | it auto-creates the top-level window too |
| 02:10.44 | ``Erik | aight, *shrug* I'm not a tcl guy |
| 02:10.57 | brlcad | so you can just start packing |
| 02:11.01 | yukonbob | brlcad: so will package require tk ;) |
| 02:11.02 | ``Erik | heh |
| 02:11.07 | brlcad | it does? |
| 02:11.29 | brlcad | ah, so it does |
| 02:11.35 | brlcad | learn something new every day |
| 02:11.43 | ``Erik | hum, heh, |
| 02:11.43 | ``Erik | % package require tk |
| 02:11.43 | ``Erik | can't find package tk |
| 02:11.46 | brlcad | Tk |
| 02:11.46 | yukonbob | Tk |
| 02:11.50 | ``Erik | okie |
| 02:12.31 | ``Erik | it might be possible to eliminate btclsh (or make it a 2 line tcl program) then *shrug* :) |
| 02:12.52 | brlcad | simh ftw! |
| 02:13.07 | ``Erik | hrm? |
| 02:13.13 | brlcad | just reading the backlog |
| 02:13.20 | ``Erik | :D |
| 02:13.27 | ``Erik | yeah, I wanna spin kermit up with a phone call |
| 02:13.43 | yukonbob | I'm stripping as much from the build as possible on my tcl/tk8.4 quest to use local libs... and it's currently puking on btclsh, but like ``Erik just said, it might be easily skipped... |
| 02:15.09 | yukonbob | does mged use btclsh? |
| 02:15.13 | brlcad | no |
| 02:15.20 | ``Erik | mged wires straight to libtcl |
| 02:15.32 | brlcad | mged is its own interpreter |
| 02:15.49 | ``Erik | TclInterp out the wazoo in src/mged/ :) |
| 02:15.58 | brlcad | rather, with an embedded tcl interpreter integrated with mged's own processing loop |
| 02:16.05 | ``Erik | just one? heh |
| 02:16.13 | brlcad | just one.. |
| 02:16.27 | yukonbob | so, might one be able to build the brl-cad suite w/o btclsh and still be reasonably functional? |
| 02:16.34 | brlcad | sure |
| 02:16.43 | ``Erik | completely functional as far as I can tell |
| 02:16.54 | yukonbob | rock'n'roll... |
| 02:16.59 | brlcad | not completely, stand-alone rtwizard uses it |
| 02:17.07 | ``Erik | ah, ok |
| 02:17.20 | ``Erik | is there a reason rtwizard couldn't use, say, mged -c ? |
| 02:17.26 | brlcad | the prep in the tclscripts dir will fail without it, so it'll rely on the existing precomputed results |
| 02:18.09 | ``Erik | (and actually... getting back to the original issue... btclsh will link against system libtcl iirc) |
| 02:18.13 | yukonbob | are those script doing CAD-ish stuff, or only Tcl-ish stuff? |
| 02:18.24 | ``Erik | src/btclsh/ is a pretty light directory |
| 02:18.43 | ``Erik | er |
| 02:18.46 | ``Erik | src/bwish |
| 02:18.54 | yukonbob | but if I've already got Tcl installed, I've got tclsh, and could just 'load libtclcad' and get "btclsh". |
| 02:19.05 | ``Erik | I mean, about a thousand lines of code for both btclsh and bwish |
| 02:19.10 | yukonbob | (if that in fact works...) |
| 02:19.32 | brlcad | yukonbob: rtwizard just does tclish stuff iirc |
| 02:19.54 | brlcad | i'm not *certain* you can just load libtclcad .. never tried it |
| 02:20.11 | yukonbob | brlcad: never a better time than the present... |
| 02:20.16 | brlcad | but in theory at least you should be able to, or at least come close to that after adding the hooks |
| 02:20.20 | ``Erik | blaze a trail, yukon :D see if it's possible, explore it! |
| 02:21.04 | brlcad | i imagine it won't work outright simply because there's nothing that says what function(s) it needs to invoke, and how to get access to the functions the lib provides |
