| 05:48.59 | Notify | 03BRL-CAD:brlcad * 56975 brlcad/trunk/bench/run.sh: reduce the default deviation to 2% and the number of frames to average across up to 5. this should allow the benchmark number to become considerably more stabilized at the expense of just a few more frames (on average) |
| 05:58.36 | Notify | 03BRL-CAD:brlcad * 56976 brlcad/trunk/NEWS: changed the benchmark defaults from 3% deviation to 2% deviation and from 3 stabilization frames to 5. this will make the benchmark take a fair bit longer on some platforms, but should give a much more stable and repeatable benchmark performance metric. |
| 07:33.44 | *** join/#brlcad d_rossberg (~rossberg@66-118-151-70.static.sagonet.net) |
| 07:45.49 | *** join/#brlcad caen23 (~caen23@92.81.204.102) |
| 08:22.38 | Notify | 03BRL-CAD:phoenixyjll * 56977 brlcad/trunk/src/libbrep/boolean.cpp: Separate the code, so that we can then update the connectivity graph after we have done the surface partitioning. |
| 08:43.20 | *** join/#brlcad kesha (~kesha@49.202.231.185) |
| 11:05.06 | Notify | 03BRL-CAD:tbrowder2 * 56978 brlcad/trunk/src/util/dsp_add.c: prepare for use in auto-man-page process |
| 11:08.02 | Notify | 03BRL-CAD:tbrowder2 * 56979 brlcad/trunk/src/util/dsp_add.c: simplify progname handling |
| 11:09.59 | Notify | 03BRL-CAD:tbrowder2 * 56980 brlcad/trunk/src/util/dsp_add.c: ws, style |
| 11:15.57 | Ch3ck | tryinjg to write a comb_rest() routine which restores the original matrix transformation for a combination while moving up the tree but i wish to understand clearly how it is on a combination. |
| 11:17.07 | Ch3ck | I mean the matrix is stored on the combmbination, I also wish to know if i could do that by calling rt_db_get_internal(..,..,mat_t, ..); like this? |
| 11:23.21 | Notify | 03BRL-CAD:tbrowder2 * 56981 brlcad/trunk/src/util/dsp_add.c: rename function because of new usage var |
| 11:28.33 | Notify | 03BRL-CAD:tbrowder2 * 56982 brlcad/trunk/src/util/dsp_add.c: simplify: use Usage for no args |
| 11:34.44 | Notify | 03BRL-CAD:tbrowder2 * 56983 brlcad/trunk/src/util/dsp_add.c: style |
| 11:46.31 | Notify | 03BRL-CAD:phoenixyjll * 56984 brlcad/trunk/src/libbrep/boolean.cpp: More connectivity graph support: update after surface partitioning. |
| 11:48.44 | Notify | 03BRL-CAD Wiki:Phoenix * 6009 /wiki/User:Phoenix/GSoc2013/Reports: /* Week 10 */ |
| 11:55.55 | Notify | 03BRL-CAD:tbrowder2 * 56985 brlcad/trunk/src/util/dsp_add.c: add option descriptions |
| 13:03.19 | Notify | 03BRL-CAD:tbrowder2 * 56986 brlcad/trunk/src/util/yuv-pix.c: setting up for auto-man-page |
| 13:06.06 | Notify | 03BRL-CAD:tbrowder2 * 56987 brlcad/trunk/src/util/yuv-pix.c: style |
| 13:12.27 | Notify | 03BRL-CAD:tbrowder2 * 56988 brlcad/trunk/src/sig/ustats.c: prep for auto-man page; simplify usage; hardwire progname |
| 13:15.15 | Notify | 03BRL-CAD:tbrowder2 * 56989 brlcad/trunk/src/shapes/wire.c: prep for auo-man-page; simplify; hardwire progname; style |
| 13:16.17 | Notify | 03BRL-CAD:tbrowder2 * 56990 brlcad/trunk/misc/auto-man-page/auto-man-progs.txt: reorder dsp_add |
| 13:30.09 | Notify | 03BRL-CAD:tbrowder2 * 56991 (brlcad/trunk/doc/burst/Make-docs.sh brlcad/trunk/doc/burst/Makefile brlcad/trunk/doc/burst/run_doclifter.sh): pref and option changes |
| 13:38.20 | *** join/#brlcad mpictor_ (~mpictor_@2600:1015:b105:15a5:0:48:2fe4:8e01) |
| 14:07.56 | *** join/#brlcad mpictor_ (~mpictor_@2600:1015:b105:15a5:0:48:2fe4:8e01) |
