| 01:20.46 | *** join/#brlcad milamber (~devlin@d118-75-244-176.try.wideopenwest.com) | |
| 01:24.09 | *** join/#brlcad milamber (~devlin@d118-75-244-176.try.wideopenwest.com) | |
| 02:21.51 | *** join/#brlcad milamber (~devlin@d118-75-244-176.try.wideopenwest.com) | |
| 03:16.54 | *** join/#brlcad milamber (~devlin@d118-75-244-176.try.wideopenwest.com) | |
| 03:47.51 | brlcad | ``Erik: shouldn't be giving any off-by errors any more |
| 03:48.43 | brlcad | at least, you'll have to be more specific |
| 03:49.19 | brlcad | I do think that the solids regression is going to need to have one pixel updated due to increased root solver precision, but it is exactly one pixel |
| 03:50.29 | brlcad | I did walk that one through a debugger to understand it a bit more a few weeks ago |
| 04:01.47 | Notify | 03BRL-CAD:brlcad * 54851 (brlcad/trunk/regress/solidspix.asc brlcad/trunk/src/libbn/poly.c): restore the increased M_SQRT3 precision change that carl had made to the polynomial root solver, which broke the solids regression test. walking the math through the debugger showed that a slightly different root was being returned causing a ray to intersect the massively distorted eto in the solids regression at enough of |
| 04:01.49 | Notify | a delta to cause different light computations. this resulted in a single pixel off-by-many, which we also update in the reference image so that it reflects the new more accurate answer. while we're at it, increase the precision of 1/3 and 1/27th by computing them during static init. |
| 04:19.55 | *** join/#brlcad milamber (~devlin@d118-75-244-176.try.wideopenwest.com) | |
| 04:30.03 | Notify | 03BRL-CAD:brlcad * 54852 brlcad/trunk/TODO: push second iteration tasks further back since we're way off track and release is the priority. richard noticed massive memory leakage in mged blowing out bu_malloc when drawing large geometry, so need to sort that out asap. push the rest back. |
| 09:46.20 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) | |
| 09:50.09 | jordisayol | hello |
| 09:51.07 | jordisayol | with the last commit 54852, rtwizard returns this error http://paste.debian.net/243894/ |
| 11:05.05 | *** join/#brlcad milamber (~devlin@d118-75-244-176.try.wideopenwest.com) | |
| 12:07.46 | *** join/#brlcad milamber (~devlin@d118-75-244-176.try.wideopenwest.com) | |
| 12:35.38 | brlcad | starseeker: presumably related to the tcl/tk upgrade? |
| 13:19.29 | *** join/#brlcad milamber (~devlin@d118-75-244-176.try.wideopenwest.com) | |
| 14:27.45 | *** join/#brlcad caen23_ (~cezar@92.83.177.51) | |
| 14:56.50 | *** join/#brlcad milamber (~devlin@d118-75-244-176.try.wideopenwest.com) | |
| 15:05.53 | *** join/#brlcad merzo (~merzo@136-6-133-95.pool.ukrtel.net) | |
| 15:08.54 | *** join/#brlcad yiyus (1242712427@je.je.je) | |
| 16:05.21 | starseeker | brlcad: probably - I think I got that bit, but I'm not sure what's up with libpng... |
| 16:06.05 | starseeker | oh, you mean rtwizard |
| 16:06.11 | starseeker | checks... |
| 16:08.10 | starseeker | that looks more like a conflict between built and installed tcl/tk |
| 16:08.52 | starseeker | when jordisayol comes back, we need to ask him the details of what he did - clearing the install directory, for example? |
| 16:09.33 | starseeker | it's clearly not happy about the versions, but unless he's got a system install of 8.5.9 lurking somewhere that smells like stale files of some kind |
| 16:11.19 | starseeker | let me build here and I'll check rtwizard |
| 16:12.47 | starseeker | brlcad: sorry, guess that was probably not the best time to do that update... |
| 17:03.10 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) | |
| 17:46.02 | *** join/#brlcad merzo (~merzo@43-210-201-46.pool.ukrtel.net) | |
| 18:03.32 | *** join/#brlcad milamber (~devlin@d118-75-244-176.try.wideopenwest.com) | |
| 18:32.32 | ``Erik | brlcad: sorry, benchmark test shows it on m35, 74 off by 1. iirc, they all happen to be along the wheel torii, skimming hit type issue |
| 18:36.42 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) | |
| 18:40.50 | starseeker | hmm... |
| 18:40.58 | starseeker | yeah, rtwizard is having trouble here too |
| 18:41.00 | starseeker | what the.. |
| 18:41.49 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) | |
| 18:42.21 | starseeker | ah, wait |
| 18:42.34 | starseeker | jordisayol: I think I may know what's going on with rtwizard - one sec... |
| 18:42.48 | jordisayol | starseeker: ok |
| 18:43.51 | Notify | 03BRL-CAD:starseeker * 54853 (brlcad/trunk/src/other/tcl/CMakeLists.txt brlcad/trunk/src/other/tk/CMakeLists.txt): Update CMake's variables for Tcl/Tk versions |
| 18:46.02 | starseeker | note to self - grepping for tcl/tk version number in src code is insufficient... |
| 18:46.23 | starseeker | jordisayol: give it a go |
| 18:47.21 | jordisayol | starseeker: ok. |
| 19:11.09 | jordisayol | starseeker: fixed! many thanks! :-) |
| 19:35.52 | *** join/#brlcad merzo (~merzo@246-240-132-95.pool.ukrtel.net) | |
| 19:45.51 | *** join/#brlcad merzo (~merzo@226-7-132-95.pool.ukrtel.net) | |
| 20:00.52 | *** join/#brlcad milamber (~devlin@d118-75-244-176.try.wideopenwest.com) | |
| 20:41.58 | *** join/#brlcad merzo (~merzo@243-144-132-95.pool.ukrtel.net) | |
| 20:49.12 | *** join/#brlcad crdueck_ (~cdk@24.212.219.10) | |
| 20:50.49 | *** join/#brlcad merzo (~merzo@105-64-133-95.pool.ukrtel.net) | |
| 21:37.33 | *** join/#brlcad merzo (~merzo@183-72-132-95.pool.ukrtel.net) | |
| 21:49.18 | *** join/#brlcad milamber (~devlin@d118-75-244-176.try.wideopenwest.com) | |
| 21:50.39 | *** join/#brlcad crdueck (~cdk@24.212.219.10) | |
| 22:01.16 | *** join/#brlcad merzo (~merzo@141-19-132-95.pool.ukrtel.net) | |
| 22:48.32 | *** join/#brlcad merzo (~merzo@187-63-133-95.pool.ukrtel.net) | |
| 23:30.32 | *** join/#brlcad merzo (~merzo@71-5-132-95.pool.ukrtel.net) | |
| 23:55.09 | *** join/#brlcad milamber (~devlin@d118-75-244-176.try.wideopenwest.com) | |