| 02:21.34 | brlcad | when mged uses it, it just calls the function in C-land, and it gets passed the already created interpreter |
| 02:25.24 | brlcad | woo hoo, .bz's ISP agreed to let me keep my bandwidth at the same price .. so looks like I'll be doing the upgrade after all |
| 02:25.59 | brlcad | ``Erik: any special setup requests/suggestions (w.r.t. freebsd)? likely choose stable 6 otherwise |
| 02:32.22 | ``Erik | a modern 6 (6.2-stable) is a good place for an os to be |
| 02:32.52 | ``Erik | my request is not letting it go so outdated that updated ports fail :D and of course, I have 8 zillion ports I'd strongly recommend |
| 02:33.13 | ``Erik | portmanager, portaudit, uh, uptimed, ... bash cuz I'm lame... |
| 02:34.13 | ``Erik | but, hey, it's your box :) if you want, I'll help where I can, but I aint' got no final say, yo |
| 02:35.43 | ``Erik | awesome that they're grandfathering you in |
| 02:36.23 | ``Erik | might make it worth putting effort into putting up some fun simh with anonymous telnet (throttled, of course) |
| 02:37.41 | brlcad | it moved to 6 within just a month or two of getting installed |
| 02:38.46 | brlcad | otherwise, you've seen as much of the port upgrades as I have .. there were at least 6 full-port updates over the last two years |
| 02:39.13 | brlcad | and only that infrequent because they became harder and harder to do |
| 02:39.35 | brlcad | I think I'll pay the extra per month and upgrade the net link as well |
| 02:46.03 | ``Erik | 5.21 was more than 6mo before 6.0... like, 18 months |
| 02:46.33 | brlcad | i mean, it was the end of the stable line |
| 02:46.42 | brlcad | then a month or two later, it swapped to 6 |
| 03:10.37 | ``Erik | 5.3 and 5.4 were stables before 6.0 went stable |
| 03:10.46 | ``Erik | and there's a 5.5 out, too |
| 03:25.23 | yukonbob | libtool: link: warning: library `/usr/pkgsrc/bch/brlcad/work/.buildlink/lib/tcl/itk3.3/libitk33.la' was moved. |
| 03:25.42 | yukonbob | means |
| 03:25.42 | yukonbob | ... building mged... |
| 03:26.10 | yukonbob | I got similar when I was building bwish (which I've now just skipped), but I _do_ want mged... |
| 03:51.41 | ``Erik | 'was moved'? O.o |
| 03:52.27 | yukonbob | ya -- it's weird -- ..and then I get some basic Xlib (I think) missing references -- might be a pkgsrc thing -- I just updated my Makefile to see what shakes out... |
| 03:52.50 | yukonbob | do you know what it means? (/me guesses "no" based on your eyes O.o |
| 03:52.52 | yukonbob | ) |
| 03:52.59 | brlcad | it probably really was moved, something pkgsrc is doing maybe (where it was compiled for one path, but finally put into another, then libs linked with it in the final location) |
| 03:53.35 | yukonbob | the .la looks "normal" to me, though... |
| 03:53.47 | yukonbob | just failed again... get ready for short flood: |
| 03:53.50 | brlcad | it means what I wrote in the parenthetical .. which would be either an unclean build, or it really was moved, or you're getting something to link against two different itk33 libs |
| 03:53.54 | ``Erik | um |
| 03:53.57 | ``Erik | pastebin.bzflag.bz |
| 03:55.38 | yukonbob | http://www.pastebin.ca/725142 |
| 03:57.54 | ``Erik | hrmmmm what os again? debian/ubuntu? |
| 03:58.25 | yukonbob | netbsd |
| 03:58.31 | ``Erik | O.o hrm |
| 03:58.54 | brlcad | it's missing -Xi |
| 03:58.58 | brlcad | er, -lXi |
