IRC log for #brlcad on 20111118

00:50.33 CIA-109 BRL-CAD: 03starseeker * r47540 10/brlcad/trunk/misc/CMake/ (BRLCAD_Util.cmake ThirdParty.cmake ThirdParty_TCL.cmake): Make the third party options a bit more informative
01:08.31 *** join/#brlcad velociostrich (~nicholas@c-24-0-153-224.hsd1.pa.comcast.net)
01:12.30 CIA-109 BRL-CAD: 03starseeker * r47541 10/brlcad/trunk/ (TODO.cmake misc/CMake/BRLCAD_Util.cmake): Get the on/off toggles with auto option displaying their actual state, not just 'AUTO'
01:17.50 CIA-109 BRL-CAD: 03starseeker * r47542 10/brlcad/trunk/misc/CMake/BRLCAD_Util.cmake: AUTO isn't always in the front with new labels
01:23.34 CIA-109 BRL-CAD: 03starseeker * r47543 10/brlcad/trunk/CMakeLists.txt: Be more informative about what's going on with the CPU type, too
01:24.18 starseeker brlcad: hopefully that'll be more informative
01:27.19 CIA-109 BRL-CAD: 03starseeker * r47544 10/brlcad/trunk/TODO.cmake: add a todo item for the alias mechanism
02:49.23 CIA-109 BRL-CAD: 03starseeker * r47545 10/brlcad/trunk/ (CMakeLists.txt misc/CMake/BRLCAD_Util.cmake): Proof-of-concept implementation of a BRLCAD_OPTION macro that supports aliases for an option and appends documentation to a txt file.
02:50.16 starseeker brlcad: before I go too much farther with that I'd appreciate some feedback (in particular, how to write out the documentation)
02:54.04 CIA-109 BRL-CAD: 03starseeker * r47546 10/brlcad/trunk/misc/CMake/BRLCAD_Util.cmake: tweak
02:57.48 CIA-109 BRL-CAD: 03starseeker * r47547 10/brlcad/trunk/CMakeLists.txt: correct global example
03:04.42 *** join/#brlcad abhi2011 (~chatzilla@117.200.86.6)
03:06.09 CIA-109 BRL-CAD: 03starseeker * r47548 10/brlcad/trunk/ (CMakeLists.txt misc/CMake/BRLCAD_Util.cmake): Wait - not handling things quite right with the macro. FORCE shouldn't be needed for that var in the GLOBAL file.
03:14.32 *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol)
03:17.02 CIA-109 BRL-CAD: 03starseeker * r47549 10/brlcad/trunk/ (CMakeLists.txt src/other/CMakeLists.txt): Fix some of the stray variables that should be advanced
03:18.56 CIA-109 BRL-CAD: 03starseeker * r47550 10/brlcad/trunk/CMakeLists.txt: Hmm, that one is easier to change than I thought. Static libs, not the whole thing static
03:36.22 abhi2011 starseeker: there are mismatched if/endif pairs in brlcad_config.h : http://bin.cakephp.org/view/430932163
03:37.07 abhi2011 there are 2 endifs at the end
04:21.11 *** join/#brlcad abhi2011 (~chatzilla@117.200.89.252)
04:35.11 *** join/#brlcad packrat (~packrator@c-98-209-146-133.hsd1.mi.comcast.net)
04:52.56 abhi2011 probably its the cmakecache again, will clear and check
05:20.15 *** join/#brlcad abhi2011_ (~chatzilla@117.200.89.252)
05:39.19 starseeker abhi2011: yeah, that's usually the cache or a stale file, and after that series of changes to the build logic this evening you'll probably have to clear out and reconfigure
09:21.00 *** join/#brlcad abhi2011 (~chatzilla@117.200.89.252)
09:48.34 *** join/#brlcad Technicus (~Technicus@DSLPool-net208-2.wctc.net)
14:25.25 *** join/#brlcad abhi2011 (~chatzilla@117.200.80.155)
14:45.25 *** join/#brlcad n_reed (~molto_cre@BZ.BZFLAG.BZ)
15:23.10 CIA-109 BRL-CAD: 03starseeker * r47551 10/brlcad/trunk/src/libpkg/example/CMakeLists.txt: Add libbu to the link lines...
