| 00:12.57 | *** join/#brlcad kintel (~kintel@unaffiliated/kintel) | |
| 00:21.24 | *** join/#brlcad infobot (ibot@rikers.org) | |
| 00:21.24 | *** topic/#brlcad is GSoC students: if you have a question, ask and wait for an answer ... responses may take minutes or hours. Ask and WAIT. ;) | |
| 00:26.19 | *** join/#brlcad srjbekrqtqgiejvs (~armin@dslb-088-066-130-190.088.066.pools.vodafone-ip.de) | |
| 01:22.52 | brlcad | DaRock: did you make any progress on build cleanup? |
| 01:30.36 | brlcad | ``Erik: thanks for saving the day (thrice) |
| 01:36.20 | brlcad | DaRock: can you summarize what problems you ran into building? I know (and read) that you tried a lot of different things, but what were the underlying issues specifically? |
| 01:46.34 | brlcad | sees where the grazing point collapse on the tgc went wrong |
| 01:59.49 | *** join/#brlcad milamber (~devlin@2602:306:8094:9990:28c2:d95a:f922:f50) | |
| 02:16.57 | Notify | 03BRL-CAD:brlcad * 69706 brlcad/trunk/include/vmath.h: %g defaults to 6 significant digits meaning it can unhelpfully truncate digits after the decimal point. instead, use %.6f which prints 6 past decimal and gets us precision past our calculation tolerance, near single precision limit. |
| 02:21.27 | Notify | 03BRL-CAD:brlcad * 69707 brlcad/trunk/src/librt/primitives/tgc/tgc.c: clean up the convergence failure reporting, consistently report just 10 times. |
| 02:55.31 | Notify | 03BRL-CAD:brlcad * 69708 brlcad/trunk/src/librt/primitives/tgc/tgc.c: VEXCHANGE() is unnecessary, call VSWAP() instead |
| 03:04.15 | *** join/#brlcad Plazma (~plazma@freenode/staff-emeritus/plazma) | |
| 03:22.58 | brlcad | hello Plazma, you can get 2D drawings trivially from a 3D model using our "rtedge" or "rtwizard" applications. example: http://brlcad.org/~maths22/books/en/images/hacking_brlcad_havoc_views_c.png |
| 03:22.59 | gcibot_ | [ BRL-CAD: Open Source Solid Modeling ] |
| 03:23.34 | brlcad | another: https://brlcad.org/gallery/var/albums/renderings/havoc_rtedge.png |
| 03:44.58 | *** join/#brlcad gabbar1947 (uid205515@gateway/web/irccloud.com/x-xviilmgzxzujfcmj) | |
| 05:39.53 | Notify | 03BRL-CAD:brlcad * 69709 brlcad/trunk/src/librt/primitives/tgc/tgc.c: implement a better solution to the array overrun false positive when eliminating side cylinder hits beyond the end. also wasn't correctly yanking the second point with the new iteration, fixed with this new approach. we tally up a skip list and do just one copy (potentially) instead of nlog(n) copies. also fixed npts <0 when reporting convergence |
| 05:39.55 | Notify | failures (zero isn't a failure). inadvertently includes other cleanup too. |
| 05:39.57 | Notify | ... |
| 05:42.54 | brlcad | starseeker: tgc should be fixed now ... regress and bench passing |
| 05:44.38 | brlcad | it's now back to considering single grazing hits as a hit instead of a miss, so we'll have to be on the lookout for changes we're not testing |
| 05:45.20 | brlcad | it was a problem earlier because it wasn't reducing out-of-bounds cylinder hits fully, which is what solids caught |
| 05:46.53 | brlcad | only time it should happen now is if root solver ever returned 1 hit (which I'm not sure it can) or if we somehow perfectly hit an edge such that we hit one surface but numerically miss the other, just outside tolerance |
| 05:52.06 | Notify | 03BRL-CAD:brlcad * 69710 brlcad/trunk/BUGS: much has changed since 2007, plus recent work should make one hit tgc errors impossible. |
| 06:18.58 | *** join/#brlcad teepee_ (~teepee@unaffiliated/teepee) | |
| 07:41.32 | *** join/#brlcad nmz787_ (~nmz787@unaffiliated/nmz787) | |
| 07:55.30 | *** join/#brlcad KimK (~Kim__@2600:8803:7a81:7400:116f:4b0:3f7c:601) | |
| 08:57.16 | *** join/#brlcad Caterpillar (~caterpill@unaffiliated/caterpillar) | |
