| 00:54.21 | *** join/#brlcad dan_falck (~danfalck@pool-71-111-62-241.ptldor.dsl-w.verizon.net) |
| 02:15.25 | *** join/#brlcad PKMOBILE (~Matthew@130.85.253.237) |
| 03:07.58 | *** join/#brlcad ChanServ (ChanServ@services.) |
| 03:07.59 | *** mode/#brlcad [+o ChanServ] by irc.freenode.net |
| 05:05.48 | *** join/#brlcad PrezKennedy (~Matthew@resnet-253-237.resnet.umbc.edu) |
| 05:18.26 | *** join/#brlcad ibot (ibot@apt.bot.TimRiker.active.supporter.pdpc) |
| 05:18.27 | *** topic/#brlcad is http://brlcad.org/ || BRL-CAD is now Open Source! || Screenshots: http://sourceforge.net/project/screenshots.php?group_id=105292 || http://brlcad.org/images/mged.jpg || Release 7.2.0 is posted (20050305) |
| 07:07.06 | *** join/#brlcad d_rossberg (~c28bf505@bz.bzflag.bz) |
| 12:20.41 | narnia | learner, there is a good reason for writing a compiler. the current part21 parser is the one originally written by nist. it is written in c++. not lex/flex or yacc/bison but straight c++. |
| 12:22.59 | narnia | also from looking at it again it does nothing to check the local or global 'where' clauses. |
| 12:43.21 | narnia | i will figure something just nevermind. |
| 13:40.48 | *** join/#brlcad polyspin (~butler@bz.bzflag.bz) |
| 13:41.31 | polyspin | twingy: Does the observer work on the mac? Could we demo tomorrow using a mac laptop? |
| 14:33.22 | CIA-3 | BRL-CAD: 03brlcad * 10brlcad/src/libfb/if_ogl.c: M-x indent-region and other ws |
| 16:14.33 | CIA-3 | BRL-CAD: 03brlcad * 10brlcad/src/libfb/fb_log.c: indent-region, error if we can't log at compile-time |
| 16:22.27 | CIA-3 | BRL-CAD: 03brlcad * 10brlcad/src/libfb/fb_log.c: bah, preprocessor goo begone |
| 17:01.21 | *** join/#brlcad polyspin (~butler@bz.bzflag.bz) |
| 17:25.05 | CIA-3 | BRL-CAD: 03brlcad * 10brlcad/src/libfb/if_ogl.c: |
| 17:25.05 | CIA-3 | BRL-CAD: do not let any key press or mouse release make the lingering framebuffer |
| 17:25.05 | CIA-3 | BRL-CAD: disappear (!). also do not close stderr since that is what the child uses when |
| 17:25.05 | CIA-3 | BRL-CAD: fb_logging. stub in the same behavior as the X24 interface where it reports the |
| 17:25.05 | CIA-3 | BRL-CAD: rgb value at a particular coordinate. |
| 17:58.04 | CIA-3 | BRL-CAD: 03brlcad * 10brlcad/src/libfb/ (if_4d-old.c if_ogl.c if_ogl_win32.c): BRL frambuffers are just framebuffers now |
| 18:19.20 | CIA-3 | BRL-CAD: 03brlcad * 10brlcad/src/util/ (pl-X.c pl-X10.c): return a value and sleep less before existing |
| 18:22.19 | CIA-3 | BRL-CAD: 03brlcad * 10brlcad/src/libfb/if_ogl.c: throttle the linger idle event loop so that we don't end up calling XCheckWindowEvent as fast as possible. |
| 18:23.17 | CIA-3 | BRL-CAD: 03brlcad * 10brlcad/NEWS: throttle cpu usage on lingering OpenGL frambuffers |
| 18:32.17 | learner | we're so late.. sheesh |
| 18:39.41 | *** join/#brlcad ``Erik (erik@pcp0011474399pcs.chrchv01.md.comcast.net) |
| 18:39.47 | *** join/#brlcad polyspin (~butler@bz.bzflag.bz) |
| 18:48.03 | CIA-3 | BRL-CAD: 03brlcad * 10brlcad/configure.ac: print out the right results for the java test |
| 18:52.45 | learner | we're no less stable than previous releases, in fact we're more stable among other improvements |
| 18:53.13 | learner | that's the whole point of the timely periodic iterations |
| 18:54.48 | learner | there's nothing that prevents more releases as more gets fixed too |
| 18:59.32 | learner | speaking of bugs, I've tracked down the longstanding opengl framebuffer bug.. pure blind luck that I found the cause |
| 19:00.34 | learner | it's the bug that causes ogll to bus error out on the mac |
