| 00:23.08 | *** join/#brlcad ibot (i=ibot@rikers.org) |
| 01:51.21 | Twingy_ | mmm |
| 01:51.28 | Twingy_ | clone svideo display works now |
| 01:52.38 | Twingy_ | only 2 hours left for optimized kd-tree |
| 01:53.21 | Maloeran | Two hours? Shouldn't you have the whole night left? :) |
| 01:53.47 | Twingy_ | hrm? |
| 01:54.09 | Maloeran | Ah, I thought you had some deadline, but I understand that was an estimate |
| 01:54.26 | Twingy_ | I need to get a video of my ray-tracer ready by tommorrow |
| 01:54.35 | Twingy_ | going to make a quicktime |
| 01:55.06 | Twingy_ | just finished configuring my other machine with freebsd, it has a svideo dongle |
| 01:55.14 | Twingy_ | and dorking with X for the better part of 2 hours |
| 01:55.18 | Maloeran | Here, I dump raw frames on stdout and have mencoder encode on the fly from stdint |
| 01:55.39 | Twingy_ | right, but then I have to encode time intervals |
| 01:55.46 | Twingy_ | since all frames are not evenly spaced |
| 01:55.52 | Maloeran | Ah, okay. |
| 01:56.23 | Twingy_ | this way they can see me starting up the app and driving it |
| 01:56.30 | Twingy_ | with the mouse and what not |
| 01:57.16 | Twingy_ | the kd-tree caching kicks butt with the distributed setup |
| 01:57.26 | Twingy_ | the only downside is I gotta kick more data around to the nodes |
| 01:57.39 | Twingy_ | 300MB (geom + kdtree) to each node |
| 01:57.52 | Twingy_ | so that only requires 24 seconds of data transfer |
| 01:58.07 | Twingy_ | for my 15 mil polygon model |
| 01:58.13 | Maloeran | Good deal |
| 01:58.26 | Twingy_ | openrt is kinda lame like that |
| 01:58.36 | Twingy_ | they require you to have the file(s) on the nodes as well |
| 01:59.04 | Twingy_ | being able to just shoot the data to the nodes, which have no disks, is nice |
| 01:59.13 | Maloeran | I think I'll implement distributed processing that isn't even visible to the library user |
| 01:59.29 | Twingy_ | it's called MPI |
| 01:59.35 | Twingy_ | it was done 30 years ago |
| 01:59.37 | Twingy_ | :) |
| 01:59.41 | Maloeran | Fine :) |
| 02:00.03 | Twingy_ | my network code is larger than the actual engine |
| 02:00.14 | Twingy_ | since I'm handling all the little cases of what can go wrong |
| 02:00.36 | Maloeran | Ah yes, question : what "kind" of scene was Reschenov rendering to get 10 millions? A single object with abundant void around it? |
| 02:00.40 | Twingy_ | allowing nodes to connect whenever, dissapear, etc... |
| 02:00.53 | Twingy_ | he was walking around inside an office building |
| 02:01.17 | Maloeran | So the count of visible polygons was much lower than 10m at any time |
| 02:01.24 | Twingy_ | yah |
| 02:01.38 | Twingy_ | he didn't show any outdoor scenes |
| 02:01.42 | Maloeran | That's precisely the kind of scene my ray-tracer was thought for |
| 02:01.56 | Twingy_ | cool |
| 02:02.19 | Twingy_ | exactly the type of scene we don't do at ARL :) |
| 02:02.56 | Twingy_ | (but I try to support that sort of stuff off the clock) |
| 02:03.07 | Maloeran | Fine, it doesn't perform too bad in scenes where rays pass near most of the geometry either :) |
| 02:03.20 | Twingy_ | yep yep |
| 02:03.30 | Maloeran | I think I'll reach that 15m primary rays per second soon, for the tank model |
| 02:03.42 | Twingy_ | you should get some new geometry :) |
| 02:03.49 | Twingy_ | I've found some cases were my tracer bottomed out |