| 14:22.06 | Notify | 03BRL-CAD:vladbogo * 56992 brlcad/trunk/src/libdm/dm-qt.cpp: Process events before generating a new update request event - this is necessary so that the window is exposed when performing the first update request event. |
| 15:07.20 | *** join/#brlcad Ch3ck_ (~Ch3ck@66-118-151-70.static.sagonet.net) |
| 15:07.37 | *** join/#brlcad ejno (~ejno@unaffiliated/kazaik) |
| 15:09.49 | Notify | 03BRL-CAD:erikgreenwald * 56993 brlcad/trunk/src/conv/step/ON_Brep.cpp: strcmp -> bu_strcmp |
| 15:33.30 | ejno | brlcad: it seems to support double |
| 15:34.41 | ejno | but I'm currently working on problems with opencl 1.1 vs 1.2 |
| 15:42.31 | ejno | actually it seems that was something else |
| 15:46.03 | *** join/#brlcad kesha (~kesha@49.249.16.159) |
| 15:48.25 | ejno | but I am having other problems such as black output. The rt output is not black |
| 16:21.51 | brlcad | all black output or black in areas not expected? |
| 16:22.06 | brlcad | hickoryknoll: did you get the build system integration sorted out? |
| 16:24.27 | hickoryknoll | brlcad: yes. Right now I'm basically rewriting it to get rid of a huge amount of junk I had in there from troubleshooting and fiddling and stuff. |
| 16:32.40 | Notify | 03BRL-CAD:ejno * 56994 (brlcad/branches/opencl/src/librt/CMakeLists.txt brlcad/branches/opencl/src/librt/primitives/sph/sph.c brlcad/branches/opencl/src/librt/primitives/sph/sph_shot.cl): changes for other systems |
| 16:35.58 | brlcad | ``Erik: did you upgrade or fix denyhosts? |
| 16:35.58 | brlcad | hickoryknoll: great |
| 16:36.23 | brlcad | if it's compiling, you can go ahead and commit it |
| 16:36.40 | brlcad | then make more commits as you make improvements |
| 16:37.17 | ``Erik | brlcad: it got reinstalled when I did the big ports purge/rebuild, that's all I can think of |
| 16:37.45 | brlcad | ah cool. looks like it's actually working now |
| 16:38.04 | ``Erik | <-- didn't know that it wasn't working O.o |
| 16:38.35 | brlcad | i noticed and mentioned it a few months back, but didn't have time to investigate how to fix it |
| 16:38.48 | brlcad | it was writing out the firewall rules to a file.. but nothing was reading that file |
| 16:38.52 | Notify | 03BRL-CAD:starseeker * 56995 brlcad/trunk/src/conv/step/ON_Brep.cpp: Put the context at the beginning, break it into its own function, fix a problem where application instances were set to garbage info, set up oriented edge code to have more than the number of edges in openNURBS. |
| 16:39.45 | ``Erik | huh, wonder if something was messed up with the python install or something *shrug* |
| 16:42.27 | brlcad | yep, dunno, dunt care :) |
| 16:42.33 | brlcad | it works and thats great |
| 16:42.44 | brlcad | no longer needs to be my scripted hack |
| 16:49.13 | ejno | brlcad: all black output |
| 16:49.22 | ejno | rt works fine, though |
| 16:56.20 | brlcad | ejno: have you confirmed that a single ray misses? |
| 16:56.42 | brlcad | running "nirt" in mged will fire a ray through the view center and tell you if it hits |
| 17:00.51 | ejno | brlcad: it hits/misses correctly. There is debug output from the opencl function, so it is running. Possibly the data is not being returned/received correctly |
| 17:04.41 | ejno | nvm, forgot to re-build it |
| 17:06.32 | Notify | 03BRL-CAD:brlcad * 56996 brlcad/trunk/TODO: group the attribute tasks together except for the binary-incompatible one |
| 17:07.54 | ejno | brlcad: no, it's not hitting/missing correctly. Currently investigating |