| 19:01.15 | learner | it basically ends up trying to blit to an ogl context that for "some reason" yet determined isn't ready to blit |
| 19:01.36 | learner | if you turn on the multiple window flag (which was my blind luck), it works just fine |
| 19:02.16 | learner | multiple window forces a direct ogl context whereas non uses indirect |
| 19:02.40 | polyspin | hrm |
| 19:02.59 | learner | er, other way around.. multiple is indirect, non is direct |
| 19:03.05 | learner | direct is not ready for some reason |
| 19:03.41 | learner | debugging that led to a handful of other debugging/fixes too |
| 19:03.52 | learner | like why it eats up 100% cpu/io on linger |
| 19:04.01 | learner | why it won't output debug after raytrace starts |
| 19:04.20 | learner | why mouse/key events killed the window |
| 19:05.18 | learner | only one I didn't try to fix was the old "performance bottoms out" bug .. encountered it on wopr with postage stamp renders |
| 19:06.17 | learner | ``Erik, what's the configure line you use for jdk? |
| 19:08.43 | ``Erik | ../../configure --prefix=/usr/brlcad --with-jdk=/usr/local/jdk1.5.0 |
| 19:08.49 | ``Erik | (freakin' query mode) |
| 19:08.53 | learner | :) |
| 19:08.54 | learner | k |
| 19:09.07 | learner | you know that prefix is the default, right? |
| 19:09.16 | ``Erik | yeah, but I'm a pedantic sob |
| 19:16.49 | learner | bingo |
| 19:17.07 | learner | glXCreateContext was being told to make a direct context |
| 19:17.38 | learner | if it can't/won't do direct, it doesn't fail, it just doesn't give you a direct context so you have to check whether the context given was direct |
| 19:17.45 | learner | that's why it crashed |
| 19:18.07 | learner | the context was never acquired/released since it was assumed to be direct |
| 19:29.33 | CIA-3 | BRL-CAD: 03brlcad * 10brlcad/src/libfb/if_ogl.c: |
| 19:29.33 | CIA-3 | BRL-CAD: glXCreateContext might not be able to make a direct context, so don't presume |
| 19:29.33 | CIA-3 | BRL-CAD: that it will. if it's not, we need to turn on the multiple_windows flag so that |
| 19:29.33 | CIA-3 | BRL-CAD: the context is properly acquired/released. this fixes a bus error crash on the |
| 19:29.33 | CIA-3 | BRL-CAD: macs, as well as other oddities on some other platforms. this is possibly |
| 19:29.33 | CIA-3 | BRL-CAD: related to sf bug 925427 where the entire X display would go black until the |
| 19:29.35 | CIA-3 | BRL-CAD: raytrace completes. |
| 19:30.52 | polyspin | Does anyone know if there is an automake variable for the "current directory under top_srcdir"? |
| 19:30.53 | learner | srcdir? |
| 19:31.14 | learner | builddir is where you're building |
| 19:31.15 | polyspin | I'm trying to set up VPATH based builds for the regression tests |
| 19:31.35 | polyspin | I want to access $top_srcdir/current_directory/filename in the makefiles |
| 19:31.55 | polyspin | For example: sh $(top_srcdir)/regress/solids.sh |
| 19:32.10 | polyspin | Should the directory "regress" ever change, I want it to track |
| 19:32.33 | polyspin | This is invoked via "make test" in the root directory of the build |
| 19:33.41 | learner | srcdir is supposed to be that, I believe, though bugs in older versions sometimes makes that unusable |
| 19:34.40 | learner | easy enough to pinpoint the regress dir if it moves, that's should work regardless of versions |
| 19:34.50 | learner | (using top_srcdir |
| 19:36.41 | learner | polyspin: you did see that I added a top-level test target that calls the regression tests a while back? |
| 19:37.13 | polyspin | yes, I saw that. "make test" |
| 19:37.58 | learner | only made it cd regress and make regress there, thought about making it work like the benchmark but didn't get to it |