| 03:59.10 | ``Erik | what lib is, say, XOpenDevice in there? pehaps Xi? |
| 03:59.11 | ``Erik | heh |
| 03:59.13 | ``Erik | beat me to it |
| 04:00.35 | brlcad | that should be fixed on head already |
| 04:01.01 | brlcad | assuming you really do have a usable Xi and that our test isn't flawed |
| 04:01.11 | yukonbob | we'll find out ;) |
| 04:01.38 | brlcad | grep X_LIBS Makefile |
| 04:01.43 | brlcad | what does it report? |
| 04:02.14 | yukonbob | is building ... 1 sec. |
| 04:02.35 | brlcad | Tk functionality test also includes Tcl |
| 04:02.48 | brlcad | the Tk lib test is an independent symbol search |
| 04:02.58 | brlcad | shouldn't need tcl for that |
| 04:03.25 | brlcad | (if you did, pretty much all the lib symbol tests would fail for more than a dozen libs) |
| 04:09.07 | yukonbob | thx. for help w/ all this folks -- I hope something good comes of it for both of us ;) |
| 04:09.45 | yukonbob | brl-cad supermodel edition *Now Less Calories!!* |
| 04:10.04 | yukonbob | actually "supermodel" is a good name, in this case... |
| 04:10.13 | yukonbob | (tm)! |
| 04:11.56 | yukonbob | so, is archer X11 too, or just Win/Mac? |
| 04:12.41 | louipc | it's buggy in linux |
| 04:13.05 | louipc | well buggy to the point I can't get anything other than a little grey box |
| 04:13.12 | yukonbob | where linux == linux/bsd/solaris/*nix -- or just linux? |
| 04:13.35 | louipc | I don't have those other *nix so i couldn't say |
| 04:13.48 | louipc | but that would seem like a logical assumption |
| 04:14.03 | yukonbob | is tcl/tk, fltk, gtk, other? |
| 04:14.33 | louipc | tcl/tk |
| 04:15.14 | yukonbob | idea is it's primarily WIMP interface, vs. mged's keyboarding? |
| 04:15.46 | louipc | I hope it has command line! |
| 04:25.56 | yukonbob | !yaY |
| 04:53.15 | brlcad | yukonbob: archer is pure tcl/tk |
| 04:53.33 | yukonbob | nice... what are the plans for mged? |
| 04:53.47 | brlcad | louipc: I just fixed archer a couple days ago... it should work everywhere now if I got took care of all the loading issues |
| 04:54.37 | brlcad | archer actually includes a half embedded and half-rewritten mged |
| 04:55.06 | louipc | oh sweet |
| 04:55.16 | yukonbob | where did the name "archer" come from? |
| 04:55.16 | brlcad | it was a prototype interface intended to maybe some day replace mged or at least show what's possible even via tcl/tk with a different style gui |
| 04:55.45 | brlcad | it was part of this joint (army, navy, air force) SBIR funded project called "Crossbow" |
| 04:56.03 | brlcad | archer was one of the tools in the project, one of several similarly named tools |
| 04:56.08 | yukonbob | ahh |
| 04:56.29 | yukonbob | arrow, target, robinhood, apple? |
| 04:56.44 | brlcad | the project dealt with vulnerability/lethality calculations, so naming it with things that throw projectiles was quite appropriate |
| 04:56.52 | louipc | the marines don't need cad eh? |
| 04:57.10 | brlcad | marines don't do research :) |
| 04:57.16 | yukonbob | heh |
| 04:58.14 | brlcad | archer still keeps a command console, with most of the core mged commands still available, and shows how things could work with various tabbed information panels, gui buttons, and more |
| 04:58.27 | brlcad | navy uses brl-cad |
| 04:59.43 | brlcad | they're just late to the V/L game, it took a catastrophic event like a ship bombing to get them thinking about paying more attention to their vulnerabilites |
