| 00:03.02 | *** join/#brlcad Twingy (n=justin@74.92.144.217) | |
| 01:06.34 | brlcad | hmm |
| 01:22.24 | iraytrace | Hmmm? |
| 01:23.11 | iraytrace | :( :P :D 9.9 :) |
| 01:25.40 | poolio | :'( |
| 01:47.15 | *** join/#brlcad minusinsk (n=jishi@83.234.35.158) | |
| 03:55.37 | *** join/#brlcad Twingy (n=justin@74.92.144.217) [NETSPLIT VICTIM] | |
| 03:55.37 | *** join/#brlcad vedge (i=vedge@leary.csoft.net) [NETSPLIT VICTIM] | |
| 03:55.37 | *** join/#brlcad PrezKennedy (i=Matt@74.86.45.130) | |
| 03:55.37 | *** join/#brlcad alex_joni (n=juve@emc/board-of-directors/alexjoni) [NETSPLIT VICTIM] | |
| 03:55.37 | *** join/#brlcad b0ef (n=b0ef@062016141231.customer.alfanett.no) [NETSPLIT VICTIM] | |
| 03:55.37 | *** join/#brlcad SWPadnos (n=Me@emc/developer/SWPadnos) [NETSPLIT VICTIM] | |
| 03:55.37 | *** join/#brlcad CIA-32 (n=CIA@208.69.182.149) [NETSPLIT VICTIM] | |
| 05:13.36 | *** mode/#brlcad [+o brlcad] by ChanServ | |
| 06:51.54 | CIA-32 | BRL-CAD: 03brlcad * r30473 10/brlcad/trunk/ (NEWS src/libfb/if_ogl.c): (log message trimmed) |
| 06:51.54 | CIA-32 | BRL-CAD: remove the protections for the 'm' mode in the opengl framebuffer as well as |
| 06:51.54 | CIA-32 | BRL-CAD: requesting indirect buffers, now always requesting direct and seeing what we |
| 06:51.55 | CIA-32 | BRL-CAD: get. my always doing 'multiple window mode', the code now (correctly) acquires |
| 06:51.55 | CIA-32 | BRL-CAD: and releases the opengl context for all of the drawing so that there aren't |
| 06:51.57 | CIA-32 | BRL-CAD: multithreaded crashes on mac and linux. bob and I tag-teamed tracking down the |
| 06:51.59 | CIA-32 | BRL-CAD: cause and a fix for this bug over a couple days. interestingly, the crash on |
| 06:55.56 | *** join/#brlcad Z80-Boy (n=clock@217-162-111-118.dclient.hispeed.ch) | |
| 06:56.54 | CIA-32 | BRL-CAD: 03brlcad * r30474 10/brlcad/trunk/src/ (6 files in 6 dirs): common.h always comes first, there's a regression test that checks for this now too |
| 08:27.29 | *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch) | |
| 09:48.56 | *** join/#brlcad yukonbob (i=1000@s142-179-54-198.bc.hsia.telus.net) | |
| 11:36.23 | *** join/#brlcad yukonbob_ (i=1000@s142-179-54-198.bc.hsia.telus.net) | |
| 12:09.22 | *** join/#brlcad d_rossberg (n=rossberg@bz.bzflag.bz) | |
| 13:55.36 | ``Erik | hrm |
| 14:01.23 | ``Erik | blah blah blah |
| 14:01.39 | ``Erik | brlcad, can we just axe vas4? I'm seeing issues with it again |
| 14:05.29 | clock_ | ``Erik: do you know which function or *.c file translates cyliders into the line for display when I do the "B" command? |
| 14:05.34 | clock_ | into the lines |
| 14:06.28 | ``Erik | yeah |
| 14:06.29 | ``Erik | um |
| 14:06.41 | ``Erik | src/librt/g_tgc.c um, "plot" |
| 14:07.07 | clock_ | Thanks |
| 14:07.12 | ``Erik | rt_tgc_plot, line 1774 |
| 14:08.01 | ``Erik | you'll notice all it does is create a linked list of points... in mged, there's a routine to take those, do matrix crunch on each point and rasterize a line... |
| 14:08.04 | clock_ | great |
| 14:10.26 | clock_ | is this linked list created every time the picture is redrawn? |
| 14:10.35 | clock_ | Or is it kept in memory and just the matrix crunch is changing? |
| 14:10.39 | ``Erik | um, I'm not sure |
| 14:10.52 | ``Erik | I'd hope the latter, and that plot is only called when you 'e' something up |
| 14:12.13 | ``Erik | if you look at the mged code, lost in a maze of twisty spaghetti code, all looke alike |
