| 00:43.12 | *** join/#brlcad andrecastelo (n=chatzill@189.71.36.61) |
| 00:43.36 | andrecastelo | good evening folks |
| 01:13.24 | *** join/#brlcad quentusrex (n=quentusr@c-71-197-244-228.hsd1.or.comcast.net) |
| 01:34.03 | *** join/#brlcad dtidrow_ (n=dtidrow@c-69-255-182-248.hsd1.va.comcast.net) |
| 01:39.15 | andrecastelo | hey ``Erik |
| 03:36.41 | learner | howdy andrecastelo |
| 03:36.47 | learner | que tal? |
| 03:37.07 | andrecastelo | learner: hey what's up? |
| 03:37.30 | learner | not much, just finished munching on some boiled peanuts, new fresh batch :) |
| 03:37.40 | andrecastelo | learner: i see |
| 03:38.01 | andrecastelo | learner: you're using translators now?? :b |
| 03:48.20 | learner | tanslators? |
| 03:48.28 | learner | er translators? |
| 03:48.59 | andrecastelo | [00:36] <learner>que tal? |
| 03:49.10 | andrecastelo | i just thought :b |
| 03:50.35 | learner | jej, ya entiendo |
| 03:50.52 | learner | y no :P |
| 03:57.04 | learner | andrecastelo: so just about ready to go? :) |
| 04:34.22 | brlcad | apparently not |
| 04:35.07 | pacman87 | hi brlcad |
| 04:35.12 | brlcad | howdy pacman87 |
| 04:35.18 | brlcad | you ready? :) |
| 04:35.24 | pacman87 | pretty much |
| 04:35.28 | brlcad | great |
| 04:35.34 | pacman87 | got a nice 3 monitor display set up :) |
| 04:35.42 | brlcad | so got it all working eh |
| 04:35.43 | pacman87 | 4480x1050 |
| 04:35.47 | brlcad | sweet |
| 04:35.56 | pacman87 | well, not exaclty how i want it, but close enough |
| 04:36.42 | pacman87 | i didnt get a chance to discuss the deliverables with d_rossburg before he left |
| 04:37.19 | brlcad | sense *gasp* private e-mails :) |
| 04:38.47 | pacman87 | and for patches to get to commit access, i'm planning on continuing the hyperboloid |
| 04:39.00 | brlcad | great |
| 04:39.25 | pacman87 | though most of the patching i'd do is changes to the g_hyp.c file, which is new |
| 04:39.30 | brlcad | nods |
| 04:40.47 | brlcad | pacman87: fyi, for future reference, you can/should combine all those patch files into just one patch file |
| 04:40.57 | brlcad | svn diff > your_changes.patch |
| 04:41.07 | brlcad | if you svn add the new file, you can even bundle in new files |
| 04:41.42 | pacman87 | right, i was keeping them seperate so i could keep track of the files i changed, but i can do that just as easily by reading the diff :) |
| 04:43.29 | brlcad | yep |
| 04:43.45 | brlcad | and patches are usually for the patcher's convenience/use, not the submitter's ;) |
| 04:44.37 | pacman87 | i'll pull them all and put up a unified patch on my next change |
| 04:44.49 | pacman87 | working out exactly what tess() does |
| 04:50.15 | brlcad | no need to rework the patches, that's just a pedantic detail "for next time" and/or others |
| 04:52.17 | pacman87 | well, if i'm doing a svn diff, they should all be there, right? |
| 04:52.24 | brlcad | yep |
| 04:52.49 | brlcad | you should run the diff for patches as well as before (all) commits too |
| 04:53.05 | brlcad | so you can do the review before you commit |
| 04:53.16 | pacman87 | right |
| 04:53.21 | pacman87 | needs to svn up |
| 04:54.01 | brlcad | which you can do now |
| 04:54.58 | pacman87 | updated to revision 31147 :) |
| 04:55.32 | brlcad | wasn't referring to the svn up :) |
| 04:56.19 | pacman87 | i'm checking to make sure nothing broke on the update |
| 04:59.37 | brlcad | always a good thing :) |
| 05:00.12 | brlcad | you should do a quick test to make sure it's set up proper |
| 05:08.02 | pacman87 | seems i have some svn setup to do; "svn diff | grep SOURCE | less" gives 1242 lines :( |
| 05:11.42 | brlcad | just usually does "svn diff | less" .. and reads |
| 05:12.20 | brlcad | it's mostly a quick review of "are these the changes/files that you intended" |
| 05:13.13 | brlcad | coupled with a compilation test and maybe a run-time test depending on the change, that's usually plenty reasonable |
| 05:14.39 | pacman87 | did src/librt/debug.h get moved somewhere? |
