IRC log for #brlcad on 20090715

00:29.24 *** join/#brlcad CIA-30 (n=CIA@208.69.182.149)
00:37.23 CIA-30 BRL-CAD: 03starseeker * r35120 10/brlcad/trunk/ (include/opennurbs_cleanup.h src/librt/opennurbs_cleanup.cpp): More CurveTree building tweaks
00:42.17 CIA-30 BRL-CAD: 03starseeker * r35121 10/brlcad/trunk/src/librt/opennurbs_cleanup.cpp: Er - don't need call to isLinear in the if condition that exists for the Linear cases...
00:54.49 ``Erik awesome, an accelerated emacs course: xslt.2:${PORTSDIR}/textproc/libxslt
00:54.52 ``Erik grrr
00:55.36 ``Erik http://edward.oconnor.cx/2009/07/learn-emacs-in-ten-years
00:55.37 ``Erik http://edward.oconnor.cx/2009/07/learn-emacs-in-ten-years
00:55.54 ``Erik fuckin' A, my X11.app isn't talking to the system clipboard quite right anymore ://
01:47.04 CIA-30 BRL-CAD: 03n_reed * r35122 10/brlcad/trunk/src/ (libdm/dm-rtgl.c mged/Makefile.am): fixed mged dm-rtgl bus error on exit
02:08.54 starseeker hmm - still some difference in how the surface tree is being built
02:08.57 starseeker arrrrgh
02:25.36 CIA-30 BRL-CAD: 03starseeker * r35123 10/brlcad/trunk/ (include/opennurbs_cleanup.h src/librt/opennurbs_cleanup.cpp): (log message trimmed)
02:25.39 CIA-30 BRL-CAD: Ah. Need to subdivide even flat nurbs surfaces in order to assign small subsets
02:25.41 CIA-30 BRL-CAD: of the trimming curves to small areas of the nurbs surface for evaluation
02:25.43 CIA-30 BRL-CAD: purposes. I wonder, would it be 'cleaner', in cases like this, to do some sort
02:25.47 CIA-30 BRL-CAD: of test of points against the curve tree? i.e. for a uv point, if inside the uv
02:25.49 CIA-30 BRL-CAD: bounds of a node, find subnodes that it is inside of, repeat until leaf node(s)
02:25.51 CIA-30 BRL-CAD: are found, then work on the leaf nodes? Probably more expensive for a trimming
04:42.17 *** join/#brlcad Patmcc19 (n=chatzill@71-223-60-113.phnx.qwest.net)
05:34.37 *** join/#brlcad stevegt_1 (n=stevegt@c-24-130-122-25.hsd1.ca.comcast.net)
07:45.58 *** join/#brlcad _clock_ (n=_sushi_@77-58-151-159.dclient.hispeed.ch)
08:15.21 Ralith woo
08:15.27 Ralith QDialog says hello :D
08:15.36 Ralith still have to work out why ogre seems to be misaligned though
08:15.51 Ralith er, wait
08:16.01 Ralith that's.... funny
08:18.05 CIA-30 BRL-CAD: 03ralith * r35124 10/rt^3/trunk/src/g3d/ (6 files): "Naive" approach to the Ogre/Qt conflict: Added the test widget as a child of the OgreGLWidget. Test widget ends up as a separate window.
08:22.50 Ralith my mistake:
08:22.51 Ralith 01:22:59 < HuHa> Ralith: the docs explicity say: "QDialog... uses the parent widget slightly differently from other classes in Qt. A dialog is always a top-level widget, ..."
08:25.09 CIA-30 BRL-CAD: 03ralith * r35125 10/rt^3/trunk/src/g3d/ogretest.cxx: Tested with a QPushButton which, unlike QDialog, is *not* documented as "always a top-level widget," and which allegedly should not need ->show() to become visible. No luck yet.
08:30.36 CIA-30 BRL-CAD: 03ralith * r35126 10/rt^3/trunk/src/g3d/ogretest.cxx:
08:30.38 CIA-30 BRL-CAD: Working pushbutton overlayed! Something about manually setting the size of the OgreGLWidget with ->resize(1024, 768); threw it off; my
08:30.41 CIA-30 BRL-CAD: guess is that my (tiling) wm's enforced resize produced a layout that resulted in the widget being offscreen. It may be worth testing
08:30.44 CIA-30 BRL-CAD: previously failed options with said resize call removed.