| 14:27.49 | brlcad | ``Erik: what's the prob? |
| 14:29.36 | brlcad | I believe it calls plot each time |
| 14:29.39 | brlcad | plot really takes no time to recompute for all prims except the old nurbs |
| 14:30.59 | brlcad | plus when you e something up, since all db actions are transactional, if the object was B'd or d's otherwise invalidated, it's removed from memory and reloaded from disk (in case it changed) |
| 14:34.01 | clock_ | brlcad: but if I change something in the scene it doesn't redraw by itself |
| 14:35.16 | ``Erik | old bsd tty api, fbsd7 freaks out a bit on it |
| 14:35.59 | ``Erik | and we never define HAVE_XOPEN for the termios.h style, and if I change that to HAVE_TERMIOS_H, it freaks out because it uses struct pieces that aren't there |
| 14:42.06 | brlcad | ``Erik: aiight, axe |
| 14:44.02 | brlcad | clock_: that's the mged side of things and depends on what you change and how you change it -- some actions force a redraw, some don't |
| 14:44.36 | clock_ | OK |
| 14:44.59 | brlcad | the redraws are intentionally minimized since once you get to real models, it might take several seconds to load the data (i.e. the db read is fast, the plot is fast, but takes forever to blit the damn lines over a remote X11 connection to the display manager) |
| 14:45.24 | clock_ | My model takes forever even over a local X11 connection |
| 14:46.26 | clock_ | You or ``Erik can try how fast it is on your super-machines look at http://ronja.twibright.com/3d/comparison.g and do "B comparison" |
| 14:46.42 | clock_ | That's the one that was too slow |
| 14:46.56 | clock_ | Provided that you have enough time for such experiments, of course. |
| 14:47.17 | brlcad | i'll have to try it in about an hour, this box doesn't have a gui-mged ready to go .. testing something else |
| 14:48.24 | ``Erik | took like 2 seconds for me |
| 14:48.46 | ``Erik | 3.5, sorry |
| 14:49.39 | brlcad | on a completely unrelated topic, some previews of the new BZFlag tank model being worked on: http://bzflag.bz/tmp/glasstank.png |
| 14:50.10 | brlcad | http://www.opencombat.net/files/tankv5_sides.png |
| 14:50.20 | ``Erik | wow, so I'm not gonna be able to say "at least ktank is better than the bzflag model" for much longer? :D *duck* |
| 14:51.07 | brlcad | http://bzflag.bz/tmp/newtank.png |
| 14:51.07 | brlcad | various incarnations, still working some tweaks out |
| 14:51.07 | ``Erik | is that oval on top supposed to be a hatch or something? |
| 14:51.11 | brlcad | like how big the sidewall armor plating is, where the turret actually sits, etc |
| 14:51.16 | brlcad | heh yeah |
| 14:51.45 | brlcad | there's lots of "artistic freedom" going on, but has always been loosely based on our beloved bradley |
| 14:51.48 | clock_ | ``Erik: you have 3GHz? |
| 14:52.00 | brlcad | http://bzflag.bz/tmp/m1bzoverlay.png |
| 14:52.07 | ``Erik | yeah, clock |
| 14:52.30 | clock_ | ``Erik: did you use the "B" command or some other one? Does this command take advantage of the 8 CPUs? |
| 14:52.33 | ``Erik | bradly or abrams? O.o *cough* |
| 14:52.37 | clock_ | You said you have 8 CPUs |
| 14:52.48 | ``Erik | I used 'e', and I don't believe plot is threaded |
| 14:52.56 | brlcad | sry, abrams |
| 14:53.21 | brlcad | yeah, bradley would suck as a bz tank :) |
| 14:53.31 | ``Erik | bradley would suck as any tank *cough* |
| 14:53.50 | brlcad | now that we can actually load separate tank models |
| 14:54.06 | brlcad | I'm looking forward to seeing a battlefield of abrams vs t72's :) |