| 05:15.35 | brlcad | ayep |
| 05:21.05 | brlcad | to find out what happened to a file that was potentially deleted with svn, you can run "svn log --verbose src/librt | less" .. look for the src/librt/debug.h line and it'll tell you when the file was last modified, what the modification was, etc |
| 05:25.36 | pacman87 | so it was removed, not moved |
| 05:28.55 | brlcad | yep |
| 05:29.05 | brlcad | it was empty, obsoleted a long long time ago |
| 05:31.17 | pacman87 | seems there's a div by zero possibility in my rt_hyp_norm() i didn't notice the first time :\ |
| 05:32.00 | brlcad | there were a few minor issues like that |
| 05:33.38 | pacman87 | is NEAR_ZERO() the usual way to handle this? |
| 05:33.48 | brlcad | yeah |
| 05:34.01 | brlcad | all floating point equality comparisons should use that |
| 05:35.15 | pacman87 | i wasn't talking about a comparison, though |
| 05:35.37 | pacman87 | the problem happens when z=0 (around the 'waist' of the hyperboloid) |
| 05:36.04 | brlcad | nods |
| 05:36.11 | pacman87 | in that case, the normal is in the x-y plane, and i can use a different algorithm |
| 05:37.34 | pacman87 | so would it be better to check z or what i'm dividing by? |
| 05:46.30 | brlcad | usually what you're diving by, or the code leading up to it |
| 05:46.43 | brlcad | since it's the div-by-zero that will cause a run-time floating point exception/crash |
| 05:46.59 | brlcad | otherwise you'll want to check both |
| 05:47.14 | brlcad | you've got a slew of potential divide-by-zero's |
| 05:47.28 | brlcad | with the right parameters |
| 05:48.16 | brlcad | e.g. if a == b == sqrt(.5) |
| 05:48.25 | brlcad | (.7071067..) |
| 05:49.26 | pacman87 | which function? |
| 05:50.00 | brlcad | at least norm, noticed a few other potentials in shot too iirc, but I didn't look too hard |
| 05:50.09 | brlcad | work in progress ;) |
| 05:50.18 | brlcad | you should still apply what you have now |
| 05:50.30 | brlcad | and just work the mods/fixes/improvements incrementally |
| 05:50.32 | pacman87 | it does compile :) |
| 05:50.46 | brlcad | that's an important first step :) |
| 05:50.59 | pacman87 | but compile != correctness |
| 05:51.26 | brlcad | sure, but it's a start and you intend to improve it right? |
| 05:51.31 | pacman87 | of course |
| 05:51.53 | brlcad | better to see the work progress incrementally |
| 05:52.06 | brlcad | you've put in a lot of good effort so far |
| 05:53.30 | brlcad | so if it wasn't obvious .. |
| 05:53.46 | brlcad | you should just commit what you have now, and keep going |
| 05:54.13 | pacman87 | right, now i'm trying to fix svn |
| 05:54.26 | brlcad | fix? |
| 05:54.40 | brlcad | you said it compiles no? :) |
| 05:54.45 | pacman87 | svn diff gives a whole lot more lines than needed |
| 05:54.49 | brlcad | ah |
| 05:55.03 | pacman87 | fix svn, not the code therein |
| 05:55.17 | brlcad | svn revert will undo changes to files you didn't intend to modify |
| 05:55.54 | brlcad | it'll also undo delete, add, and other file "events" |
| 05:57.53 | pacman87 | the full svn diff is 39189 lines, so i'll revert and reapply my patches |
| 05:58.23 | brlcad | how so? what's in the diff? |
| 05:58.50 | brlcad | "svn status" will tell you the files that are actually modified |
| 05:58.53 | pacman87 | ~pastebin |
| 05:58.54 | ibot | [~pastebin] A "pastebin" is a web-based service where you can paste anything over 3 lines without flooding the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://www.rafb.net/paste |
| 06:01.28 | pacman87 | too big for pastebin, but here it is if you want it: https://webspace.utexas.edu/trv82/www/brlcad.diff |
| 06:02.35 | brlcad | yep, the usual |
| 06:03.26 | brlcad | a lot of windows line-terminated files were converted to unix line endings, usually caused by overly ambitious/invasive GUI IDE's |
| 06:03.49 | brlcad | svn revert src/other && find . -name \*.dsp -exec svn revert {} \; |
| 06:03.56 | brlcad | that'll take care of most of them |