15:37.18 *** join/#brlcad abhi2011_ (~chatzilla@117.200.83.178)
16:03.08 *** join/#brlcad Elrohir (~kvirc@p5B149B04.dip.t-dialin.net)
16:03.08 abhi2011 starseeker: the other errors are gone now, however 3 link errors are still there for libpkg : http://bin.cakephp.org/view/1734456229
16:25.56 starseeker abhi2011: does commit 47551 fix it?
16:41.10 abhi2011 starseeker: yes thanks :), that did the trick
17:22.45 CIA-109 BRL-CAD: 03starseeker * r47552 10/brlcad/trunk/misc/CMake/ (BRLCAD_Util.cmake ThirdParty.cmake ThirdParty_TCL.cmake): Tweaks and corrections to new CMake labeling.
17:25.00 CIA-109 BRL-CAD: 03n_reed * r47553 10/brlcad/trunk/doc/bison_to_lemon.txt: update on alias substitution
17:37.12 *** join/#brlcad abhi2011_ (~chatzilla@117.200.83.178)
18:01.22 abhi2011 nope spoke too early :P
18:01.26 abhi2011 same error
18:01.52 abhi2011 http://bin.cakephp.org/view/412812688
18:03.40 abhi2011 well libged compiles , so I can proceed with testing the simulate command :)
18:11.48 abhi2011 so when a ray is travelling through a air region , is there some flag set in the struct partition for it (that is returned in the hit call back)
18:12.43 abhi2011 or is only way to check that a ray segment is travelling through air, to check the order of the out and in primitives
18:32.31 CIA-109 BRL-CAD: 03bob1961 * r47554 10/brlcad/trunk/src/ (libtclcad/tclcad_obj.c tclscripts/lib/Ged.tcl): These mods add the functionality to apply snap-to-grid to data polygon editing.
19:19.04 *** join/#brlcad yukonbob (~bch@S01060024a5c9dad4.ok.shawcable.net)
19:19.09 yukonbob hello, #brlcad
19:19.12 brlcad hello
19:19.34 yukonbob hey brlcad :) -- how're things?
19:19.47 yukonbob seeing lots of interesting discussions in newsletters.
19:20.25 brlcad usual, busy, fun, frustrating, interesting, difficult, routine, new, etc
19:20.59 yukonbob what's "new"?
20:22.27 *** join/#brlcad velociostrich (~nicholas@c-24-0-153-224.hsd1.pa.comcast.net)
20:53.36 CIA-109 BRL-CAD: 03n_reed * r47555 10/brlcad/trunk/src/other/perplex/ (parser.y perplex.c perplex.h scanner.re template.c): pass app data to lemon parser; first try at embellishing input
21:03.04 starseeker yukonbob: any luck with making the tcltk 8.6 branch of BRL-CAD/
21:03.10 starseeker ?
21:03.31 yukonbob starseeker: haven't even tried yet.
21:04.31 yukonbob I have a local copy and installation; has been a while since I've played w/ the build, so I'd retry that, confirm it works. The code base will be out-of-date w/ current repo, *but* should be a working piece, which is the important part.
21:04.45 CIA-109 BRL-CAD: 03brlcad * r47556 10/brlcad/trunk/src/rt/view.c: retStatus is unused
21:05.09 yukonbob from there, pull changes from trunk -> 8.6, test, fix, test, fix, commit, lather, rinse, repeat.
21:09.02 CIA-109 BRL-CAD: 03brlcad * r47557 10/brlcad/trunk/src/librtserver/rtserverTest.c: remove/comment out unused code. removes -a option since use_air isn't used.
21:09.25 starseeker if the changes for 8.6 are something we can integrate back into trunk, it might make your life easier
21:10.13 CIA-109 BRL-CAD: 03starseeker * r47558 10/brlcad/trunk/src/libpkg/example/CMakeLists.txt: Ah, right - need special compile flags for MSVC... (ugh)
21:11.48 yukonbob starseeker: the work I've got so far is as it relates to my work w/ NetBSD, modularization, and BRLCAD. I'll see how it works; I know that I'm not at tip of repo, and that I've tried/failed w/ more recent code; I'll take stock though, and get something (with a description of that "something") committed.
