| 01:26.28 | *** join/#brlcad cad24 (n=4347bd2c@bz.bzflag.bz) |
| 01:35.08 | starseeker_ | PrezKennedy: heh, no not yet |
| 01:35.37 | starseeker_ | doubts IRC bots have gotten good enough with AI to take over cars, but you never know... |
| 01:56.22 | *** join/#brlcad Twingy (n=justin@74.92.144.217) |
| 03:00.34 | brlcad | starseeker_: doc-book wiki is a project I've been involved with loosely for years -- tis a great project, but unfortunately not a great wiki |
| 03:01.18 | brlcad | the mediawiki and drupal docbook integrations are more interesting since they can hook into what we already have better |
| 03:01.41 | brlcad | and yes, doclifter is the one |
| 03:04.52 | brlcad | starseeker_: and Tkhtml3 should be more than enough to display our stuff, you really shouldn't need a complicated comprehensive html support package |
| 03:05.11 | brlcad | in fact the smaller/simpler the better/easier it will be to maintain and integrate |
| 03:16.19 | brlcad | starseeker_: also, Tkhtml at http://www.hwaci.com/sw/tkhtml/ is not GPL, it's LGPL |
| 03:51.52 | starseeker_ | brlcad: They seem to have both GPL and LGPL statements in their tarball: http://www.hwaci.com/sw/tkhtml/tkhtml.tar.gz |
| 03:51.59 | starseeker_ | brlcad: and the README is saying GPL |
| 03:52.03 | starseeker_ | is confused |
| 03:52.22 | brlcad | well, sources say lgpl, and those are all we'd need |
| 03:52.40 | starseeker_ | which is the better one to look at do you think? |
| 03:53.18 | brlcad | well, it's certainly a lot simpler than tkthml3 |
| 03:53.53 | starseeker_ | nods. |
| 03:54.01 | brlcad | I'd give tkhtml3 a try, see if it's simple to integrate |
| 03:54.11 | brlcad | if not, fall back to the one that you know should be simple |
| 03:54.16 | starseeker_ | Right |
| 03:55.05 | starseeker_ | Don't know if you've had a chance to look - I got above and below working, but at the cost of yet another tree walking routine. |
| 03:55.33 | brlcad | saw the commit |
| 03:56.05 | starseeker_ | is thinking when the tree walker rework comes he'll rip all that out, but leave it for now? |
| 03:56.41 | brlcad | that's the idea for all the walkers eventually |
| 03:56.48 | brlcad | at least most of them |
| 03:59.43 | starseeker_ | nods |
| 04:00.06 | starseeker_ | was actually a good experience - taught me something about how general a really good tree walker will need to be |
| 04:00.30 | starseeker_ | should I rip out the nested_region type? |
| 04:04.21 | brlcad | does the added value outweigh the added complexity and maintenance burden? |
| 04:04.30 | brlcad | if it does, keep it; if not, rip it |
| 04:05.16 | starseeker_ | Added complexity is virtually ziltch for that type. |
| 04:05.34 | starseeker_ | at least, given the need to detect assembly types too ;-) |
| 04:06.15 | starseeker_ | Actually was able to show Mike that option to help him get at a model problem today |
| 04:06.22 | starseeker_ | he seemed to like it |
| 04:06.34 | starseeker_ | even though that wasn't his actual problem ;-) |
| 04:06.55 | brlcad | remember everything that's involved, not just code |
| 04:07.34 | brlcad | documentation over a lifetime, etc |
| 04:08.17 | starseeker_ | nods. I'd say that it's actually much simpler to document that option than it is to explain search -below -type region -type region |
| 04:08.43 | starseeker_ | so in that sense (docs included) it's probably a net savings |
| 04:09.10 | brlcad | and the fact that it breaks the consistency (it's not a type like the others, it's sort of a metatype) |
| 04:09.17 | starseeker_ | so is assembly |
| 04:09.25 | brlcad | yep |
| 04:09.44 | brlcad | that's not reason to keep it, that's reason to do something about both of them ;) |
| 04:10.18 | brlcad | that said, if you think it's really worth it -- then i'll go with that |
