| 00:11.22 | IriX64 | the fairies just dropped in, going to be a good spring, summer, fall and winter haven't made up their minds yet |
| 00:58.52 | *** join/#brlcad ki6ijg (n=wegstar@h69-128-116-234.69-128.unk.tds.net) | |
| 05:37.12 | *** join/#brlcad clock_ (i=clock@84-72-89-172.dclient.hispeed.ch) | |
| 06:38.37 | *** join/#brlcad louipc_ (n=louipc@bas8-toronto63-1177614277.dsl.bell.ca) | |
| 06:47.32 | *** join/#brlcad SWPadnos_ (n=Me@dsl245.esjtvtli.sover.net) | |
| 06:50.56 | *** join/#brlcad louipc (n=louipc@bas8-toronto63-1128543684.dsl.bell.ca) | |
| 07:38.57 | *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch) | |
| 07:40.44 | CIA-21 | BRL-CAD: 03d_rossberg * 10brlcad/src/librt/g_brep.cpp: |
| 07:40.44 | CIA-21 | BRL-CAD: VC++ 6.0 does not like near, far (obsolete keyword error) |
| 07:40.44 | CIA-21 | BRL-CAD: operator= should return a reference to *this |
| 11:40.00 | *** join/#brlcad dtidrow_work (n=dtidrow@host66.objectsciences.com) | |
| 12:46.52 | *** join/#brlcad louipc (n=louipc@bas8-toronto63-1128543684.dsl.bell.ca) [NETSPLIT VICTIM] | |
| 12:46.52 | *** join/#brlcad Maloeran (n=maloeran@glvortex.net) [NETSPLIT VICTIM] | |
| 13:55.04 | CIA-21 | BRL-CAD: 03erikgreenwald * 10rtcmp/ (tri.h adrt/adrt.c): use linear fastf_t buf for tri data instead of typedef on typedef on typedef |
| 13:58.24 | CIA-21 | BRL-CAD: 03erikgreenwald * 10rtcmp/tri.c: |
| 13:58.24 | CIA-21 | BRL-CAD: Move to linear fastf_t buffers. Radical simplification. Proper endian |
| 13:58.24 | CIA-21 | BRL-CAD: safety added/tested. |
| 14:59.26 | Maloeran | You aren't the only one |
| 14:59.34 | Maloeran | http://news.bbc.co.uk/2/hi/technology/6506027.stm - Yay |
| 15:01.37 | *** join/#brlcad louipc_ (n=louipc@bas8-toronto63-1088754359.dsl.bell.ca) | |
| 15:08.42 | ``Erik | oh, but they'll install the wrong linux, not only htat, they'll install it all wrong |
| 15:08.50 | ``Erik | :D |
| 15:17.15 | *** join/#brlcad Elperion (n=Elperion@p5487493B.dip.t-dialin.net) | |
| 15:22.09 | clock_ | ``Erik: and then all will turn into hardcore trolls and accuse each other from trolling in a huge flamewar tsunami |
| 15:22.31 | ``Erik | O.o |
| 15:23.10 | clock_ | ``Erik: do you mean the same modding as in slashdot? |
| 15:23.15 | ``Erik | yeah :D |
| 15:23.17 | clock_ | What does -1 mean? |
| 15:23.23 | ``Erik | down a point... |
| 15:23.52 | ``Erik | usually it's like "+1 insightful" or "-1 troll" |
| 15:23.58 | clock_ | can you mod -2? |
| 15:24.07 | ``Erik | but, uh, relevance seems to be a bad thing in slashdot |
| 15:24.07 | clock_ | aha -1 relevant hehe |
| 15:24.25 | *** join/#brlcad dtidrow_work (n=dtidrow@host169.objectsciences.com) | |
| 15:24.30 | clock_ | yeah and relevance also increases the probability that you'll be accused from trolling |
| 15:24.30 | ``Erik | no, but if enough people mod up, the score goes up |
| 15:25.03 | clock_ | and then you can see the statistics how many people said troll how many relevant how many insightful etc.? |
| 15:25.28 | clock_ | interesting that projects like gEDA, BRL-CAD or Ronja usually have an easy discussion |
| 15:25.54 | clock_ | and projects like Linux or OpenBSD or Gentoo are very sensitive to exploding into a trolllabelling fit |
| 15:26.07 | ``Erik | people get religion about funny things |
| 15:26.25 | clock_ | i. e. in-depth discussion is impossible |