21:13.35 starseeker archivist: does 47558 do the trick?
21:13.40 starseeker yukonbob: sounds good!
21:14.02 starseeker when you say "modularization", what are you referring to?
21:15.04 brlcad yukonbob: more nurbs, physics engine, geometry editing goodness, cmake stabilizing, archer revitalizing, ...
21:15.34 brlcad getting 8.6 tested and udpated would be pretty helpful
21:15.56 yukonbob will need to come to know nurbs. and is also still pretty interested in (and may have work for) dsp.
21:16.18 yukonbob dsp still a stable, maintained, first-class citizen in BRLCADville?
21:16.25 brlcad dsp is awesome
21:16.37 brlcad underutilized and underdocumented, but good stuff nonetheless
21:16.50 starseeker for head-scratching values of "maintained" :-P
21:16.58 yukonbob dsp was one of my first forrays into brlcad, and I think my first bug-report ;)
21:17.26 yukonbob anyway -- I'll get my scratch-work massaged and committed.
21:17.42 brlcad lots of ways it can be improved further still, but hard-pressed to find better in-memory solid geometry terrain representation on the scale it supports
21:17.50 yukonbob was not a fan of 8.5, but -is- a fan of 8.6, so happy to have that a side-effect of his work.
21:18.01 starseeker sweet
21:18.06 CIA-109 BRL-CAD: 03n_reed * r47559 10/brlcad/trunk/src/other/perplex/ (scanner.re template.c): use re2c syntax for setting conditions
21:18.12 starseeker can build 8.6 with CMake - took some pains to be sure I could, actually
21:18.23 brlcad well, as you noted.. you do still have commit ability :)
21:18.39 starseeker needs to get a TIP written up - would REALLY love to have the CMake build logic just "there" in 8.6 final
21:21.28 starseeker yukonbob: I need to set up a virtual machine with NetBSD so I can try a build - any recommendations on pre-cooked VirtualBox images?
21:23.21 yukonbob starseeker: no...
21:23.40 yukonbob 1s.
21:24.14 yukonbob starseeker: virtualbox == that sun vm?
21:24.34 yukonbob sees "yes".
21:25.14 CIA-109 BRL-CAD: 03brlcad * r47560 10/brlcad/trunk/ (9 files in 9 dirs):
21:25.15 CIA-109 BRL-CAD: rename BRLCAD-CPU_TYPE and CMAKE_CPU_TYPE to BRLCAD-WORD_SIZE and
21:25.15 CIA-109 BRL-CAD: CMAKE_WORD_SIZE respectively so as not to imply chip type or architecture. CPU
21:25.15 CIA-109 BRL-CAD: may be multimode as could the compiler, so it's not a good moniker. also
21:25.15 CIA-109 BRL-CAD: consistently use either ##BIT or ##-bit styles when referring to the size in
21:25.15 CIA-109 BRL-CAD: text.
21:26.02 starseeker brlcad: awesome, thanks!
21:26.08 brlcad starseeker: so I'm leaning more towards renaming all the BRLCAD- cases to BRLCAD_ grouping all vars together, for improved usability
21:26.35 starseeker nods - sounds good. I have no particularly strong feelings on that, so whatever you feel is best
21:26.51 brlcad it'd be nice to have subgroupings, but not at that usability crux since we'd still want to document what it really is
21:27.02 brlcad then the aliases can just be shorthand and typo preventions
21:27.14 starseeker nods
21:27.15 brlcad not every _- permutation :)
21:29.32 CIA-109 BRL-CAD: 03starseeker * r47561 10/brlcad/trunk/src/libpkg/example/ (client.c server.c): Hmm... trying to send a message back to the client from the server. Doing something wrong...
21:31.43 starseeker I suppose I should have known it wouldn't be that simple...
21:36.58 yukonbob starseeker: re: nbsd -- nothing I see off top of head.
21:37.08 starseeker no problem
21:37.54 yukonbob brl-cad still maintains own build server?