| 04:11.09 | starseeker_ | I'm willing to concede that anyone likely to appreciate (or care about) nested regions and assemblies can probably understand examples explaining how to detect them |
| 04:12.01 | starseeker_ | Tell you what - if I do a tutorial-style doc for the search command that includes subsections on spotting assemblies and nested regions, I'll support yanking them as types. |
| 04:12.19 | starseeker_ | 's main concern is to make sure usability is there |
| 04:13.14 | CIA-4 | BRL-CAD: 03brlcad * r32674 10/brlcad/trunk/src/mged/typein.c: make sure creating points with the in command will not result in empty point sets. also make sure they match the number specified at the beginning, not the number provided (which could be larger) |
| 04:13.24 | starseeker_ | reflects on the deafening irony of that concern given MGED's usability status, but still... |
| 04:13.39 | starseeker_ | cheers the point primitive |
| 04:17.42 | CIA-4 | BRL-CAD: 03brlcad * r32675 10/brlcad/trunk/include/ged.h: missing declaration of wdb_stub_cmd for windows (it's used by mged) |
| 04:24.52 | CIA-4 | BRL-CAD: 03brlcad * r32676 10/brlcad/trunk/ (NEWS src/libbu/vls.c): |
| 04:24.52 | CIA-4 | BRL-CAD: fix a bug in bu_vls_gets where a fixed buffer size of BUFSIZE*10 was incorrectly |
| 04:24.52 | CIA-4 | BRL-CAD: returning truncated/short lines. this caused nulls to get inserted during reads |
| 04:24.52 | CIA-4 | BRL-CAD: that in turn caused commands like mged's 'in' command to fail. encountered this |
| 04:24.52 | CIA-4 | BRL-CAD: problem when creating a new pnts object (though it conceivably affects *any* |
| 04:24.55 | CIA-4 | BRL-CAD: mged command. |
| 04:29.22 | brlcad | is impressed to find out that you can actually display half a million points interactively |
| 04:29.38 | starseeker_ | wow |
| 04:29.49 | starseeker_ | that is pretty impressive |
| 04:29.54 | brlcad | at least half a million, haven't seen how high it'll go |
| 04:30.02 | starseeker_ | heh |
| 04:30.14 | starseeker_ | feels pity for brlcad's poor computer |
| 04:34.41 | starseeker_ | brlcad: You might find some interesting stuff here: http://www.mpa-garching.mpg.de/Virgo/data_download.html |
| 04:35.53 | starseeker_ | This sucker is 280 megs, not sure how many points that equates to: http://www.vis.uni-stuttgart.de/eng/research/fields/current/pointclouds/virgo.pcld.gz |
| 04:39.10 | brlcad | even displaying them as spheres works relatively reasonably implying that it should be fine to display upwards of about 10M points "interactively" |
| 04:39.42 | starseeker_ | jeez |
| 04:40.00 | starseeker_ | how many points are we looking at for a full vehicle scan? |
| 04:40.19 | brlcad | nobody has ever done one |
| 04:40.39 | starseeker_ | heh - well, that's one point ;-) |
| 04:40.52 | brlcad | it's not even clear (to me) that it's still anywhere close to being a viable/efficient modeling approach (scanning) |
| 04:41.12 | brlcad | but it is useful for visualization and as reference points for "real modeling" |
| 04:41.28 | starseeker_ | That's my thought |
| 04:41.46 | starseeker_ | Dwayne can follow a point cloud around making primitives for the surface |
| 04:43.58 | starseeker_ | I really don't see how a point cloud can serve as anything but (maybe) a basis for a single mesh |
| 04:44.15 | brlcad | now the cool trick would be to leverage what we did with the vulcan scanner |
| 04:45.24 | brlcad | click a button, select a couple points from the scan, and it'd automatically make primitives for you |
| 04:46.15 | starseeker_ | Hmm. Maybe have it prompt for a user selected list of points to "fit" a user selected primitive type to? |
| 04:46.57 | starseeker_ | or a version of in where it's a option to use the mouse to sample the point cloud and have it print the xyz coords to the in command? |