| 02:03.56 | Twingy_ | and made smarter in the kd-tree algorithm |
| 02:04.21 | Maloeran | The pipe scene isn't far from the worse cases I could imagine :) |
| 02:05.16 | Twingy_ | yep |
| 02:05.40 | Maloeran | At least for my ray-tracer, BSP doesn't care as much about spatial organisation and distribution |
| 02:07.01 | Twingy_ | welp, time to poke at this video some |
| 02:23.55 | silvap_ | so sick.. |
| 02:24.07 | Twingy_ | o.O |
| 02:24.33 | Twingy_ | hrm |
| 02:24.38 | silvap_ | think i have mono |
| 02:24.39 | Twingy_ | erik undid the delacktimes |
| 02:24.40 | silvap_ | :( |
| 02:24.44 | Twingy_ | bummer dude |
| 02:26.18 | Twingy_ | or not |
| 02:54.12 | Twingy_ | hrm |
| 02:54.27 | Twingy_ | 3.85 hours and going |
| 03:03.19 | Twingy_ | very tempting to wait for it to finish |
| 03:03.32 | Twingy_ | should be done within the hour |
| 03:03.58 | Twingy_ | then I could just put the video together in the morning |
| 03:04.03 | Twingy_ | hrms |
| 03:04.45 | Twingy_ | I think that's what I'll do |
| 03:05.31 | Twingy_ | okay, be back in about 7 hours |
| 03:06.51 | Twingy_ | ooh! |
| 03:06.54 | Twingy_ | it just finished |
| 03:07.05 | Twingy_ | happy dance! |
| 03:07.11 | Twingy_ | talk about timing |
| 03:23.53 | Maloeran | Ah :), that spares you the doubt of waking up to see a Segmentation fault |
| 10:00.54 | *** join/#brlcad cad028 (n=c139f720@bz.bzflag.bz) |
| 10:41.39 | *** join/#brlcad learner (n=brlcad@pcp0011649376pcs.aberdn01.md.comcast.net) |
| 10:41.39 | *** mode/#brlcad [+o learner] by ChanServ |
| 11:05.00 | Twingy_ | . |
| 11:21.02 | learner | ./ |
| 11:24.06 | Maloeran | :/ |
| 12:16.08 | Twingy_ | mmm |
| 12:16.16 | Twingy_ | sleeeeee3333eeeep |
| 12:23.39 | *** join/#brlcad d_rossberg (n=c28bf505@bz.bzflag.bz) |
| 12:47.08 | brlcad | yay, time to release soon |
| 15:10.37 | *** join/#brlcad cad241 (n=c0a7ccfd@bz.bzflag.bz) |
| 15:13.39 | *** join/#brlcad AlekosP (n=c0a7ccfd@bz.bzflag.bz) |
| 15:15.40 | *** join/#brlcad AlekosP (n=c0a7ccfd@bz.bzflag.bz) |
| 15:17.12 | AlekosP | need some help on building brlcad with Microsoft Visual C; anyone can help |
| 15:18.37 | brlcad | hello AlekosP |
| 15:18.53 | brlcad | what sort of help? |
| 15:22.08 | AlekosP | I tried to build with Vusual C++, but most of .dsp have problems; I think I fixed at least a part, but something strange happens with my TK-windows; is there anyone I can send the result of my efforts? |
| 15:23.06 | AlekosP | by the way, it's my first time on a chat, so my netiquette is poor: hello brlcad! |
| 15:23.33 | brlcad | heh, not a problem |
| 15:23.38 | brlcad | and welcome |
| 15:23.53 | brlcad | some actually live on irc ;) |
| 15:24.20 | brlcad | as for the build, that's not incredibly surprising -- there's two primary devs working on the windows builds |
| 15:25.03 | brlcad | there were a handful of fixes that have been made to cvs recently, you would be better off using the latest cvs than using a source tarball from the project site |
| 15:25.25 | brlcad | there's also build files in the misc/ directory that look like are more up to date |
| 15:26.44 | AlekosP | for that matter, I just checked on the cvs that the .dsp for the liboptical stuff is still missing the "r"s , so Microsoft becomes Micosoft and Target is spelt as Taget (and the project file isn't recognized!)... |
| 15:27.43 | AlekosP | ... I mean the one in the src/liboptical folder... |