| 15:27.11 | ``Erik | it's the people with little to no knowledge that really get flamey about things, usually |
| 15:27.32 | clock_ | or personality disorder developers like theo de raadt |
| 15:27.48 | ``Erik | they'll spot one minor wart on whatever isn't their favorite and focus on it, ignoring the 8 zillion good things |
| 15:27.55 | clock_ | ;-) |
| 15:27.58 | ``Erik | and conveniently ignoring the warts on their own |
| 15:28.41 | clock_ | You mean realloc is slow when it has to move the block? |
| 15:29.50 | ``Erik | doug leas malloc does slight of hand to remap pages, so realloc across page boundries is pretty cheap... phkmalloc forces memory contiguous, so if you can't take the next page, it allocates in a completely new area, copies the old data, then frees the old area |
| 15:30.32 | ``Erik | the pathological case gives you a malloc/copy/free cycle every realloc |
| 15:31.24 | ``Erik | if you have a program that reads a file stupidly, like read a page, realloc, read the next page, ad nauseum... it shows :) (but if you were to, say, stat the file, allocate the size and read it... or mmap, it ain't an issue) |
| 15:31.39 | *** join/#brlcad joevalleyfield (n=joevalle@bz.bzflag.bz) | |
| 15:32.26 | ``Erik | chaining buffers linked list style also "fix" it at the cost of losing contiguouity (sp?) during the read |
| 15:32.37 | ``Erik | oi, timbert |
| 15:33.44 | joevalleyfield | my presence is random |
| 15:35.23 | ``Erik | well, I d'no |
| 15:35.33 | ``Erik | he hides in a dark cave with a club and screams at children |
| 15:37.44 | ``Erik | and EATS TEHIR BONES |
| 15:37.55 | clock_ | BONES!!!!11111 |
| 15:38.04 | ``Erik | !!@!~@~!!!~!ONE!~@!~ |
| 15:39.04 | clock_ | actually I thought it should be possible to use BRL-CAD to produce some kind of industrial 3D rendered art |
| 15:39.25 | ``Erik | in what sense? |
| 15:39.34 | clock_ | I saw often posters for technical things including some kind of artistically processed blueprints, diagrams or wireframe models |
| 15:40.06 | clock_ | so one could either use rt-edge to produce such linedraw pictures, or straight rt to make colour pictures |
| 15:40.13 | clock_ | modeling of stylized technical elements should be easy |
| 15:40.17 | ``Erik | oh, sure, um, the plot routines generate rough wireframe, or you could tesselate and generate wireframe... and if you want to raytrace things, the shaders are fairly easy to write, check out src/liboptical |
| 15:40.46 | clock_ | can one write a script / little program to generate "landscapes" in ASCII? |
| 15:40.56 | clock_ | To make e.g. repeating changing patterns etc.? |
| 15:41.20 | ``Erik | um, there is a program to generate random dsp's |
| 15:41.28 | clock_ | what is a random dsp? |
| 15:41.31 | ``Erik | which are good for gnenerating terrain |
| 15:41.36 | ``Erik | dsp is kinda like a heightfield map |
| 15:41.44 | clock_ | no I don't mean real landscape |
| 15:42.09 | ``Erik | like if you generated a 2d plasma image and fed that to POV as a heightfield |
| 15:42.11 | clock_ | I mean for example a bunch of low cylinders with different colours, positions and diameters |
| 15:42.31 | ``Erik | oh, I'd imagine a tcl script could do that easily enough |
| 15:42.45 | ``Erik | or if you want to get fancy, a procdb |