| 04:49.10 | brlcad | that's sort of what lee and I did |
| 04:49.23 | brlcad | the first one at least |
| 04:49.45 | starseeker_ | first one is probably more useful :-) |
| 04:49.49 | brlcad | the vulcan is a point-sample cmm device, gives you individual points in 3space |
| 04:50.26 | brlcad | so we came up with a system where you could create primitives using a set of conventions |
| 04:51.24 | brlcad | so like for an N-sided "plate", for example, you'd click a point on each corner of the top, and then one point for the depth |
| 04:51.35 | brlcad | and it'd automatically make the corresponding solid |
| 04:51.56 | brlcad | (which in that exaple best fit a sketch and an extrude) |
| 04:52.35 | starseeker_ | nice |
| 04:52.35 | brlcad | if you had a cylindrical object, you could select three points on one of the ends and a fourth point for the length, and it'd make the tgc/rcc |
| 04:53.50 | CIA-4 | BRL-CAD: 03starseeker * r32677 10/brlcad/trunk/src/libged/ (search.1 search.c): |
| 04:53.50 | CIA-4 | BRL-CAD: Now that above and below are working, there is no need for special case logic to |
| 04:53.50 | CIA-4 | BRL-CAD: handle assemblies and nested regions in the type section. Since those are |
| 04:53.50 | CIA-4 | BRL-CAD: likely to be common search cases, include examples of how to use below to find |
| 04:53.51 | CIA-4 | BRL-CAD: nested regions and assemblies have been added to the man page. |
| 04:54.27 | brlcad | there was a dozen or so defined, mged has it now as (the undocumented, as it's pre-alpha) parse_points command |
| 04:54.44 | starseeker_ | sweet |
| 04:54.46 | brlcad | expecting labels and points from a file |
| 04:55.14 | brlcad | which is what src/mged/points is |
| 04:55.32 | starseeker_ | should those live in libged eventually? |
| 05:00.05 | starseeker_ | brlcad: Well, this mortal must seek sleep - thanks for talking me into the type handling change. That does seem cleaner. |
| 05:01.04 | brlcad | i wasn't trying to talk you into one way or the other, just to think about all the tradeoffs |
| 05:01.18 | starseeker_ | either way, good thoughts :-) |
| 05:01.58 | starseeker_ | also realized that with .mgedrc you could simply alias a findnestedregions command to a specific search |
| 05:02.39 | starseeker_ | same for a wide variety of common searches - worth mentioning as a labour saving technique in the main docs. |
| 05:06.16 | starseeker_ | shutdown -h brain |
| 05:13.38 | brlcad | hrm |
| 05:13.57 | brlcad | kicks CIA-4 |
| 05:13.57 | CIA-4 | ow |
| 05:19.50 | CIA-4 | BRL-CAD: 03brlcad * r32681 10/brlcad/trunk/src/librt/namegen.c: beware of c99isms, ws. add to dist. |
| 05:20.08 | CIA-4 | BRL-CAD: 03brlcad * r32682 10/brlcad/trunk/src/librt/Makefile.am: beware of c99isms, ws. add to dist. |
| 05:29.28 | CIA-4 | BRL-CAD: 03brlcad * r32683 10/brlcad/trunk/src/libbu/fnmatch.c: don't call it isblank, don't rely on it being a define either. just roll own. |
| 05:45.34 | CIA-4 | BRL-CAD: 03brlcad * r32684 10/brlcad/trunk/src/mged/ (edsol.c mged.h): dead code cleanup |
| 06:18.58 | CIA-4 | BRL-CAD: 03brlcad * r32685 10/brlcad/trunk/src/libged/search.c: ws |
| 06:24.37 | Ralith | ws? |
| 06:24.59 | CIA-4 | BRL-CAD: 03brlcad * r32686 10/brlcad/trunk/src/libged/search.c: quell warnings and fix a bug with case-insensitive matching (it was always matching case-insensitive). |
| 06:30.01 | CIA-4 | BRL-CAD: 03brlcad * r32687 10/brlcad/trunk/NEWS: (log message trimmed) |
| 06:30.02 | CIA-4 | BRL-CAD: bob finished implementing a preliminary bot-stl exporter that dumps out bots as |
| 06:30.02 | CIA-4 | BRL-CAD: stl data but without evaluating the csg booleans, solidity, normals, etc. this |