| 05:01.46 | louipc | I'd have thought that would be the first thing you'd think of when building any type of military vehicle heh |
| 05:02.41 | yukonbob | louipc: are you a brl-cad dev? |
| 05:03.06 | louipc | nope just an interloper |
| 05:03.13 | yukonbob | :) |
| 05:03.48 | brlcad | louipc: it was federally mandated that everyone *has* to think about it before vehicles go into use |
| 05:04.01 | brlcad | but the level of that thought varies widely, as do the needs |
| 05:04.20 | brlcad | the V/L characteristics of a ship vs. a plane vs. a tank are *massively* different |
| 05:04.38 | brlcad | i mean, if a plane is hit, it's pretty much a critical kill |
| 05:04.59 | brlcad | if a tank is hit, it very well might actually bounce off |
| 05:05.21 | yukonbob | is that a function of the plane/tank, or just their geometric properties, though? |
| 05:05.46 | louipc | both I think |
| 05:05.48 | yukonbob | tank == thicker angled armour, plane == thin aluminum tubes or carbon fiber, etc. |
| 05:05.50 | brlcad | depends on both, and varies heavily asset to asset |
| 05:05.59 | brlcad | as well as the threat |
| 05:07.08 | louipc | like the tail prop of a chopper is a critical yet delicate part. if that's taken out the whole thing comes down.. even if the rest is intact |
| 05:07.41 | yukonbob | well, brlcad wouldn't know that, though, would it? |
| 05:07.43 | brlcad | either way, the vast *majority* of all V/L analyses are used by a single code that directly hooks into BRL-CAD for geometric analyses answering questions like "what components are along this shotline", so the more V/L work going on, the more folks (in DoD) are using BRL-CAD (directly and indirectly) |
| 05:08.18 | louipc | brlcad does the modelling... |
| 05:08.56 | brlcad | the modeling, and the *geometric* analysis (which is almost entirely what drives the *V/L* analyses, which then add in their own algoritms) |
| 05:09.34 | brlcad | they deal with the physics of penetrations, simulating interactions, velocities, energy, etc .. BRL-CAD answers all the questions about the geometry |
| 05:10.33 | brlcad | like what material is the object, where is it physically, what regions is it comprised of, what mass/volume/moments/etc does it have, how thick is something from a given angle, what's along a shotline, etc |
| 05:11.03 | brlcad | yukonbob: cool, was there much backport changes? |
| 05:11.23 | yukonbob | mostly the build process... though I've only been running it ~30s |
| 05:11.34 | brlcad | no tcl/code changes? |
| 05:11.39 | yukonbob | not yet. |
| 05:12.00 | yukonbob | ... but if it's OK with you, I'd really appreciate 8.4 support, and can help w/ determining needs to that end... |
| 05:12.45 | yukonbob | (where support == consideration in future code for 8.4 compat) |
| 05:13.27 | brlcad | I could possibly motivate/convince a back-port to 8.4 to the folks that will care, but I'd need to know the exact steps/patching needed and test it myself first |
| 05:16.04 | yukonbob | btw, 'load ...libtclcad.so' yields "couldn't find procedure Tclcad_Init". |
| 05:16.08 | louipc | is tcl beta in brlcad now? |
| 05:18.22 | brlcad | ah, cool -- figured they'd auto-search for some _Init() function |
| 05:18.31 | brlcad | no, beta's not in yet |
| 05:18.49 | brlcad | wanna add it? :) |
| 05:19.32 | louipc | :D |