| 06:07.21 | brlcad | again, that's where "svn status" comes in handy, it'll list the modified files and you can revert any that aren't intentional |
| 06:07.38 | pacman87 | that's exactly what i'm doing :) |
| 06:08.12 | brlcad | oh, and that should have been svn revert --recursive src/other |
| 06:08.31 | pacman87 | yeah, i picked that up when nothing happened |
| 06:15.17 | pacman87 | my g_hyp.c has mime-type: text/plain and eol-style: native, is this correct? |
| 06:18.15 | brlcad | yes |
| 06:20.07 | pacman87 | ok, here comes the commit! |
| 06:23.44 | CIA-21 | BRL-CAD: 03pacman87 * r31148 10/brlcad/trunk/ (6 files in 2 dirs): start of code to support hyperboloid of one sheet, finished prep(), print(), shot(), norm(), uv(), and free(); still need to check for div by zero cases |
| 06:23.54 | brlcad | woo hoo |
| 06:23.59 | pacman87 | :D |
| 06:24.14 | brlcad | the start of something great ;) |
| 06:24.38 | pacman87 | thanks for all your help, and sorry to bother you with the trivial stuff |
| 06:24.39 | brlcad | now you can employ the "commit early, commit often" ;) |
| 06:24.58 | brlcad | np, that's what I'm here for |
| 06:27.34 | pacman87 | 1:30am, bedtime |
| 06:27.37 | pacman87 | back tomorrow |
| 06:27.45 | brlcad | cya! |
| 06:27.58 | brlcad | should do the same, hrmph |
| 06:32.11 | CIA-21 | BRL-CAD: 03brlcad * r31149 10/brlcad/trunk/include/bu.h: remove register keywords from public interface |
| 06:40.38 | *** join/#brlcad vedge (n=vedge@205-237-251-204.ilesdelamadeleine.ca) |
| 06:52.55 | *** join/#brlcad alex_joni (n=juve@emc/board-of-directors/alexjoni) |
| 07:36.21 | *** join/#brlcad louipc (n=louipc@76-10-146-181.dsl.teksavvy.com) |
| 09:49.56 | *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6) |
| 09:58.19 | *** join/#brlcad Zenit (n=alex@212.45.226.240) |
| 10:02.40 | *** part/#brlcad Zenit (n=alex@212.45.226.240) |
| 11:51.35 | *** join/#brlcad Elperion (n=Bary@p54876A17.dip.t-dialin.net) |
| 12:36.44 | *** join/#brlcad PrezzKennedy (i=Matthew@74.86.45.130) |
| 12:41.29 | *** join/#brlcad vedge (n=vedge@205-237-251-204.ilesdelamadeleine.ca) |
| 12:48.25 | *** join/#brlcad thing0 (n=ric@123.208.224.99) |
| 13:20.31 | *** join/#brlcad andrecastelo (n=chatzill@189.71.15.11) |
| 13:21.09 | andrecastelo | good morning everyone :} |
| 14:22.46 | pacman87 | morning, andrecastelo |
| 16:33.19 | *** join/#brlcad Elperion (n=Bary@p54876A17.dip.t-dialin.net) |
| 16:40.50 | *** join/#brlcad andrecastelo_ (n=chatzill@189.71.58.243) |
| 17:13.49 | *** join/#brlcad docelic (n=docelic@78.134.205.150) |
| 18:03.03 | poolio | alloo |
| 18:19.45 | PrezKennedy | guten tag |
| 18:23.03 | poolio | heh, it's already the afternoon :) |
| 18:51.35 | alex_joni | day's almost over over here |
| 20:00.00 | *** join/#brlcad cad30 (n=4d778bdc@bz.bzflag.bz) |
| 20:01.09 | *** join/#brlcad cad30 (n=4d778bdc@bz.bzflag.bz) |
| 21:10.00 | brlcad | but but .. the night is just getting started alex_joni |
| 21:30.46 | *** join/#brlcad elite01 (n=elite01@dslc-082-082-095-235.pools.arcor-ip.net) |
| 21:38.06 | *** join/#brlcad andrecastelo__ (n=chatzill@189.71.58.243) |
| 22:06.10 | *** join/#brlcad elite01_ (n=elite01@dslb-088-070-014-041.pools.arcor-ip.net) |
| 22:21.25 | CIA-21 | BRL-CAD: 03pacman87 * r31150 10/brlcad/trunk/src/librt/g_hyp.c: fix div by zero in norm() for g_hyp.c |
| 22:26.48 | CIA-21 | BRL-CAD: 03pacman87 * r31151 10/brlcad/trunk/src/librt/g_hyp.c: note to self: compile first, then check in |
| 22:26.54 | brlcad | :) |
| 22:27.13 | pacman87 | missed the second arg to NEAR_ZERO() |
| 22:42.34 | poolio | pacman87: for divide by zero, wouldn't equality work? Or are you just afraid that really tiny values will lead to way too large results? |
| 22:43.40 | pacman87 | the second one |
| 22:44.06 | pacman87 | i'm using ( x/z, y/z, 1 ) as a vector, then normalizing |
| 22:45.52 | pacman87 | except, of course, when NEAR_ZERO(z) |
| 22:48.43 | poolio | ah fair enough |