| 15:42.51 | clock_ | can I specify the geometry in ASCII without trying to parrot the ASCII dump? |
| 15:43.23 | ``Erik | by 'ascii dump', do you mean the .asc form of a .g file? |
| 15:43.24 | joevalleyfield | write an mged script and pipe it to mged? |
| 15:43.39 | clock_ | that's an idea |
| 15:43.46 | clock_ | can I do just cat script | mged |
| 15:43.48 | clock_ | ? |
| 15:43.48 | ``Erik | mged -c is nifty :) |
| 15:44.04 | clock_ | can't get man mged is there a page? |
| 15:44.15 | ``Erik | MANPATH=/usr/brlcad/man man mged |
| 15:45.48 | clock_ | wow |
| 15:46.19 | clock_ | hmm mged -c asks the user at start for something, is it gonna work? |
| 15:46.27 | clock_ | Or do I need to prepend an answer? |
| 15:46.37 | joevalleyfield | it should work |
| 15:46.47 | joevalleyfield | it should detect that the input isn't coming from a terminal |
| 15:47.13 | clock_ | oh yeah seems so and then it doesn't display the graphics |
| 15:47.18 | ``Erik | $ echo 'tops' | mged -c ktank.g |
| 15:47.18 | ``Erik | _GLOBAL ktank/ metaball.r/R |
| 15:47.18 | ``Erik | air/ ktank-facets |
| 15:47.24 | clock_ | that's actually quite handy |
| 16:44.12 | CIA-21 | BRL-CAD: 03erikgreenwald * 10rtcmp/partm.c: internal functions should be static |
| 16:45.48 | CIA-21 | BRL-CAD: 03erikgreenwald * 10rtcmp/dry/dry.c: return negative radius for bounding sphere, as this is not a legit engine |
| 17:23.32 | *** join/#brlcad louipc_ (n=louipc@bas8-toronto63-1128543513.dsl.bell.ca) | |
| 17:29.32 | CIA-21 | BRL-CAD: 03erikgreenwald * 10rtcmp/rtcmp.h: add notion of views |
| 17:30.19 | CIA-21 | BRL-CAD: 03erikgreenwald * 10rtcmp/perfcomp.c: add "golden rays", begin move to orthos view instead of "same ray" |
| 17:54.40 | CIA-21 | BRL-CAD: 03erikgreenwald * 10rtcmp/perfcomp.c: add some more "interesting" ray directions |
| 17:55.59 | CIA-21 | BRL-CAD: 03erikgreenwald * 10rtcmp/rtcmp.c: show ray packets for visual inspection... will be automated later |
| 17:57.46 | CIA-21 | BRL-CAD: 03erikgreenwald * 10rtcmp/perfcomp.c: uh, normalize the right dir vectors... |
| 18:05.12 | CIA-21 | BRL-CAD: 03erikgreenwald * 10brlcad/src/tclscripts/mged/openw.tcl: File->_R is raytrace, not renderview. Fixes bug 1687710 |
| 18:18.01 | ``Erik | amusing, the mailing lists on sf don't escape the body, so'z the it tries to parse C diffs as html in the web thingy O.o |
| 18:45.23 | brlcad | heh, or just mged -c ktank.g tops |
| 18:45.56 | brlcad | anything after the db throws it into single command mode |
| 19:00.07 | ``Erik | the idea was to have a long script that he can cat into it :) |
| 19:00.51 | *** join/#brlcad louipc (n=louipc@bas8-toronto63-1128544250.dsl.bell.ca) | |
| 19:01.08 | ``Erik | mged -c < cmdsoutthewazoo |
| 19:04.12 | louipc | eh? |
| 19:07.53 | brlcad | yeah, but then echo doesn't show that well :) |
| 19:08.10 | brlcad | a herenow doc also works well |
| 19:08.20 | brlcad | mged -c file.g <<EOF |
| 19:08.21 | brlcad | ls |
| 19:08.22 | brlcad | tops |
| 19:08.23 | brlcad | EOF |
| 19:11.10 | ``Erik | the echo was to vrfy tims statement :) |
| 19:13.55 | brlcad | ah, true dat |
| 20:04.30 | ``Erik | perhaps not o.O :D |
| 20:06.21 | *** join/#brlcad Bariton (n=Elperion@p548760E6.dip.t-dialin.net) | |
| 22:29.26 | *** join/#brlcad Ben (n=43be4c56@bz.bzflag.bz) | |
| 22:38.06 | *** join/#brlcad Benny (n=43be4c56@bz.bzflag.bz) | |