| 05:20.03 | louipc | I'd love to but other things are grabbing my time |
| 05:21.30 | yukonbob | now my rtarea works, presumably so will rtedge, among others ;) |
| 05:23.03 | yukonbob | wireframe in mged gfx window seems slower (compared to 7.8.x) -- anybody else notice this? |
| 05:23.15 | yukonbob | *slower rotation w/ x,y,z |
| 05:25.56 | yukonbob | for arb*, does it matter what order the points are entered? |
| 05:41.17 | brlcad | yes, it does |
| 05:41.33 | brlcad | gfx may actually be slower if you compile optimized vs non-optimized |
| 05:41.41 | brlcad | or with/without opengl |
| 05:41.57 | brlcad | otherwise, it's all in your head |
| 05:42.19 | brlcad | it reports a fps in the faceplate |
| 05:42.32 | brlcad | (on the misc options in the menu) |
| 06:14.56 | *** join/#brlcad Z80-Boy (i=clock@77-56-81-225.dclient.hispeed.ch) |
| 07:09.46 | *** join/#brlcad ``Erik (i=erik@c-69-250-155-85.hsd1.md.comcast.net) |
| 07:20.12 | *** join/#brlcad Z80-Boy (n=clock@zux221-122-143.adsl.green.ch) |
| 09:41.18 | *** join/#brlcad elite01 (n=elite01@dslb-088-070-017-073.pools.arcor-ip.net) |
| 11:48.48 | *** join/#brlcad thing0 (n=ric@203-59-35-171.dyn.iinet.net.au) |
| 12:35.12 | *** join/#brlcad elite01 (n=elite01@dslb-088-070-023-115.pools.arcor-ip.net) |
| 13:59.28 | ``Erik | unthrottled.... |
| 13:59.28 | ``Erik | Abs vgr2 2261.62 1051.17 809.07 771.05 1003.79 1179.34 |
| 13:59.29 | ``Erik | *vgr vgr2 16.50 15.67 14.42 14.44 14.19 15.04 |
| 14:32.41 | Z80-Boy | Now the Ronja videos are recompiled without the flashes and uploaded to the Ronja page, woohoo :) |
| 14:33.42 | archivist | one for ``Erik as a connesiur of amusing pics http://www.nickscipio.com/funstuff/archive8/2005-05-14_butttrumpet.html |
| 14:51.33 | Maloeran | Argh, please don't post stuff like that |
| 15:05.49 | *** join/#brlcad elite01 (n=elite01@dslb-088-070-023-115.pools.arcor-ip.net) |
| 15:16.48 | ``Erik | heh |
| 15:16.59 | ``Erik | I'll assume it's not work safe and will avoid opening it O.o |
| 15:20.13 | ``Erik | such awesome hardware, with 45 minute renderings |
| 15:30.31 | archivist | a motorcycle, a bare butt, a can of beans and a traffic cone |
| 15:31.00 | *** join/#brlcad elite01 (n=elite01@dslb-088-070-023-115.pools.arcor-ip.net) |
| 15:33.18 | ``Erik | that, um... |
| 15:33.19 | ``Erik | uh... |
| 16:48.35 | *** join/#brlcad poolio (n=poolio@c-71-206-215-46.hsd1.pa.comcast.net) |
| 16:53.40 | Z80-Boy | brlcad: basically all ronja models - including those with complicated screws - render fast |
| 16:53.54 | Z80-Boy | except for 2 models where a screw is cut apart, which render orders of magnitude slower |
| 16:53.59 | Z80-Boy | Isn't that suspicious? |
| 16:54.11 | Z80-Boy | Even when the screw actually occupies only tiny fraction of the picture |
| 17:04.18 | ``Erik | how is it cut? |
| 17:04.43 | ``Erik | the 'cut' primitive can make things really really really slow |
| 17:05.27 | MinuteElectron | brlcad: I guess I wasn't here 'in a few minutes' lol. |
| 17:05.27 | Z80-Boy | it's a boolean operation |
| 17:05.27 | Z80-Boy | with a rpp |
| 17:05.27 | Z80-Boy | either subtract or and, both slow |
| 17:05.49 | ``Erik | hum, depending on how the tree is built, you might be doing way more weaves than you need to be |
| 17:10.53 | ``Erik | (and brlcad probably wont' be here for several hours... he might even be too busy until mondayish) |