| 17:08.43 | Notify | 03BRL-CAD:brlcad * 56997 brlcad/trunk/TODO: separate out a separate section for attributes |
| 17:27.29 | Notify | 03BRL-CAD:tbrowder2 * 56998 brlcad/trunk/src/util/fix_polysolids.c: style |
| 17:28.17 | Notify | 03BRL-CAD:brlcad * 56999 brlcad/trunk/TODO: expand the results of our (nick and my) brainstorming session for how to fix our attribute system issues. we need versioning and a means to upgrade a file to newer conventions. shouldn't be much more work to use that same system for downgrading too. include several tasks related to this envisioned plan. comments and discussion welcome. |
| 17:28.56 | brlcad | hickoryknoll: does it compile cleanly? |
| 17:38.32 | Izak_ | brlcad:Did you say you tested prep and shot for the hrt already? |
| 18:04.23 | Notify | 03BRL-CAD:brlcad * 57000 brlcad/trunk/BUGS: attr set should always do exactly as its told |
| 18:05.15 | Notify | 03BRL-CAD:carlmoore * 57001 brlcad/trunk/regress/repository.sh: remove trailing blank, and in the 'for' line remove a duplicate occurrence of 'stricmp' |
| 18:06.16 | brlcad | woot, 57000 |
| 18:06.29 | brlcad | Izak_: why would I say that? :) |
| 18:06.38 | brlcad | no, I did not say that |
| 18:06.48 | brlcad | I may have asked you if you tested them |
| 18:07.03 | ``Erik | can't compile hrt, isgreater() is a c99 construct |
| 18:07.36 | ``Erik | src/librt/primitives/hrt/hrt.c:248: warning: implicit declaration of function 'isgreater' |
| 18:07.39 | ``Erik | make[3]: *** [src/librt/CMakeFiles/librt.dir/primitives/hrt/hrt.c.o] Error 1 |
| 18:07.53 | brlcad | Izak_: so that needs fixing asap, but I also had to make substantial changes to prep() as half the logic you had in there was completely unused |
| 18:08.22 | brlcad | so I doubt it works unless you just forgot to remove code |
| 18:09.24 | Izak_ | brlcad:I am modifying the prep() function now |
| 18:13.25 | Ch3ck_ | brlcad: for walking up the tree i'm thinking of using db_functree() is this ok? or must I use db_preorder_traverse since i think it'll be easier calling the functions directly? |
| 18:14.35 | brlcad | Ch3ck_: just was replying to your message |
| 18:15.25 | brlcad | db_functree() is fine |
| 18:15.32 | brlcad | any traversal method is fine really, so long as it's stopping at the right nodes in the right order |
| 18:16.01 | brlcad | looks like db_functree will stop at both leaves and combs, so you're good to go to use it |
| 18:16.17 | brlcad | but see my message, you do not want to start with leaves |
| 18:18.53 | Ch3ck_ | ok db-functree walks just fine here is my code for pulling combinations on pastebin http://pastebin.com/2UYmyTGN . Wish to know if the logic is ok? |
| 18:19.43 | brlcad | Ch3ck_: cannot get to pastebin.com, would you use a different pastebin |
| 18:19.49 | brlcad | plus they suck |
| 18:20.51 | Ch3ck_ | :) yeah will do.. |
| 18:21.09 | ``Erik | I'm fond of paste.lisp.org, some like paste.kde.org, pastebin.ca, ... |
| 18:21.32 | ``Erik | or make a file in your public_html/ :) |
| 18:21.35 | brlcad | and I have no idea if the logic will be okay other than blatant errors - you need to have test code that proves it works |
| 18:23.17 | Ch3ck_ | http://paste.kde.org/pe4239999/ |
| 18:23.54 | *** join/#brlcad caen23 (~caen23@92.81.204.102) |
| 18:24.36 | Ch3ck_ | ok just wanted you to haveo since we have to understand that our logic works before our compiler confirms |