| 09:23.29 | *** join/#brlcad merzo (~merzo@user-94-45-58-139.skif.com.ua) | |
| 09:56.02 | DaRock | brlcad: I have a running program now - I'm still learning the system, so I can't specifically say whether I have any run faults yet :) |
| 09:58.40 | DaRock | brlcad: it began with libpng errors in the normal port on FreeBSD - conflict between the system lib and the builtin lib, it couldn't find the builtin. Also urt port is no longer available, and therefore brlcad port was due for removal soon. Since urt is actually builtin, then building the urt port was pointless. |
| 10:02.09 | DaRock | brlcad: specifically to your interest, running the build with bundled option off ran into several issues. First, it couldn't find termlib. It looks for other options, finds them as far as can tell, but refuses to use them. FreeBSD uses the ncurses system, but links the headers and libs to show curses, ncurses, and termcap. It even has libs for termlib, but no header; so in looking for termlib, the build found the lib, |
| 10:02.24 | DaRock | brlcad: but not the header and so failed. |
| 10:05.07 | DaRock | brlcad: A fix was made to FindTERMLIB.cmake to get it to set the variables properly, but it still had issues as you could see with the header. I ended up just linking the termlib.h due to temporal restrictions |
| 10:10.11 | DaRock | brlcad: next was tcl/tk. It wouldn't use the system headers at all - the build system specifically states that it wants to guarantee the builtins. Worse, it only did that for tcl - not tk - so of course there would be conflict in versions. That was a wash really - I hacked the CMakeList.txt to libtclcad to check if the system libs are set to be used and use them and specifically ignore the builtin tcl/tk. |
| 10:14.05 | DaRock | brlcad: I also had to hack it so that it found the tcl/tk include dir, as well as the subfolder as is already done for the builtin - ie tclX.x/generic. I found setting that system wide created issues elsewhere, and could only be specifically set in that cmake file. |
| 10:16.20 | DaRock | brlcad: ideally this stuff should be possible to set from the cli using an env variable like -DTCL_INCLUDE_DIRS or such, as if you are not going auto or bundled then you must be a dev/porter and know what you are doing and can resolve it if things go wrong. My 2c - save the tcl/tk conundrum at least, anyway... ;) |
| 10:17.05 | DaRock | support on here has been good to btw, thanks all.... now if only I could find some good tuts on how to use it.... |
| 10:49.50 | *** join/#brlcad teepee (~teepee@unaffiliated/teepee) | |
| 11:31.13 | *** join/#brlcad merzo (~merzo@94.45.58.139) | |
| 11:33.54 | *** join/#brlcad teepee (~teepee@unaffiliated/teepee) | |
| 11:58.01 | *** join/#brlcad ``Erik_ (~erik@pool-100-16-14-17.bltmmd.fios.verizon.net) | |
| 12:00.56 | *** join/#brlcad teepee (~teepee@unaffiliated/teepee) | |
| 12:52.06 | *** join/#brlcad kintel (~kintel@unaffiliated/kintel) | |
| 12:57.30 | *** join/#brlcad yorik (~yorik@2804:431:f721:afd0:290:f5ff:fedc:3bb2) | |
| 14:23.52 | *** join/#brlcad yorik (~yorik@2804:431:f721:afd0:290:f5ff:fedc:3bb2) | |
| 14:24.37 | *** join/#brlcad teepee (~teepee@unaffiliated/teepee) | |
| 15:08.40 | *** join/#brlcad teepee (~teepee@unaffiliated/teepee) | |
| 16:29.41 | ``Erik | starseeker: https://github.com/froggey/Iota LLVM to CL transpiler |
| 16:29.41 | gcibot_ | [ GitHub - froggey/Iota: LLVM to Common Lisp transpiler. ] |
| 17:35.57 | *** join/#brlcad gabbar1947 (uid205515@gateway/web/irccloud.com/x-jceriaygzjhopbfc) | |
| 22:21.02 | *** join/#brlcad merzo (~merzo@95.132.50.179) | |
| 22:26.16 | *** join/#brlcad teepee (~teepee@unaffiliated/teepee) | |
| 23:34.23 | Notify | 03BRL-CAD:starseeker * 69711 (brlcad/trunk/doc/docbook/system/mann/bot.xml brlcad/trunk/src/libged/bot.c): Expose the bot solidity test through the bot command. |