21:38.19 starseeker um... you mean brlcad's setup?
21:38.42 yukonbob not sure -- I've had an account on what I thought was a brl-cad machine in past.
21:39.33 starseeker yeah, that's probably it - yeah, he's still got it up
21:39.47 yukonbob only wondering for same reasons you're wondering about nbsd -- If there was some other box that's linux or fbsd or $whatever, just another opportunity to test build/run, shake out bugs, etc., etc.
21:40.08 starseeker hmm... well, here's a bunch of images... no netbsd though http://virtualboxes.org/images/
21:40.38 starseeker come to think of it, that might be the opensolaris image I grabbed
21:41.05 starseeker hopes he remember how to get into that - it was set up for building, and it's a long shot whether that could be done again
21:41.06 yukonbob ah -- my account is still good.
21:41.27 yukonbob oh wow. timewarp.
21:41.29 yukonbob :)
21:41.34 starseeker hehe
21:41.44 yukonbob s'all good in the 'hood. ;)
21:58.40 brlcad starseeker: fyi, I did a quick check and confirmed that the server should be able to send back to the client bidirectionally
21:58.48 brlcad your comment yesterday didn't sound right..
21:59.34 brlcad fbserv sends back error notifications to the client being one relatively obvious example
22:00.51 brlcad that's a pretty nifty virtualization OS list .. would be great for automated compilation testing for someone to set up
22:01.21 brlcad run vm with each OS image, add brl-cad sources, compile, report to dashboard
22:01.45 brlcad shudders at the awesome power of that
22:05.16 starseeker was discussing that with some of the GSoC guys
22:05.49 brlcad interesting, I was too
22:05.58 starseeker Jenkins is apparently one of the relevant tools for managing something like that...
22:06.16 brlcad more for gci purposes, though -- provide an image completely preconfigured, go!
22:06.18 yukonbob a build dashboard?
22:06.27 starseeker nods
22:06.39 yukonbob cbuild
22:07.08 yukonbob err. ctest.
22:07.33 starseeker oh - we don't have ctest integrated yet
22:07.38 yukonbob http://www.vtk.org/Wiki/CMake_Testing_With_CTest <--- w/ cdash, or dart...
22:07.56 yukonbob has never actually used these, but is aware.
22:08.49 yukonbob out.
22:08.54 yukonbob chat later, cadheads.
22:08.57 starseeker later!
22:09.04 brlcad jenkins is one of the top ones
22:09.17 starseeker hmm https://wiki.jenkins-ci.org/display/JENKINS/VirtualBox+Plugin
22:09.18 brlcad hudson is a fork that seems more promising
22:09.38 brlcad cruisecontrol is the old steadfast best
22:09.54 starseeker um... thought jenkins was a fork of hudson?
22:09.57 brlcad buildbot is one of the newercomers with some neat features
22:10.45 brlcad sorry, that's right
22:11.48 brlcad "Both the Jenkins and Hudson projects appear to consider the other to be a fork."
22:12.04 brlcad more Oracle stupidity
22:13.09 CIA-109 BRL-CAD: 03bob1961 * r47562 10/brlcad/trunk/ (3 files in 3 dirs): Added the mechanism for sketching out elliptical shaped polygons.
22:13.11 brlcad I think any one of those four choices would be perfect for our needs, it's mostly just needing someone to champion setting it up proper
22:13.36 starseeker brlcad: OK, I'll take a look at fbserv
22:14.26 brlcad suggest just trying to figure it out with your own example still :)
22:14.59 brlcad master of your own domain and whatnot
22:15.14 brlcad plus it then requires really understanding what state both client and server are in
22:15.43 starseeker doen't want to understand pkg... just wants it to work and go away...
22:15.44 brlcad put printing statements before and after all of your send() and recv() statements in both client and server, so you can see who is waiting on what
22:16.53 brlcad don't think that's a productive mindset
22:17.00 brlcad it's not a means to an end, we use it already
22:17.24 brlcad so you either want to learn it so you can/could use it or debug it or extend it, or you leave it to someone else...