| 18:24.50 | brlcad | Ch3ck_: "restore" implies that a matrix was previously there, which is not necessarily the case |
| 18:25.33 | ``Erik | wonders if tdd would be helpful |
| 18:25.45 | Ch3ck_ | brlcad: which means I could just rename it comb_pull() ? |
| 18:26.06 | brlcad | Ch3ck_: unless you're going to write pseudocode, it's not efficient to review code that hasn't been passed through the compiler |
| 18:27.36 | brlcad | and what ``Erik said is pretty much what I assume most people should be doing by default .. if you're not, you should |
| 18:27.41 | brlcad | tdd == test driven development |
| 18:27.53 | Ch3ck_ | yeah.. |
| 18:28.05 | brlcad | i.e., write a little test program like you did with the unit tests (before the functions you're calling even exists) |
| 18:28.19 | Ch3ck_ | ok |
| 18:28.20 | brlcad | then implement your functions |
| 18:29.09 | brlcad | makes you think about the problem in the most simplest terms first, without getting distracted by structs and types and available API |
| 18:29.18 | brlcad | how simple should it be |
| 18:29.31 | brlcad | as an actual working program |
| 18:31.34 | Izak_ | ``Erik:I already fixed the 'isgreater' thing earlier today. Doing a fresh checkout to build before commiting. |
| 18:31.37 | brlcad | fwiw, you do have some blatant syntax errors, and I don't understand the reslationship of those two functions |
| 18:32.20 | ``Erik | Izak_: commit often, update even more often :) |
| 18:32.27 | brlcad | Izak_: I have a fresh checkout and it's still there |
| 18:32.52 | brlcad | and fresh checkout is not needed, just make sure one is up-to-date (svn up) |
| 18:33.48 | brlcad | AH .. I see what you were saying, my confusion by the wording |
| 18:34.00 | brlcad | YOU are doing a fresh checkout .. got it |
| 18:34.49 | Izak_ | brlcad: I mean I am still doing the chekout |
| 18:34.51 | Ch3ck_ | well brlcad: will get started on some unit tests for all the routines i'm to implement. |
| 18:35.37 | brlcad | Izak_: yeah, I understand now .. |
| 18:36.08 | brlcad | Izak_: you'll definitely want to minimize checkouts! .. that can take several minutes, just verify a clean build and verify what you're committing before you commit |
| 18:36.37 | brlcad | you can test a commit by running "svn diff path/to/file1 path/to/file2 | less" read it, make sure it's only the changes intended |
| 18:37.10 | brlcad | "svn diff src/librt/primtivies/hrt" for example to commit everything changed in the hrt dir |
| 18:37.27 | brlcad | rather, to SEE everything changed |
| 18:37.33 | brlcad | then change diff to commit if it's right |
| 18:38.01 | Izak_ | brlcad: I minimise them . Just don't know what went wrong with the one I had. I build and had error |
| 18:38.05 | Izak_ | svn checkout https://svn.code.sf.net/p/brlcad/code/brlcad/trunk brlcad |
| 18:38.34 | Izak_ | CMakeFiles/librt.dir/primitives/hrt/hrt.c.o: file not recognized: File truncated |
| 18:39.09 | Izak_ | brlcad: Sorry not the "svn ....". The error was "CMakeLIst...." |
| 18:39.30 | ``Erik | might try rm'ing the offending file and build again |
| 18:40.03 | ``Erik | (almost looks as if gcc were killed strangely and didn't clean up a partial file) |
| 18:40.15 | ``Erik | s/were/was/ |
| 18:40.49 | brlcad | Izak_: if you've not done so, create an empty directory and run cmake from there |
| 18:41.03 | Izak_ | ``Erik: It could be that |
| 18:41.07 | brlcad | then if/when something goes wrong like that, you can just delete the entire build directory (or run make clean) |