08:30.46 Ralith brlcad: we have success!
08:30.48 Ralith of a sort.
08:31.01 Ralith should work well enough to move on, certainly
08:34.42 Ralith I could really do with some testing from others, btw
08:34.53 Ralith I want to find out whether this weird ogre misalignment is a side effect of my tiling wm or what
08:46.18 *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6)
08:47.56 Ralith logs with excessively nested bullet points
08:48.00 CIA-30 BRL-CAD: 03Ralith 07http://brlcad.org * r1571 10/wiki/User:Ralith: Logs for 2008-07-12 and 2008-07-14. We now have usable Qt widget overlaying!
08:49.33 Ralith brlcad: let me know if you want me to continue to play with alternative approaches, especially considering the implications of discoveries relating to QWidget::show(), or whether I should just dive ahead into building a useful GUI.
08:52.27 brlcad insane timing, love it but have to read/respond later today :)
08:52.43 Ralith 'kay :)
08:52.45 Ralith night
09:02.40 CIA-30 BRL-CAD: 03Ebautu 07http://brlcad.org * r1572 10/wiki/More_Changelog: July 13-14
09:43.45 *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6)
09:45.19 *** join/#brlcad LarsG (n=lars@137.132.207.11)
09:45.22 *** part/#brlcad LarsG (n=lars@137.132.207.11)
10:11.52 *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6)
10:13.36 d-lo Morning all!
10:14.12 d-lo Ralith: Grats :)
11:05.53 CIA-30 BRL-CAD: 03ebautu * r35127 10/web/trunk/htdocs/more/sites/all/modules/brlcad/ (brlcad.module scripts/7.14/metadata.txt): Improved php-mged interface. Added feature to retrieve file list.
11:20.07 *** join/#brlcad mafm (n=mafm@83.42.152.74)
11:38.50 CIA-30 BRL-CAD: 03jdoliner * r35128 10/brlcad/trunk/ (include/bn.h src/libbn/multipoly.c): added some nuts and bolts bivariate polynomial manipulation functiuons
11:42.24 CIA-30 BRL-CAD: 03davidloman * r35129 10/rt^3/trunk/ (19 files in 17 dirs): CMake build fixes, cleanup, WS removal and style consistancy changes.
11:45.00 *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-181.sbndin.btas.verizon.net)
11:48.49 starseeker Ralith: so this is not using the shared OpenGL context, correct?
12:11.27 *** join/#brlcad stevegt_1 (n=stevegt@c-24-130-122-25.hsd1.ca.comcast.net) [NETSPLIT VICTIM]
12:11.27 *** join/#brlcad Axman6 (n=Axman6@61.69.25.81) [NETSPLIT VICTIM]
12:11.32 CIA-30 BRL-CAD: 03d_rossberg * r35130 10/brlcad/trunk/src/libged/ (dg_obj.c editit.c edmater.c rt.c rtcheck.c):
12:11.35 CIA-30 BRL-CAD: MSVC does not provide all POSIX headers
12:11.37 CIA-30 BRL-CAD: handled this with the HAVE_~_H macros
12:20.59 *** join/#brlcad jdoliner (n=jdoliner@68.51.75.169)
12:37.15 brlcad g'morning jdoliner
12:37.44 brlcad Ralith: yeah, agree -- we should get a couple people to test things out
12:37.50 brlcad screenshot?
12:38.03 d-lo hah! Pics or it never happened :)
12:38.34 CIA-30 BRL-CAD: 03davidloman * r35131 10/rt^3/trunk/ (78 files in 14 dirs): More WS, header and Style consistancy fixes
12:46.28 brlcad d-lo: curious commit there..
12:46.42 d-lo How so?
12:47.49 brlcad aside from the subtle change to the standard header, dropping the file identifier, we moved away from using rcs variables a long while back because they cause hell on branching and merging (and is information contained within the vcs anyways)
12:48.54 d-lo So you recommend not using svn;keywords at all?