| 19:38.56 | learner | benchmark is a good example. finds itself on exec for locating resources and outputs from the callers location so a build tree run should work better |
| 19:41.06 | CIA-3 | BRL-CAD: 03brlcad * 10brlcad/NEWS: prevent indirect OpenGL framebuffers from crashing |
| 20:04.02 | CIA-3 | BRL-CAD: 03lbutler * 10brlcad/regress/ (main.sh nightly.sh Makefile.am lights.sh spdi.sh): |
| 20:04.02 | CIA-3 | BRL-CAD: Updates for running regression test from remote directory when build tree |
| 20:04.02 | CIA-3 | BRL-CAD: and source tree are different. |
| 20:04.02 | CIA-3 | BRL-CAD: Added nightly.sh and main.sh which are the foundation for the nightly runs. |
| 20:04.02 | CIA-3 | BRL-CAD: More to come in these two. |
| 20:24.07 | polyspin | should I expect a lag between when I check a file in and when I see it in SF? |
| 20:24.26 | polyspin | I've committed "regress/main.sh" and didn't get it back on a subsequent checkout |
| 20:27.47 | polyspin | Hmmm. If I get it as me, the files are there. If I get it as anonymous, they are not. |
| 20:28.38 | learner | anonymous lags |
| 20:28.44 | learner | by "up to" several hours |
| 20:29.20 | learner | after all of the sf.net upgrades they should be in exact sync, but they have never been yet |
| 20:29.57 | learner | 4 hours is the current time window, iirc |
| 20:31.03 | learner | ~nslookup ftp.brlcad.org |
| 20:31.13 | learner | ~nslookup 66.111.56.56 |
| 20:31.32 | learner | there we go.. rdns is set |
| 20:33.47 | *** join/#brlcad archivist (~archivist@host217-35-103-47.in-addr.btopenworld.com) |
| 20:39.41 | archivist | will/does brlcad have a "solidworks" or similar front end? |
| 20:40.38 | polyspin | archivist: We can read/write STL. Don't know if that's the sort of thing you meant |
| 20:41.32 | archivist | No i mean a graphical editor ala solidworks solidedge etc |
| 20:42.08 | polyspin | MGED is our graphical geometry editor. |
| 20:42.20 | polyspin | (cough) "mged" |
| 20:44.29 | archivist | hmm, will it build on my alpha, and ram requirement, and does it have 2d output from the 3d modles |
| 20:44.39 | learner | archivist, volume II "Introduction to MGED" on http://brlcad.org is a pretty comprehensive walkthrough/tutorial on how to use mged |
| 20:45.14 | learner | it should build on your alpha, ram requirements are tiny |
| 20:45.15 | archivist | ok will have a look |
| 20:45.28 | polyspin | It's pretty much a 3D package. There is very limited support for "drafting" applications. |
| 20:45.39 | learner | if it doesn't build, that can be fixed asap.. it's been a while since we've had an alpha to play with |
| 20:45.47 | polyspin | It built on the VAX, so the alpha should be little problem ;-) |
| 20:46.45 | learner | archivist, do you mean 2d renderings or something more like drawings from a drafting program? does the prior including line renderings, but not the latter ala blueprinting |
| 20:47.22 | polyspin | archivist: What do you want to use it for? |
| 20:47.58 | archivist | I'm used to doing everything 3d nowadays but manufacturers still expect 2d line drawings |
| 20:48.12 | archivist | mechanical design |
| 20:48.43 | polyspin | We don't have support for dimensioning at the moment. It's been on the wish list for a while. |
| 20:49.08 | learner | have you tried out qcad and pythoncad? they are both 2d editors under development by other groups |
| 20:49.23 | learner | they're nowhere near mature in the 3d arena, but depending on your need, might help |
| 20:49.48 | archivist | no ive got a 3d head i cant go back to 2d |