| 14:54.19 | ``Erik | got a blender .py exporter for bz? |
| 14:54.26 | brlcad | yeah, somewhere |
| 14:54.37 | ``Erik | what format is bz using for tank models? |
| 14:54.48 | ``Erik | cusom? |
| 14:54.51 | ``Erik | custom, even? |
| 14:54.52 | brlcad | basic polygonal |
| 14:54.55 | brlcad | was custom |
| 14:55.03 | brlcad | now there's actually a mini obj loader |
| 14:55.06 | ``Erik | cp g-stl.c g-bz.c |
| 14:55.09 | ``Erik | *cougH* O:-) |
| 14:55.26 | brlcad | g-obj.c already does the trick ;) |
| 14:55.35 | ``Erik | hehehe, cool |
| 14:55.41 | ``Erik | wish I had the time to slap together an obj-g.c |
| 14:55.44 | brlcad | annoying we don't have an obj importer though |
| 14:55.54 | brlcad | keep having to have jeff dump to dxf |
| 14:58.43 | CIA-32 | BRL-CAD: 03erikgreenwald * r30475 10/brlcad/trunk/src/vas4/Makefile.am: disable build (but leave in dist for this release) |
| 14:59.05 | ``Erik | obj is a really simple format, but no one has time to do it :/ |
| 14:59.31 | ``Erik | an excellent "new developer" task that'd be good bragging rights for anyone looking to help with coding |
| 14:59.36 | ``Erik | *cough* *looks around* |
| 15:04.46 | CIA-32 | BRL-CAD: 03erikgreenwald * r30476 10/brlcad/trunk/configure.ac: favor -lpthread to -pthread |
| 15:12.07 | brlcad | ah, vas4 was already deprecated for this release .. when the changes started .. forgot about it |
| 15:13.53 | ``Erik | <-- figures let it go out as dead files this release, if no one bitches, remove it in a release or 2 |
| 15:14.04 | ``Erik | working from home today? |
| 15:22.28 | brlcad | no, bob was coming in today and tomorrow to help test/fix for release, so I'll be in |
| 15:23.29 | brlcad | good idea for vas4 (already has a doc/deprecation.txt listing) |
| 15:26.46 | brlcad | i think i'll meet up and/or come in after lunch, though .. actually got some sleep last night so i'm a bit behind today |
| 15:27.33 | ``Erik | ah, I was asking as a lunch posse notion :) |
| 15:27.49 | brlcad | k |
| 15:29.43 | ``Erik | hrm, 'make dist' is failing for me (no useful error message, just "Error code 1" after some ugly svn related schell crap) |
| 15:31.21 | CIA-32 | BRL-CAD: 03bob1961 * r30477 10/brlcad/trunk/src/tclscripts/archer/Archer.tcl: Put back a few class variables (i.e. they fell out during the ArcherCore transition) and make a few plugin related methods public. |
| 15:33.04 | brlcad | ``Erik: cd regress && make flawfinder |
| 15:33.21 | brlcad | it'd be one of the two new shell tests in the flawfinder.sh |
| 15:34.00 | ``Erik | huh, guess I need python installed |
| 15:34.18 | brlcad | python? |
| 15:34.36 | brlcad | oooh, for flawfinder, yeah |
| 15:34.42 | ``Erik | env: python: No such file or directory |
| 15:34.45 | brlcad | so you just haven't run the test in a long while |
| 15:35.08 | brlcad | could make flawfinder.sh skip the flawfinder part if there's no python |
| 15:35.16 | ``Erik | well, after upgrading that opteron box to fbsd7, I purged all the ports and started minimally adding just enough for BRL-CAD piecewise... |
| 15:35.38 | ``Erik | aside from vim, bash, screen, emacs, portmanager, and pkg_cutleaves, that is |
| 15:37.52 | CIA-32 | BRL-CAD: 03brlcad * r30478 10/brlcad/trunk/regress/flawfinder.sh: make sure we have a python interpreter before proceeding with the flawfinder test |
| 15:38.23 | brlcad | pkg_cutleaves? |
| 15:38.30 | brlcad | wasdat |