12:50.36 brlcad within the file, yeah, with a few exceptions (like build system files that use revision stamping)
12:51.29 d-lo *confused* How else can the rcs variables be used, if not 'with in the file' ?
12:51.49 brlcad the two main issues are the technical one with it causing more grief with conflict than time-savings gain conveying information; and a social impact of tending towards territorial code if files include authorship
12:52.24 brlcad within source files, there are other files they work well in where the conflict potential is very low
12:53.22 d-lo ah, okay. 'Territorial code' ? is that really a problem within the brlcad dev community?
12:53.59 brlcad that's a problem with most devs, left unchecked, but even more so for 'new' devs, which is the main reason
12:54.35 ``Erik ralith: pleased to be leaving breadcrumb trail in case anyone else picks up your "good enough" to make it "great" :D
12:54.55 d-lo Okay. I honestly don't care what is in the headers, so long as its consistant file to file.
12:55.01 brlcad new developers often won't touch a file that has an author associated with it, kinda like a dog marking his territory for fear of "changing their code" instead of treating it like project code
12:55.47 d-lo Except for the F I L E N A M E at the top. that is a pita.
12:55.56 brlcad there's a great interview with brian fitzpatrick (subversion developer, google dev, among other notabilities) and one of his cohorts on this phenomenon
12:56.12 brlcad pita howso?
12:56.19 brlcad should be autogenerated
12:56.29 brlcad that's what the template script is for
12:56.49 ``Erik or some vim or emacs fu if you want to copy and rework a file
12:56.58 ``Erik tends to do it by hand, though :/
12:57.15 d-lo going through and retro fitting a bunch of various stuff from a bunch of different devs.
12:58.20 brlcad there's actualy an intent legal motivation (however strong) to uniquely identify each file before the project, in order to imply it is an individual work that is part of a larger collective work
12:58.57 brlcad write-up from the FSF and groklaw regarding that issue, iirc
12:59.34 brlcad there was a reason for each item in the template ;)
13:00.02 d-lo And thats all well and good. Still a pita ;)
13:00.23 brlcad the header.sh and footer.sh scripts can be run individually on single or groups of files to correct them
13:01.29 brlcad 'find' is a great command to master if you haven't already ;)
13:01.56 d-lo yes, but thats not the pita i am talking about
13:02.07 brlcad and "perl -pi -e 'EXPRESSION' ..." or variants using sed/awk
13:03.33 brlcad should be a trivial mod to the header script to detect an existing header and 'make it conform' like footer.sh does
13:04.23 brlcad forget what the current behavior is, whether to detect and abort, or detect and prepend
13:05.09 brlcad think it's a detect and abort if it finds the legal clause, otherwise prepends and treats as a comment header
13:06.08 d-lo kk, thanks for the heads up on the footer/header scripts.
13:08.37 brlcad not at concerned about that as the rcs vars, more because of their long term impact
13:08.49 brlcad because they're pretty easy to fix
13:12.22 d-lo the '@file' part of the header... is that a DocBook thing?
13:12.54 *** join/#brlcad docelic (n=docelic@78.134.202.66)
13:14.10 d-lo err, Doxygen i mean.
13:16.42 d-lo brlcad: Network guys are looking at the Switch UPS. Anything you need me to tell them?
13:16.48 ``Erik brlcad: have you looked at the swiss cheese vrml issue?
13:17.11 ``Erik switch ups hasn't been an issue for us so far
13:17.28 ``Erik all our machines drop before the switch *shrug*
13:17.55 d-lo Well, 'They said 'Sean said''... maybe you can talk to them :)
13:20.21 *** join/#brlcad _sushi_ (n=_sushi_@77-58-151-159.dclient.hispeed.ch)
13:35.04 CIA-30 BRL-CAD: 03johnranderson * r35132 10/jbrlcad/trunk/src/org/brlcad/numerics/BoundingBox.java: Added check for NaN in intersect method
13:35.50 brlcad ``Erik: the switch for the networking shut off really quickly on the recent outtages
13:37.07 brlcad maybe some anamoly, maybe surge, don't know -- either way it shut off faster than any of the hosts and used to stay up for nearly an hour
13:37.38 brlcad and no, i've not looked at the vrml issue, but suspect it's needing to tweak tolerancing with those really small holes
13:38.26 CIA-30 BRL-CAD: 03davidloman * r35133 10/rt^3/trunk/ (164 files in 33 dirs): Standardizing headers.