| 17:11.15 | MinuteElectron | screwin @ me |
| 17:18.15 | *** join/#brlcad poolio_ (n=poolio@c-71-206-215-46.hsd1.pa.comcast.net) |
| 17:21.46 | yukonbob | http://www.methodlogic.net/flat/gfx/disp_error.png |
| 18:56.14 | *** join/#brlcad Z80-Boy (i=clock@77-56-91-179.dclient.hispeed.ch) |
| 19:47.31 | ``Erik | nifty, looks like vgr was running around 345Kips |
| 19:58.02 | dtidrow_work | I forget - what does the benchmark program summary mean again? |
| 20:02.40 | ``Erik | um, which part? |
| 20:02.55 | ``Erik | the vgr bit? |
| 20:03.20 | dtidrow_work | <PROTECTED> |
| 20:03.23 | dtidrow_work | arg |
| 20:03.35 | ``Erik | oh, heh :D |
| 20:03.49 | ``Erik | the Abs line is second sof computation for each image, then the mean |
| 20:03.53 | dtidrow_work | that stuff - let me get the recent one from my lappy |
| 20:04.05 | ``Erik | the *vgr line is how many vgr's worth of unf that is |
| 20:04.19 | ``Erik | where vgr was a vax 11/780 running BSD 4.2/4.3 |
| 20:04.47 | dtidrow_work | ah |
| 20:04.51 | dtidrow_work | Abs localhost.localdomain 1035035.81 467863.48 482449.25 374872.08 513705.77 540954.26 569146.77 Sat Jun 30 14:15:09 EDT 2007 |
| 20:04.51 | dtidrow_work | *vgr localhost.localdomain 7554.45 6976.78 8604.40 7025.33 7267.02 36.49 6244.07 |
| 20:04.52 | ``Erik | vgr2 is a simulated 11/780 running 4.3BSD, I'm trying to tune the sim to give me the same results as the old hw |
| 20:05.11 | dtidrow_work | that's the laptop benchmark |
| 20:05.17 | ``Erik | you should have a summary file that explains all the results |
| 20:05.38 | dtidrow_work | hmmm, the 'summay' file just had what I pasted |
| 20:05.47 | ``Erik | there should be a lot more in it O.o |
| 20:06.10 | ``Erik | sorry, I'm wrong |
| 20:06.53 | ``Erik | when ya ran 'make benchmark', it should've printed all that out... |
| 20:06.58 | ``Erik | mebbe in your run- log? |
| 20:09.08 | ``Erik | here's what my run log for my desktop looks like... http://pastebin.bzflag.bz/mbef4b9d |
| 20:22.51 | dtidrow_work | ah, looks like I clobbered the binaries to free up some disk space - and probably killed the log file along with it |
| 20:23.05 | ``Erik | 'make clean' will delete the .log and .pix files |
| 20:25.21 | dtidrow_work | ``Erik: how many cpus are on the box for that pastebin link? |
| 20:26.18 | ``Erik | um, 8 cores... two quad-core intel thingies |
| 20:27.01 | dtidrow_work | ah |
| 20:27.17 | ``Erik | it does like 27000 vgr's when optimized |
| 20:27.27 | dtidrow_work | w00t! |
| 20:28.17 | dtidrow_work | wonder what the quad-core AMD's are gonna do... |
| 20:32.26 | ``Erik | I think my dual two-core 2.0ghz opteron gets around 11500ish? |
| 20:34.56 | dtidrow_work | so the first number in the '*vgr' line is how many vgr's your box compares to? |
| 20:35.34 | ``Erik | no, um, well... only for moss.g |
| 20:35.42 | ``Erik | the last one is the average |
| 20:35.48 | dtidrow_work | ah |
| 20:36.32 | ``Erik | it's bench/run.sh if you wanna poke |
| 20:39.20 | ``Erik | <-- has no clue why you only got 36 vgr's on sphflake O.o |
| 21:23.31 | CIA-4 | BRL-CAD: 03erikgreenwald * 10brlcad/src/other/libpng/ (90 files in 11 dirs): Upgrade to 1.2.21. Fixes several bugs introduced in 1.2.19, possible overflows, and has some code cleanup. |
| 22:10.03 | *** join/#brlcad Elperion (n=Bary@p54877B93.dip.t-dialin.net) |
| 22:20.39 | *** join/#brlcad cad53 (n=8614cf1d@bz.bzflag.bz) |