| 15:41.27 | CIA-32 | BRL-CAD: 03brlcad * r30479 10/brlcad/trunk/TODO: the ogl framebuffer problem is crashed, though it still needs some cleanup. it's making way too many calls to get/release the context now so it's dog-slow .. need to try what bob had which was just protecting write() |
| 15:52.37 | CIA-32 | BRL-CAD: 03brlcad * r30480 10/brlcad/trunk/src/libfb/if_ogl.c: |
| 15:52.38 | CIA-32 | BRL-CAD: woo hoo! the performance problems were simply because several of the |
| 15:52.38 | CIA-32 | BRL-CAD: glXMakeCurrent acquire/release pairings were .. wrong. rather, they're in the |
| 15:52.38 | CIA-32 | BRL-CAD: wrong place with respect to the logic sometimes only acquiring in a specific |
| 15:52.38 | CIA-32 | BRL-CAD: scope but then releasing in a different scope. made them all consistent and |
| 15:52.38 | CIA-32 | BRL-CAD: sure enough it all behaves spiffily now .. yay for the return of the ogl |
| 15:52.40 | CIA-32 | BRL-CAD: framebuffer. |
| 15:54.18 | CIA-32 | BRL-CAD: 03brlcad * r30481 10/brlcad/trunk/src/libfb/if_ogl.c: remove dead code, just makes maintenance a bitch |
| 15:57.28 | *** part/#brlcad terrylr (i=terrylr@maria-enzersdorf.blauedonau.com) | |
| 16:01.15 | CIA-32 | BRL-CAD: 03brlcad * r30482 10/brlcad/trunk/src/libfb/if_wgl.c: more dead code removal |
| 16:07.16 | clock_ | brlcad: I still also have the problem on my home machine that the mged doesn't start. |
| 16:07.25 | clock_ | The BRL-CAD version there is the latest released. |
| 16:07.31 | clock_ | Worked fine on OpenBSD I think |
| 16:07.37 | clock_ | 7.10.4 IIRC? |
| 16:07.41 | CIA-32 | BRL-CAD: 03bob1961 * r30483 10/brlcad/trunk/src/tclscripts/lib/Display.tcl: Modify rt and rtedge methods to stop unnecessarily specifying parameters like framebuffer, width, height and aspect. |
| 16:09.35 | ``Erik | brlcad: lists all installed packages that are not depended on by anything (safe to remove), also has a recursive interactive removal mode so you can trim your package count down safely |
| 16:10.06 | ``Erik | jim and dave are interested in going out for lunch, dave said korean, fyi |
| 16:11.19 | CIA-32 | BRL-CAD: 03brlcad * r30484 10/brlcad/trunk/src/libfb/if_wgl.c: |
| 16:11.19 | CIA-32 | BRL-CAD: make the wgl interface do the same as the ogl interface, fixing the various |
| 16:11.19 | CIA-32 | BRL-CAD: incorrectly scoped wglMakeCurrent calls so that they actually pair up correctly |
| 16:11.19 | CIA-32 | BRL-CAD: with the logic. also, actually make them release the context (the code was |
| 16:11.19 | CIA-32 | BRL-CAD: wrong, they just kept it) so when parallel support is added to windows, it'll |
| 16:11.21 | CIA-32 | BRL-CAD: actually work. |
| 16:11.55 | brlcad | ``Erik: okay, just lemme know where |
| 16:12.03 | brlcad | or call on the way or something |
| 16:12.29 | brlcad | clock_: need more info, but I'm betting it's configure-related -- try adding --enable-all to the configure line |
| 16:12.41 | clock_ | brlcad: OK |
| 16:13.16 | brlcad | if you're compiling that is .. if you installed from binary, that'd be good to know if it's not working (i've not heard that) |
| 16:14.26 | clock_ | Putiting a binary on my system? Haha. |
| 16:14.35 | clock_ | Putting |
| 16:15.01 | clock_ | Binaries can contain hidden rootkits. |
| 16:15.02 | ``Erik | um |
| 16:15.06 | ``Erik | ok, actually, it's changed to olive shack |
| 16:15.12 | ``Erik | jim veto's korean |
| 16:15.26 | brlcad | boo hiss |
| 16:15.37 | ``Erik | mebbe tomorrie :) |
| 16:16.01 | ``Erik | see ya after lunch then :) |