| 18:41.18 | brlcad | without affecting your source tree or needing a new checkout |
| 18:41.42 | brlcad | this is documented in the INSTALL file |
| 18:41.52 | brlcad | and in HACKING iirc |
| 18:42.02 | brlcad | at least under the release testing steps |
| 18:44.20 | Izak_ | brlcad: Is it true that BRL-CAd is the oldest open source repository existing ? |
| 18:44.38 | brlcad | Izak_: as far as I'm aware, yes |
| 18:44.43 | Izak_ | s/BRL-CAd/BRL-CAD |
| 18:44.48 | ``Erik | there was an article on ohloh that placed us right behind emacs and gcc iirc (or in that neighborhood, at least) |
| 18:45.22 | ``Erik | http://meta.ohloh.net/2007/08/worlds_oldest_source_code_repositories/ |
| 18:45.27 | ``Erik | ah, we are listed as the oldest, w00t |
| 18:45.46 | Izak_ | Izak:So this code is really as old as my Uncle or sth |
| 18:46.28 | brlcad | Izak_: "this code" tries to characterize more than a million lines as having one age, and that's not true |
| 18:46.55 | Ch3ck_ | brlcad: could you please help me with a test i could use to test the comb_restore() or pull_comb() routine? |
| 18:47.00 | ``Erik | checking out early revisions is good fun, code spelunking! |
| 18:47.27 | brlcad | Izak_: great for statistics: http://www.ohloh.net/p/brlcad |
| 18:48.42 | brlcad | notes our scan is stalled |
| 18:51.42 | Ch3ck_ | brlcad: what the status about the Doc Camp? Since I have alot work to do on the Docs.. |
| 18:52.09 | Ch3ck_ | :) |
| 18:54.06 | brlcad | Ch3ck_: more on that will be on the list |
| 18:54.16 | brlcad | there's more that want to go than can go, so selections have to be made |
| 18:56.00 | Ch3ck_ | so when are we expecting results? |
| 18:56.08 | Ch3ck_ | or selectees? |
| 18:56.29 | brlcad | this week |
| 18:56.35 | brlcad | before friday |
| 18:57.07 | Ch3ck_ | nice. |
| 18:57.44 | Ch3ck_ | well concerning the corrections to my patches I've a already updated patch 225 and |
| 18:58.41 | Ch3ck_ | made the necessary corrections to patch 231 |
| 18:58.50 | Ch3ck_ | for the committed patches.. |
| 19:02.19 | Notify | 03BRL-CAD:ejno * 57002 (brlcad/branches/opencl/src/librt/primitives/sph/sph.c brlcad/branches/opencl/src/librt/primitives/sph/sph_shot.cl): fix problems; work with opencl 1.1; align memory used with opencl |
| 19:03.09 | Notify | 03BRL-CAD:ejno * 57003 brlcad/branches/opencl/src/librt/primitives/sph/sph_shot.cl: remove old code |
| 19:23.04 | *** join/#brlcad vladbogo (~vladbogo@188.25.238.69) |
| 19:46.34 | Notify | 03BRL-CAD Wiki:NyahCh3ck20 * 6010 /wiki/User:NyahCh3ck20/GSoc2013/Coding_Repor: /* 19 August - 25 August */ |
| 19:51.35 | Notify | 03BRL-CAD Wiki:Vladbogolin * 6011 /wiki/User:Vladbogolin/GSoC2013/Logs: /* Week 10 */ |
| 20:15.03 | *** join/#brlcad caen23 (~caen23@92.81.204.102) |
| 20:17.21 | *** join/#brlcad mpictor (~mark@2601:d:b280:b5:d63d:7eff:fe2d:2505) |
| 21:04.26 | Notify | 03BRL-CAD:iiizzzaaakkk * 57004 brlcad/trunk/src/librt/primitives/hrt/hrt.c: removed the isgreater keyword which is C99 standard. Modified logic of rt_hrt_prep: No need for max variable anylonger. My intuition tells me to make Xu and Yu shorter than Zu. |
| 21:14.19 | Notify | 03BRL-CAD Wiki:IIIzzzaaakkk * 6012 /wiki/User:Izak/GSOC_2013_logs: /* August 19th to August 24th */ |
| 21:23.40 | *** join/#brlcad KimK (~Kim__@wsip-184-176-200-171.ks.ks.cox.net) |
| 21:28.32 | Notify | 03BRL-CAD:starseeker * 57005 brlcad/trunk/src/conv/step/ON_Brep.cpp: Going to need to re-think approach to edge building. |