13:51.19 *** join/#brlcad mafm_ (n=mafm@83.42.152.74)
13:51.28 *** join/#brlcad samrose (n=samrose@24.11.214.181)
13:59.21 CIA-30 BRL-CAD: 03starseeker * r35134 10/brlcad/trunk/src/librt/opennurbs_cleanup.cpp: OK, now that we're breaking down flat surfaces too, just call GetBVChildren directly.
14:02.47 CIA-30 BRL-CAD: 03davidloman * r35135 10/rt^3/trunk/src/GE/ (14 files in 3 dirs): Missed some headers. Fixed some mangling.
14:04.42 CIA-30 BRL-CAD: 03davidloman * r35136 10/rt^3/trunk/ (5 files in 2 dirs): refactored CommunicationsManager to AbstractPortalManager.
14:21.53 CIA-30 BRL-CAD: 03davidloman * r35137 10/rt^3/trunk/ (4 files in 2 dirs): Stubs for NetSockPortalManager (subclass of AbstractPortalManager).
14:39.41 ``Erik he
14:41.11 brlcad she
14:43.31 d-lo we
14:47.31 archivist they
14:47.36 ``Erik woops, heh, d'no what I was typing that accidently went here O.o
14:47.41 ``Erik (fuckers)
14:48.26 d-lo play
14:49.53 CIA-30 BRL-CAD: 03johnranderson * r35138 10/jbrlcad/trunk/ (4 files in 3 dirs):
14:49.56 CIA-30 BRL-CAD: Minor mod to RegionInfo class to handle null "region_id" attribute.
14:49.58 CIA-30 BRL-CAD: Added a simple test for RegionInfo.
15:39.14 *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-181.sbndin.btas.verizon.net)
15:40.09 *** join/#brlcad hippieindamakin8 (n=hippiein@210.212.55.3)
15:46.54 *** join/#brlcad b0ef (n=b0ef@084202026157.customer.alfanett.no)
17:33.22 *** join/#brlcad louipc (n=louipc@archlinux/trusteduser/louipc)
17:45.27 CIA-30 BRL-CAD: 03davidloman * r35139 10/rt^3/trunk/ (11 files in 4 dirs): Basic implementation of NetSockPortal and NetSockPortalManager.
18:19.02 CIA-30 BRL-CAD: 03starseeker * r35140 10/brlcad/trunk/ (include/opennurbs_cleanup.h src/librt/opennurbs_cleanup.cpp): Alright. Start trying to PRECISELY duplicate the curvetree building of the old code, just to have a working baseline.
18:37.51 *** join/#brlcad samrose (n=samrose@adsl-99-147-180-206.dsl.lgtpmi.sbcglobal.net)
18:47.32 *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-181.sbndin.btas.verizon.net)
19:05.21 CIA-30 BRL-CAD: 03irpguardian * r35141 10/brlcad/trunk/src/proc-db/human.c:
19:05.23 CIA-30 BRL-CAD: Added 2 new functions, boundingBox and boundingRectangle.
19:05.25 CIA-30 BRL-CAD: Bounding box makes rectangles with 1 value for width/depth, and bounding rectangle makes rectangles with 2 values, one for depth, one for width.
19:05.28 CIA-30 BRL-CAD: They currently make boxes using -b, but are not rotated or positioned correctly yet.
19:24.55 CIA-30 BRL-CAD: 03starseeker * r35142 10/brlcad/trunk/src/librt/opennurbs_cleanup.cpp: Ah, blast it - + where a - should be.