| 15:28.53 | *** join/#brlcad Obscene_CNN (n=DiscoBan@66.15.214.167) |
| 15:29.20 | brlcad | heh |
| 15:29.22 | brlcad | hmm |
| 15:31.52 | brlcad | yep, looks like that one in particular had \r's stripped a really long time ago... |
| 15:32.41 | AlekosP | who is directly in charge of the Windows/Microsoft VC stuff? And is there a place where I can learn how to track/debug the flow of a tcl/tk application? For it seems to me that The problem I'm encountering are mostly of that nature |
| 15:33.59 | brlcad | I wouldn't say anyone's directly in "charge" of the VC projects, there are two devs that work on it from time to time |
| 15:34.18 | brlcad | you're more than welcome to maintain that build if you're interested ;) |
| 15:34.53 | Obscene_CNN | if your a masochist ;-) |
| 15:35.06 | AlekosP | well, I'm for sure interested, but I can't say if I'm able to... |
| 15:35.45 | brlcad | Obscene_CNN: hehe |
| 15:36.17 | brlcad | AlekosP: shouldn't be too much work to rewrite the liboptical project file, there's not much tricky about that library |
| 15:36.35 | AlekosP | Hi, Obscene_CNN, why shudder, if a project is open source, I think it should be as open as possible (of course working with MS VC isn't always comfortable) |
| 15:36.46 | brlcad | or even readd the missing r's .. i'll have to track it back through the history to see where they were stripped accidentally |
| 15:37.11 | brlcad | I should be able to restore it, but even then it may still be out of date - that goes back a couple years |
| 15:37.14 | Obscene_CNN | M$ VC doesn't always work as advertised. |
| 15:37.39 | Obscene_CNN | <PROTECTED> |
| 15:38.09 | brlcad | VC6 is pretty buggy, the newer versions are considerably better |
| 15:38.19 | AlekosP | I already have a working liboptical project file and a working mged.exe. Only it "almost" works; the tk windows behave quite strangely, get updated not when they are expected to... |
| 15:38.39 | brlcad | AlekosP: that's good to hear |
| 15:39.22 | brlcad | AlekosP: if you post up any mods to the sourceforge project patches tracker, I or one of the other guys can commit it |
| 15:39.35 | AlekosP | I agree with brlcad about vc7 beeing better than vc6 -- |
| 15:39.38 | brlcad | do that more than a couple times and it's an easy means to get your own commit access |
| 15:40.31 | brlcad | I'm always interested in getting new devs |
| 15:40.32 | AlekosP | well, brlcad. I need some help for posting my changes |
| 15:40.35 | brlcad | even if you do use windows :) |
| 15:40.44 | AlekosP | sorry ... |
| 15:40.48 | Obscene_CNN | M$ anounced they are going to ship the next version of their developer studio after the last beta. Users of it begged for another beta, but were denied. M$ also plans to introduce a major component without a beta test in the release version. |
| 15:40.50 | brlcad | hehe, j/k |
| 15:43.25 | AlekosP | so, brlcad, what if I simply pack only my modified sources (no objs or temporary things) in a tar/bz2 and I ship it to you directly? |
| 15:44.26 | AlekosP | and maybe also the executable + dll's so that you can look at the strange behavior of the gui windows...? |
| 15:46.49 | brlcad | sure |
| 15:47.31 | brlcad | or post up that tarbz to the patches tracker (without the exe/dlls) |
| 15:49.15 | AlekosP | I'll try the last one. Thanks. Bye everybody |