22:18.22 starseeker fair enough
22:18.32 brlcad that would even be the case if we picked up some 3rd party package .. it's all fine and dandy when it works, but then it's a brick wall black box when anything goes wrong (which it eventually always does)
22:19.15 velociostrich Activity on this channel? Unheard of!
22:19.23 velociostrich goes back into hiding
22:19.44 brlcad velociostrich: it's usually like this most days...
22:19.47 CIA-109 BRL-CAD: 03bob1961 * r47563 10/brlcad/trunk/src/libtclcad/tclcad_obj.c: Removed junk that was mistakenly included with the previous commit.
22:19.49 brlcad save maybe for last week
22:20.08 velociostrich I wouldn't know, I'm not in here often, but when I have been I rarely noticed activity
22:20.35 velociostrich Besides the 'e' command, is there any way to get a nicer preview of work in mged ala other cad packages, e.g., opencascade based alternatives?
22:20.41 brlcad *shrug*
22:20.41 velociostrich iirc it was 'e', anyway
22:21.02 brlcad velociostrich: you can render via raytracing (File menu)
22:21.29 velociostrich Yes, I do realize that
22:21.32 brlcad for simple models, you can use the E or ev commands to get a polygonal wireframe
22:21.42 velociostrich how do e and ev differ?
22:21.58 brlcad e just draws the fundamental wireframe
22:22.05 brlcad ev evaluates a polygonal wireframe
22:22.18 brlcad E also evaluates, but with a slightly different algorithm
22:22.43 brlcad you probably are looking for shaded display support, which is available for certain types of models but not all so it's disabled by default
22:23.01 brlcad future release will enable shaded display support for all geometry types
22:23.19 velociostrich strange, it seems that ev is having some depth buffer related issues (Does brlcad use a depth buffer?)
22:23.30 velociostrich things on top of things that shouldn't be
22:23.34 brlcad the depth buffer can be turned on/off under Misc
22:23.59 brlcad along with depth lighting, and a few other modes
22:25.31 velociostrich aha
22:25.48 velociostrich although it seems that the depth buffer doesn't work unless Z clipping and the buffer are enabled
22:25.57 brlcad starseeker: maybe you can convince bob to make that nsegs=30 dynamically adjust
22:26.09 brlcad velociostrich: yep, iirc that sounds right
22:26.45 velociostrich strange
22:28.16 velociostrich Might I ask if there is any particular reason for the use of Tk and not a (as far as I know, ignorant as I am) more modern toolkit like Qt/Gtk for Archer (which afaik is much newer than mged)?
22:30.13 CIA-109 BRL-CAD: 03brlcad * r47564 10/brlcad/trunk/ (25 files in 20 dirs): rename all of the BRLCAD- cmake variables to BRLCAD_ so we can consistently only use underscores everywhere. should help improve simplicity of docs and use.
22:30.18 brlcad velociostrich: archer is still mged, just the project name -- it's a refactoring of mged's existing code from pure tcl to itcl
22:30.33 velociostrich oooh
22:30.42 brlcad main reason is that tcl/tk was adopted long before any of the modern toolkit's existed
22:30.53 velociostrich Yeah, I assumed as much
22:30.59 brlcad qt is on our horizon for archer's eventual replacement, couple years out on our planning schedule
22:31.20 velociostrich Yeah, from and end-users' standpoint, Tk makes Archer/mged look dated unfortunately
22:31.35 velociostrich I've noticed that most people are quick to judge a book by its cover
22:31.37 brlcad still need to refactor more of mged's internal engine code down into lower library layers so the new GUI doesn't have to reinvent the wheel
22:32.15 brlcad in all fairness, Tk can be made to look like most modern GUIs... it just takes a bit of effort to get off the defaults
22:32.25 velociostrich perhaps
22:32.32 brlcad we just use the defaults because there are more pressing engine matters to develop
22:32.40 velociostrich I can understand that
22:33.02 velociostrich Especially considering the KLOCs you guys have on your hands
22:33.04 brlcad if you'd like to help in that regard (or library refactoring or whatever really), have at it
22:33.14 brlcad lemme know how I can help you get started ;)
22:33.16 starseeker archer is better that way than mged - we're mostly using the new tile/ttk widgets there
22:33.19 velociostrich If only I had time, I would :/
22:33.45 brlcad you just need a mutual itch to scratch ;)
22:33.58 brlcad a project that is interesting and tangible perhaps
22:34.03 velociostrich And time
22:34.33 brlcad time is merely a matter of priorities
22:34.43 brlcad if you had that itch, some need, you'd scratch it
22:34.53 brlcad because itchy things get scratched
22:34.54 brlcad :)
22:35.40 velociostrich You're damn right, now stop enticing me before my priorities go out of wack for a few weeks while I do that ;)
22:35.54 CIA-109 BRL-CAD: 03tbrowder2 * r47565 10/brlcad/trunk/INSTALL.cmake: give a tad bit of help for a novice cmake builder; indent the configuration args to cmake; eliminate the old autotools stuff to eliminate confusion; show as a work in progress
22:37.12 velociostrich Also, out of curiosity, do you know if the ARL still uses BRL-CAD, or have they abandoned it in favor of Autocad and friends?