19:27.41 CIA-30 BRL-CAD: 03starseeker * r35143 10/brlcad/trunk/src/librt/opennurbs_cleanup.cpp: Few more tweaks to the curve tree build - still have some definite problems somewhere but getting closer
19:53.05 CIA-30 BRL-CAD: 03starseeker * r35144 10/brlcad/trunk/src/librt/opennurbs_cleanup.cpp: Whoops, - or, not and for breakdown conditions
20:05.06 CIA-30 BRL-CAD: 03starseeker * r35145 10/brlcad/trunk/ (include/opennurbs_cleanup.h src/librt/opennurbs_cleanup.cpp): Put the bounding box adjustment back in the Children call.
20:28.30 CIA-30 BRL-CAD: 03starseeker * r35146 10/brlcad/trunk/ (include/opennurbs_cleanup.h src/librt/opennurbs_cleanup.cpp):
20:28.33 CIA-30 BRL-CAD: Hmm - looks like the memory demands of this approach are a bit too high - bump
20:28.35 CIA-30 BRL-CAD: up the BREP_SURF_SUB_FACTOR a bit, but will need to think about another approach
20:28.37 CIA-30 BRL-CAD: to passing values - I'm guessing the big ON_3dPoint and ON_3dVector arrays in
20:28.39 CIA-30 BRL-CAD: each node are to blame.
20:36.16 *** part/#brlcad brlcad (n=sean@bz.bzflag.bz)
20:40.03 ``Erik O.O
20:54.46 *** join/#brlcad stevegt_ (n=stevegt@cislunar.TerraLuna.Org)
21:11.32 *** join/#brlcad Patmcc19 (n=chatzill@71-223-60-113.phnx.qwest.net)
21:26.53 *** join/#brlcad brlcad (n=sean@bz.bzflag.bz)
21:45.10 *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-181.sbndin.btas.verizon.net)
21:45.44 Ralith starseeker: correct; it's being drawn 'on top' of the context, although still in the same top level window. Shared context may still be perfectly feasible though; it looks like a few other calls which I only just now twiddled accidentally and/or on a hunch were unexpected dealbreakers.
21:53.16 Ralith brlcad: probably easier to build it yourself than to wait for a screenshot, but okay
22:13.05 *** join/#brlcad Patmcc19 (n=chatzill@71-223-60-113.phnx.qwest.net)
22:28.20 ``Erik hm
22:40.40 Ralith I could *really* do with someone else testing this, btw
22:41.19 Ralith brlcad: http://imgur.com/TSMHY.png
22:41.37 Ralith don't mind the outlined black box in the middle, that's from my wm
22:52.10 brlcad so is the entire black area an opengl context?
22:53.02 brlcad ah, I see your reply to starseeker .. cool
22:53.16 brlcad then that's progress!
22:53.44 brlcad got to speak about brl-cad briefly during the FLOSS weekly podcast interview
22:53.55 ``Erik neat
22:54.19 brlcad not one of my best talks, by far
22:55.08 brlcad kinda frustrating, ill-timed to prepare (had all of 30 minutes or so including mic checking)
22:55.29 brlcad but nice plugs in there for brl-cad and bzflag, just didn't get to say much of the cool stuff
22:55.35 ``Erik erm, why didn't they give ya more notice and shit?
22:55.52 brlcad there was more notice, *I* was ill-prepared
22:55.57 ``Erik last minute replacement type deal?
22:55.58 ``Erik ahhh
22:56.00 ``Erik hah
22:56.02 ``Erik you suck :D
22:56.10 brlcad timing was conflicting to prepare
22:56.45 ``Erik but now it's done and you can busy yourself migrating machines, right? :D *duck*
22:56.49 brlcad otherwise, it was still just about a week's notice
22:57.37 ``Erik is very tempted to do some firewall fu, tweak the grant table and see if he can get the old machines websites using the new machines rdbms
22:57.40 brlcad now I need food and a nap .. all i've had today are two egg rolls and a glass of scotch
22:57.48 brlcad and 3 hours sleep
22:58.03 ``Erik or mebbe sneak an ssh tunnel across
22:59.16 ``Erik *shrug*
23:00.33 Ralith brlcad: the entire black area is indeed the GL context. I'm not sure why the dark grey area (Ogre's configured background color) is only partial; I'd like it if someone with a normal WM could build and test it.