| 20:50.16 | learner | if you're a dev, you're more than welcome to help add dimensioning :) |
| 20:50.54 | archivist | well i am a programmer also but more used to assembler |
| 20:51.09 | archivist | and cough php |
| 20:51.59 | polyspin | For a moment I thoought you meant "pdp" assembler and I was gonna smile |
| 20:52.54 | archivist | i almost go back that far |
| 20:53.13 | learner | archivist, the closest we have to line drawings are undemensioned outline/edge raytracings .. example: |
| 20:53.14 | learner | http://db.brlcad.org/tmp/moss.png |
| 20:53.18 | learner | http://db.brlcad.org/tmp/moss_edge.png |
| 20:54.02 | learner | and, of course, basic wireframe in the editor, which you can save as plot/postscript and/or convert to images |
| 21:02.49 | archivist | hmm mged perhaps needs a fork so mousers can use it |
| 21:04.30 | archivist | and how about near real time rotates moves |
| 21:05.00 | learner | it does via the shift-grips |
| 21:05.22 | learner | different mouse controls than you are used to, undoubtedly |
| 21:05.38 | learner | but should do all the same |
| 21:06.32 | learner | control-click and move the mouse (with button depressed) |
| 21:07.34 | learner | control+mouse to rotate, shift+mouse to translate, shift+control+mouse to zoom in "real time" |
| 21:08.19 | learner | those are customizable, but well established for brl-cad users |
| 21:08.23 | polyspin | There's a cheat sheet at http://brlcad.net/ShiftGrips.pdf |
| 21:11.42 | archivist | I downloaded the sorce a few months ago the introduction_to_mged.pdf is missing from that(7.0.4) |
| 21:12.04 | learner | "missing"? |
| 21:12.21 | learner | the pdfs aren't included in the source distributions |
| 21:13.09 | archivist | er why? |
| 21:13.23 | learner | did you see how big those pdfs are? |
| 21:13.54 | learner | it would almost double the download to include all the docs |
| 21:14.02 | archivist | no but 100meg for the source didnt frighten me |
| 21:14.04 | learner | some documentation comes in the source dist |
| 21:14.41 | learner | eventually the docs will get remerged into the sources, but not as pdfs |
| 21:14.41 | archivist | thats why i was using the download as it was easier to find |
| 21:14.51 | learner | something maintainable like opendoc or docbook is more likely |
| 21:14.59 | archivist | ok (I hate pdfs anyway) |
| 21:15.22 | learner | feel free to help with that too :) |
| 21:16.19 | archivist | hmm i need to do a trial build on my box forst |
| 21:16.35 | learner | suggest using latest cvs if/when you do |
| 21:16.52 | archivist | ok |
| 21:16.53 | learner | 7.0.4 is "old" |
| 21:25.16 | archivist | page at http://db.brlcad.org/ states there is a link to follow but its a mail addr (it could do with links to available info) |
| 21:26.35 | polyspin | Try http://brlcad.org/ |
| 21:27.00 | polyspin | don't know what db.brlcad.org is there for. |
| 21:29.54 | learner | db's a different machine that used to host brlcad.org |
| 21:30.36 | learner | db was just for that tmp link |
| 21:31.25 | archivist | lots of ppl back up to the top to see what else they can find so may be worth a mod to that page |
| 21:32.04 | learner | what would you have it say? |
| 21:32.48 | learner | also, nobody would get to db unless I give them a link in here :) |
| 21:33.14 | archivist | link back to main site or index the extras on the box |
| 21:39.43 | learner | there, changed. better? |
| 21:41.58 | archivist | yup, next a wiki? |
| 21:42.26 | archivist | i run mediawiki on my box |
| 21:42.58 | learner | there's a new website under works that's vastly improved over http://brlcad.org |