| 06:30.02 | CIA-4 | BRL-CAD: is to help make export go much faster when the modeler doesn't care about the |
| 06:30.02 | CIA-4 | BRL-CAD: geometric representation. this is *really* exceptionally bad for analysis |
| 06:30.04 | CIA-4 | BRL-CAD: purposes unless you know exactly what you're doing and the implications; it's |
| 06:30.06 | CIA-4 | BRL-CAD: fine for rendering/visualization purposes though. this addresses sf bug 2106032 |
| 06:49.08 | brlcad | ~ws |
| 06:49.11 | ibot | methinks ws is short for workstation. White Space, or the country code for Western Samoa |
| 07:00.55 | CIA-4 | BRL-CAD: 03brlcad * r32691 10/brlcad/trunk/src/mged/mged.h: bogus declaration. maybe that has something to do with the bizzare crash reported to the sf bug tracker [ 2043925 ] ERROR: NULL bn_tol pointer. |
| 09:00.20 | *** join/#brlcad b0ef (n=b0ef@062016141231.customer.alfanett.no) |
| 09:26.24 | *** join/#brlcad b0ef (n=b0ef@062016141231.customer.alfanett.no) |
| 11:21.06 | *** join/#brlcad elite01 (n=elite01@unaffiliated/elite01) |
| 11:24.11 | *** join/#brlcad mafm (n=mafm@elnet-111.lip.pt) |
| 11:25.53 | mafm | hello |
| 11:29.08 | *** join/#brlcad Elperion (n=Bary@p5B14E2DF.dip.t-dialin.net) |
| 13:10.45 | starseeker_ | brlcad: Whoops. Sorry you had to mess with the ccblank/isblank stuff |
| 13:42.19 | starseeker_ | is really liking doclifter :-) |
| 14:01.43 | *** join/#brlcad elite01 (n=elite01@unaffiliated/elite01) |
| 14:14.24 | *** join/#brlcad starseeker_ (n=CY@c-68-33-217-173.hsd1.md.comcast.net) |
| 14:14.30 | starseeker_ | scowls at xchat |
| 14:36.50 | *** join/#brlcad starseeker_ (n=CY@c-68-33-217-173.hsd1.md.comcast.net) |
| 14:36.52 | starseeker_ | grr |
| 15:27.29 | brlcad | catches th' outgoing tide, arrr! |
| 15:27.50 | brlcad | mornin' mafm |
| 15:29.33 | mafm | hello brlcad |
| 16:30.33 | *** join/#brlcad elite01 (n=elite01@unaffiliated/elite01) |
| 16:36.00 | CIA-4 | BRL-CAD: 03brlcad * r32692 10/brlcad/trunk/configure.ac: emphasize some of the more significant warnings by sleeping longer |
| 16:42.12 | CIA-4 | BRL-CAD: 03brlcad * r32693 10/brlcad/trunk/configure.ac: |
| 16:42.12 | CIA-4 | BRL-CAD: automatically search /usr/X11 for X11 components since AC_PATH_XTRA on Mac OS X |
| 16:42.12 | CIA-4 | BRL-CAD: 10.5 (where R7 is now used) isn't set up to search there yet. this should fix |
| 16:42.12 | CIA-4 | BRL-CAD: the 10.5 mac build that was previously failing in Tk because X11 was getting |
| 16:42.12 | CIA-4 | BRL-CAD: disabled when it couldn't find libXi. |
| 16:53.09 | CIA-4 | BRL-CAD: 03bob1961 * r32694 10/brlcad/trunk/include/ged.h: Added declarations for wdb_search_cmd and wdb_stub_cmd. |
| 18:57.23 | *** join/#brlcad cad71 (n=3ea87bf2@bz.bzflag.bz) |
| 19:22.11 | mafm | well, I'm going home |
| 19:22.39 | mafm | I'm gonna be mostly unreachable for a week, going to a congress in Turkey |
| 19:22.45 | mafm | be happy! |
| 19:27.53 | brlcad | cya mafm, have a great weekend! |
| 19:28.52 | mafm | thanks :) |
| 20:20.53 | CIA-4 | BRL-CAD: 03homovulgaris * r32695 10/brlcad/trunk/src/ (4 files in 2 dirs): |
| 20:20.53 | CIA-4 | BRL-CAD: trying to make the solver start looking at the current values and start |
| 20:20.53 | CIA-4 | BRL-CAD: iteration rather than from the minimum value possible, so as to facilitate |
| 20:20.53 | CIA-4 | BRL-CAD: implicit constraint checks during rt_prep for instance which might be satisfying |
| 20:20.54 | CIA-4 | BRL-CAD: the constraints already (nonfunctional), code cleanup |
| 20:29.16 | ``Erik | shiver me timbers o.O |
| 20:34.20 | prasad_1 | arrrrrrrrr |
| 21:19.33 | CIA-4 | BRL-CAD: 03erikgreenwald * r32696 10/brlcad/trunk/src/adrt/adrt.h: add load format defs |
| 21:20.22 | CIA-4 | BRL-CAD: 03erikgreenwald * r32697 10/brlcad/trunk/src/adrt/slave/load.c: stub out the various load formats |