23:01.10 ``Erik how normal is a normal wm?
23:02.01 ``Erik if you walk me through it and promise you'll still respect me in the morning, I can try it on my macbook
23:04.57 Ralith heh
23:05.00 Ralith I'm not sure it'd work on OSX at all
23:05.10 brlcad (Ralith: tell him you will, use him, get your satisfaction, then leave him to the curb)
23:05.10 Ralith let alone how to install Ogre and Qt there
23:05.16 Ralith ^^
23:05.26 brlcad it should work just fine there
23:05.26 Ralith worth a try
23:05.40 brlcad fully supported OS for both those
23:05.42 Ralith I'd hope so, but I'm in no position to walk someone through
23:05.47 Ralith goes googling
23:05.59 brlcad sure you are, "install this, install that, compile this .. "
23:06.00 ``Erik !!#!~@
23:06.02 Ralith ``Erik: I assume you've got all the usual dev tools?
23:06.02 brlcad same you did on linux
23:06.03 ``Erik I have qt
23:06.08 Ralith 'kay
23:06.09 ``Erik in /opt/local/
23:06.14 ``Erik and yeah, a full dev stack
23:06.22 Ralith got cmake?
23:06.32 ``Erik uhm, don't think so, I have a DEV stack, not a wanker stack
23:06.33 ``Erik :D
23:06.38 Ralith D:
23:06.39 Ralith get cmake
23:06.56 ``Erik yeah, updating my ports
23:07.44 ``Erik oh, I did install camke, my bad
23:08.36 ``Erik runs "cmake ."
23:09.00 Ralith okay, which bits broke? :D
23:09.25 Ralith (there are no sane defaults for OSX for almost any of the cmake package-finding modules)
23:09.31 Ralith (though if your pkg-config is working that might not matter)
23:11.36 ``Erik http://pastebin.bzflag.bz/d7f9a4b6
23:12.50 Ralith hm, that's odd
23:13.03 Ralith that's the one cmake module that I'm *not* in some way responsible for.
23:13.11 ``Erik hehehe
23:13.16 Ralith does pkg-config find Qt?
23:15.33 ``Erik hm, my qt3 didn't seem to install the pc file
23:15.59 ``Erik rebuilds it
23:16.38 ``Erik oh, wait, no it's, there heh, qt-mt
23:17.13 Ralith qt3?
23:17.13 Ralith O.o
23:17.23 ``Erik yeah, do you require 4?
23:17.47 ``Erik (aqbanking required qt3, ... *shrug*)
23:18.40 Ralith er, I'm pretty sure Qt4 is necessary, yeah
23:18.44 ``Erik <-- gtk weenie
23:18.55 ``Erik does it require a certain backend? I can do aqua or x11
23:19.02 Ralith I don't want it to
23:19.06 Ralith so let's try aqua :D
23:19.16 ``Erik aight, installing
23:19.22 Ralith ty
23:19.42 ``Erik (source based package management, this'll take a while)
23:19.48 Ralith <3 binary
23:20.14 ``Erik I d'no, the speed of installing isn't worth the versoin hell
23:20.29 ``Erik is in no hurry in his old age
23:20.53 Ralith funnily enough, last time I had version hell was on a source-based system.
23:21.28 ``Erik when I did debian systems, those went pretty well... yum/rpm based system are crap
23:21.49 ``Erik on fbsd, you get your choice, I always do source using portmanager
23:21.56 Ralith well sure
23:22.00 Ralith rpm is crap
23:22.07 Ralith but that's different from 'binaries are crap' :P
23:23.35 ``Erik hrm, I know I put a debian/ in BRL-CAD, I think I was the one who put the rpm spec file in, to
23:23.46 ``Erik don't really remember and svn sucks :D
23:28.22 *** join/#brlcad Patmcc19 (n=chatzill@71-223-60-113.phnx.qwest.net)
23:36.19 CIA-30 BRL-CAD: 03ralith * r35147 10/rt^3/trunk/src/g3d/ (OgreGLWidget.cxx OgreGLWidget.h): Let Ogre know when the Qt window moves.

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