| 15:50.37 | brlcad | you might want to comment on the devs mailing list if you can describe the problems |
| 15:51.06 | brlcad | I'll be glad to take a look at it too, see what strange behavior you mean |
| 15:54.30 | AlekosP | something like this; you open a .g file from the db directory from the command windows of mged; then you browse the geometry and say to mged that you want the root of the model to be displayed; nothing happens. But if you go to the help menu of the graphics window, select the item "commands" and simply exit the menu, suddenly you get your model displayed in the graphics window |
| 15:56.57 | AlekosP | quite strange,isn't it? That's why I think it should be easier if you get a copy of the executable, and look directly; this is for sure just one of the problems... |
| 16:00.33 | brlcad | sounds like the display manager isn't updating |
| 16:00.46 | brlcad | I suspect if you resize the window, it'd draw |
| 16:02.40 | AlekosP | brlcad, I'll try to see if resizing does the work; thanks in advance. By now, bye everybody, I'll try to get in touch tomorrow (morning, for you) ... |
| 16:02.54 | brlcad | cya |
| 16:03.19 | brlcad | by the way, you might be interested in getting a better non-web-browser based irc client |
| 16:03.44 | brlcad | they're much "better" apps, the web interface is just to get folks in quick ;) |
| 16:04.17 | AlekosP | of course, so I can let you not that I'm say "looking" or something like that. Any suggestion? |
| 16:05.19 | AlekosP | [ ... I can let you know (not NOT) ...] cya |
| 16:07.12 | brlcad | http://irchelp.org |
| 16:07.21 | brlcad | xchat is one many use |
| 16:08.07 | brlcad | mirc is another |
| 16:08.11 | brlcad | http://www.ircreviews.org/clients/platforms-windows.html |
| 16:09.34 | brlcad | http://www.psyon.org/projects/xchat-win32/ |
| 17:22.30 | *** join/#brlcad DTRemenak (n=DTRemena@DHCP-170-143.caltech.edu) |
| 20:08.54 | CIA-4 | BRL-CAD: 03twingy * 10brlcad/src/adrt/libtie/ (define.h kdtree.c tie.c): about to try some stuff. |
| 20:11.41 | CIA-4 | BRL-CAD: 03brlcad * 10brlcad/TODO: better configure section output, m4 macro in order |
| 20:19.08 | CIA-4 | BRL-CAD: 03brlcad * 10brlcad/TODO: add legal headers on the manpages (as comments), update header.sh |
| 21:20.43 | *** join/#brlcad cad635 (n=450f9442@bz.bzflag.bz) |
| 21:22.39 | brlcad | hello cad635 |
| 21:34.00 | *** join/#brlcad cad293 (n=450f9442@bz.bzflag.bz) |
| 21:36.40 | cad293 | Just looking at BRLCAD for the 1st time today |
| 21:39.22 | brlcad | sounds good |
| 21:40.30 | brlcad | if you have any questions or get stuck somewhere, feel free to ask your questions |
| 21:40.41 | brlcad | someone will eventually answer :) |
| 21:42.54 | cad293 | are there any thermal analysis capabilities? |
| 21:43.36 | brlcad | there are limited analysis capabilities in general except those related to geometric structure or signal processing |
| 21:43.41 | brlcad | but yes |
| 21:43.55 | brlcad | there's an rttherm raytracer |
| 21:44.22 | brlcad | that uses a multispectral rendering to simulate thermal energy signatures |
| 21:44.53 | brlcad | hasn't been used in a long time by anyone here, but it's there |
| 21:58.16 | cad293 | ok, thanks, i'll probably be back after I get this installed |
| 21:58.21 | *** part/#brlcad cad293 (n=450f9442@bz.bzflag.bz) |
| 23:30.42 | Twingy | grumph |
| 23:31.39 | brlcad | said the grinch |