| 21:43.08 | learner | includes a wiki |
| 21:43.19 | learner | db doesn't need one |
| 21:45.46 | archivist | they can be usefull for a community for faq and unofficial info |
| 21:45.57 | learner | new site includes a wiki, forum, news, docs, polls, search, etc |
| 21:47.41 | learner | yes, I have wikis in place for other projects .. only reason it's not up for brl-cad yet is because it's part of a much larger effort to improve the website |
| 21:48.51 | archivist | can be hard to get ppl to work on a wiki sometimes |
| 21:49.42 | learner | one of the systems I've looked into is getting the wiki integrated with docbook |
| 21:50.09 | learner | so whether the edito on-line in wiki format or directly out of cvs using their favorite editor, it's all the same |
| 21:50.20 | learner | and it's revisioned in cvs, so we have a history |
| 21:51.07 | archivist | mediawiki has history |
| 21:51.20 | learner | some wiki's can be quite less than efficient, the resistance is often with due reason |
| 21:52.06 | learner | yes, most wiki's have history, but then it's one more user auth system and it's still only through the web browser |
| 21:52.37 | learner | devs tend to live in editors like emacs/vi working out of cvs, editors like guis, users like websites |
| 21:53.07 | learner | for single-source migration, cvs/svn is the best location since the source has to be there already |
| 21:53.18 | learner | now it's just a matter of integrating a wiki with that backend |
| 21:54.30 | learner | with something like docbook on the backend, I can autogenerate the web/wiki pages, rtf, doc, pdf, etc |
| 21:54.56 | archivist | and a db for an irc bot? |
| 21:55.25 | learner | with an irc bot that groks docbook on a backend, sure |
| 21:55.36 | learner | though bots tend to mainly be useful for factoid snippets |
| 21:55.49 | learner | ~brl-cad |
| 21:55.51 | ibot | hmm... brl-cad is a powerful constructive solid geometry solid modeling system that includes an interactive geometry editor, ray tracing support for rendering and geometric analysis, network distributed framebuffer support, image and signal-processing tools. |
| 21:57.32 | learner | that would make a nice little bot project, though -- a developer bot that integrates with sf.net and gforge facilities, writes docbook on the backend, speaks rss |
| 21:58.03 | learner | CIA would be great for that, but scanline is short-visioned (or tapped for time) |
| 21:58.07 | archivist | we use one in #mysql to through users at either the main docs or the wiki for an indepth description |
| 21:58.51 | archivist | through=thow |
| 21:58.56 | archivist | through=throw |
| 22:00.10 | learner | what kind is it? |
| 22:01.13 | archivist | dont know actually |
| 22:01.13 | learner | there's a bloot and supy in here now |
| 22:01.27 | learner | guu: lart ibot |
| 22:07.53 | archivist | the #mysql bot is pynfo |
| 22:10.05 | archivist | which branch to download? |
| 22:13.13 | learner | HEAD |
| 22:15.23 | archivist | and now ive got to remember that horrid string to get tortoiseCVS to work for me |
| 22:16.28 | learner | cvs -d :pserver:anonymous@cvs.sf.net:/cvsroot/brlcad login |
| 22:16.34 | learner | cvs -d :pserver:anonymous@cvs.sf.net:/cvsroot/brlcad checkout -P brlcad |
| 22:17.02 | archivist | tk |
| 22:17.45 | CIA-3 | BRL-CAD: 03lbutler * 10brlcad/regress/main.sh: Changes to automatically track the release number recorded in configure.ac |
| 22:43.22 | CIA-3 | BRL-CAD: 03lbutler * 10brlcad/regress/ (nightly.sh main.sh): now we track the release number and do the regression test |
| 23:17.28 | *** join/#brlcad Deffie (~Defcon7@deffie.admin.nectarine) |