22:38.15 brlcad velociostrich: yep, still heavily used
22:38.50 brlcad it's pretty much core business to the DoD for the foreseeable decade
22:39.00 velociostrich interesting
22:39.03 velociostrich likely with certain extras that the public won't ever see, though
22:39.12 velociostrich namely extra ballistics bits
22:39.38 velociostrich Which I assume was the point of it being solid geometry based and having the raytracing library and all that
22:39.39 brlcad there are other CAD packages that have to interoperate with more these days than before, so strong focus on STEP import, for example, so we can bring in external CAD models without changing the representation format
22:39.57 brlcad those extra ballistic bits aren't part of brl-cad, never have been
22:40.24 velociostrich Yes, but would have used the raytracing library, correct?
22:40.29 brlcad those bits call into our core libraries, which are built for exactly that purpose (and are still pretty much the best at it barnone)
22:40.39 velociostrich Yes, that's what I meant
22:41.56 velociostrich Y'know, that makes me wonder if perhaps brlcad's raytracing library might be well suited to the games industry; from what I understand AI makes heavy use of raytracing for pathfinding and such, and they claim it's quite expensive to compute
22:42.16 velociostrich which I think is typically done using whatever partitioning scheme they're using
22:42.29 velociostrich which is less than ideal
22:42.30 brlcad it is very well suited for that purpose
22:42.57 brlcad used within DoD by a couple groups playing wargames in that way (sorta)
22:43.00 CIA-109 BRL-CAD: 03n_reed * r47566 10/brlcad/trunk/src/other/perplex/ (perplex.h scanner.re template.c): changed token text allocation scheme to be leak resistant
22:48.17 brlcad starseeker: /home/sean/brlcad/src/libpkg/example/server.c:117:10: error: variable ?bytes? set but not used [-Werror=unused-but-set-variable]
22:48.18 CIA-109 BRL-CAD: 03bob1961 * r47567 10/brlcad/trunk/src/libtclcad/tclcad_obj.c: Dynamically calculate the number of segments used to approximate circles and ellipses using the window size.
22:48.53 starseeker brlcad: hang on - will be doing a commit in a sec anyhow
22:54.03 CIA-109 BRL-CAD: 03starseeker * r47568 10/brlcad/trunk/src/libpkg/example/ (client.c server.c): Back to basics - don't worry about the file, just dirt simple back-and-forth.
22:54.14 starseeker brlcad: convinced ;-)
22:56.00 *** part/#brlcad n_reed (~molto_cre@BZ.BZFLAG.BZ)
23:12.10 starseeker ok, I see why it was doing what it was doing, I think...
23:12.41 starseeker brlcad: thanks for the print statement suggestion, that helped
23:12.59 starseeker didn't realize it didn't break out of the while loop after completing the file transfer
23:30.43 *** join/#brlcad velociostrich (~nicholas@c-24-0-153-224.hsd1.pa.comcast.net)
23:35.38 *** join/#brlcad Technicus (~Technicus@DSLPool-net208-2.wctc.net)
23:43.08 *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.