| 16:17.30 | brlcad | you all leaving soon? |
| 16:22.53 | *** join/#brlcad elite01 (n=elite01@dslc-082-082-076-143.pools.arcor-ip.net) | |
| 16:37.32 | *** join/#brlcad Elperion (n=Bary@p54874F28.dip.t-dialin.net) | |
| 16:39.24 | CIA-32 | BRL-CAD: 03bob1961 * r30485 10/brlcad/trunk/src/libfb/if_ogl.c: Try to get a direct context (speed). |
| 16:47.06 | *** join/#brlcad elite01 (n=elite01@dslc-082-082-076-143.pools.arcor-ip.net) | |
| 17:54.55 | ``Erik | O.o |
| 17:56.42 | ``Erik | brlcad, no .mgedrc http://pastebin.bzflag.bz/d3dfc9116 |
| 17:59.12 | *** join/#brlcad archivist (n=archivis@host81-149-119-172.in-addr.btopenworld.com) | |
| 19:34.13 | *** join/#brlcad Z80-Boy (n=clock@77-56-88-167.dclient.hispeed.ch) | |
| 19:54.50 | brlcad | not sure about that crash .. looks kinda like a Tk bug really |
| 19:55.06 | brlcad | using a system tk apparently, tried building from our end? |
| 19:55.34 | ``Erik | ours works |
| 19:55.35 | brlcad | if you have and it still crashes, try with --disable-freetype (I think that's the opt) in configure.ac as a tcl/tk configure option |
| 19:55.40 | brlcad | ahh |
| 19:55.43 | brlcad | weird |
| 19:55.49 | ``Erik | I d'no why system 85 doesn't |
| 19:56.23 | ``Erik | system wish85 comes up, I d'no how to use it though |
| 19:57.33 | brlcad | can you pop up that stack 17 frames and see what it's actually evaluating? |
| 19:58.20 | ``Erik | oh, I already trashed it |
| 19:58.20 | brlcad | so need a tcl debugger .. though even if we had one, it probably wouldn't work with mged since it's an interpreter in itself |
| 19:58.23 | brlcad | k |
| 19:58.40 | ``Erik | I figured I'd focus more on the metaball issues |
| 20:02.08 | CIA-32 | BRL-CAD: 03erikgreenwald * r30486 10/brlcad/trunk/TODO: note "minimal system" testing |
| 21:19.06 | CIA-32 | BRL-CAD: 03erikgreenwald * r30487 10/brlcad/trunk/src/librt/g_metaball.c: fluff up the isopotential method a bit... back to the old way, kinda |
| 21:42.43 | CIA-32 | BRL-CAD: 03bob1961 * r30488 10/brlcad/trunk/src/tclscripts/geometree/GeometryBrowser.tcl: Remove extra set of double quotes in format statement of rgbToHex method. |
| 21:44.09 | Z80-Boy | ``Erik: you mean the gcc that compiled compromised gcc and login? |
| 21:44.45 | Z80-Boy | Nice self-reference btw. Hofstadter would surely masterbate to it. |
| 21:45.32 | Z80-Boy | Write in machine code then. That's not compromised :) |
| 21:46.35 | Z80-Boy | The difference between machine code and compiler is like a skateboard and a tilting train |
| 21:46.46 | Z80-Boy | In a tilting train, the machine does the traction and tilting for you. |
| 21:47.11 | Z80-Boy | On a skateboard you have to learn how the machine feels, she won't learn it. But then - then you can do *ANYTHING* you want! |
| 21:52.30 | ``Erik | um, ken thompsons "reflections on trusting trust" |
| 21:53.21 | ``Erik | from '84? |
| 21:56.45 | alex_joni | there can be no ultimate trust |
| 21:57.05 | ``Erik | only as much as you can have ultimate provability |
| 21:58.34 | alex_joni | you might have some trust for your own code |
| 21:58.52 | ``Erik | trust it to do what? O.o :D |
| 22:00.02 | alex_joni | well.. probably something |
| 22:00.14 | alex_joni | most likely not what it was intended to do :) |
| 22:31.10 | CIA-32 | BRL-CAD: 03erikgreenwald * r30489 10/brlcad/trunk/src/mged/typein.c: set a default "goo" value on metaballs |
| 23:32.10 | *** join/#brlcad vedge (i=vedge@leary.csoft.net) | |
