| 00:00.12 | ``Erik | there, ah went an' dun sed it |
| 00:02.04 | CIA-28 | BRL-CAD: 03starseeker * r36034 10/brlcad/trunk/ (5 files in 4 dirs): Add an extrusion brep routine, now that sketch is working. |
| 00:02.10 | ``Erik | I might resurrent that "simple timer" thingie, and mebbe have it so any programming->else->programming moves 30m from programming to else to compensate for wind/unwind O.o |
| 00:50.18 | brlcad | even with time over-represented (which was rather intentional), that just means all the things that can be done really quickly that we think will have a decent payoff will get done up-front |
| 00:50.30 | ``Erik | aight |
| 00:50.33 | ``Erik | *shrug* |
| 00:50.54 | ``Erik | I'm just nervous about... expectation management? |
| 00:51.02 | brlcad | part of the point is to minimize how far in advance we're having to estimate time .. even if we get "less than a week" wrong .. it's still probably gonna get done |
| 00:52.09 | ``Erik | "technical debt", too |
| 00:52.32 | ``Erik | <-- just noting his concerns |
| 00:52.38 | brlcad | nods |
| 00:53.17 | ``Erik | I think I've been hooked with 2 things this cycle :/ the surprise bug from russ is... concerning |
| 00:53.33 | brlcad | 36% is completely arbitrary and actually unimportant -- could have picked any random number out of our ass between 1 and 28 |
| 00:53.36 | ``Erik | (why the hell didn't they msg me earlier? are they talking about the scalloping bug I fixed?) |
| 00:54.02 | brlcad | yeah, not enough info there |
| 00:54.05 | brlcad | probably |
| 00:54.16 | brlcad | his info does tend to be a quarter late when it comes to our bugs |
| 00:55.31 | brlcad | the burn rate follows our actual effort, so it will auto-adjust even if our estimates are always wrong |
| 00:56.01 | CIA-28 | BRL-CAD: 03erikgreenwald * r36035 10/brlcad/trunk/NEWS: mention scalloping bug in metaballs. |
| 00:57.07 | brlcad | if 10 units of work take 20 units of time, our estimates are low; if we only complete 5 units of work, our burn rate needs to be lowered |
| 00:57.41 | ``Erik | I wonder if the overhead of a context switch should be called out explicitely |
| 00:57.42 | brlcad | the two then converge over time as estimates are adjusted and the burn rate follows our capacity |
| 01:00.08 | ``Erik | *shrug* we'll give it a whack, as it goes, our different impressions will be expressed until it normalizes, we'll see what happens :) ain't no use spending more than 10 seconds mentioning a concern until we're a few cycles deep, just wanted to throw my view out nice and early |
| 01:01.15 | starseeker | growls - looks like revolve is going to take a slightly different version of the sketch code |
| 01:01.23 | starseeker | figures |
| 01:01.38 | ``Erik | starseeker: relax, step away, work on something like, say, fractal quaternions instead :D |
| 01:03.22 | starseeker | ``Erik: heh |
| 01:03.55 | ``Erik | you disagree? that was an awfully cool article :) |
| 01:06.03 | starseeker | indeed :-) |
| 01:06.21 | starseeker | isn't sure what he would work on on that though |
| 01:18.04 | ``Erik | <-- not sure how it'd related to BRL-CAD, just saw the article on hn, thought it was awesome, had raytracing AND quaternions... AND fractals... AND pretty pictures |
| 01:23.05 | starseeker | might make a very interesting debugging tool... |
| 01:24.23 | ``Erik | debugging, or stress test? |
| 01:26.06 | starseeker | heh - probably both |
| 01:26.57 | CIA-28 | BRL-CAD: 03starseeker * r36036 10/brlcad/trunk/ (5 files in 4 dirs): Not fully worked out yet, but this does do a revolution of sorts. |
| 01:28.28 | starseeker | ok, that'll do for tonight |
| 01:28.55 | starseeker | reflects that revolve isn't fully "done" as a supported primitive anyway... |
| 01:29.20 | starseeker | that leaves pipe, dsp, hyp and part |
| 01:31.16 | CIA-28 | BRL-CAD: 03erikgreenwald * r36037 10/isst/trunk/src/load_g.c: gtk.h is no longer required for isst.h |
| 01:31.30 | starseeker | raises his eyebrows |
| 01:31.32 | starseeker | nice |
| 01:31.54 | ``Erik | shut up |
| 01:31.55 | ``Erik | :D |
| 01:32.16 | brlcad | heh |
| 01:33.09 | starseeker | decides he'd better hit the road... |
| 01:34.30 | ``Erik | at vodka thirty? eh? |
| 01:37.11 | starseeker | heh |
| 01:37.55 | starseeker | ok, lock and load (doors and cars that is...) |
| 01:39.28 | ``Erik | hm |
| 01:52.54 | ``Erik | sunny beaches. |
| 01:53.17 | ``Erik | why won't this sneeze happen? |
| 01:59.00 | ``Erik | so, uh, yeah |
| 01:59.08 | ``Erik | ur mom, r0x ur s0x 0ff |
| 02:00.00 | CIA-28 | BRL-CAD: 03erikgreenwald * r36038 10/isst/trunk/src/ (Makefile.am load_g.c): this is now a lib func |
| 02:01.26 | ``Erik | stupid effin' rejection. :/ |
| 02:01.55 | ``Erik | yet more ways svn sucks. |
| 02:03.10 | ``Erik | blank, ready prompt... that is NOT the right way to signal error! |
| 03:18.42 | *** join/#brlcad PrezKennedyII (i=Matthew@whitecalf.net) |
| 10:31.17 | CIA-28 | BRL-CAD: 03brlcad * r36039 10/brlcad/trunk/src/libged/qray.c: another instance where not being initialized is apparently okay. seems like wrapper material. |
| 10:47.55 | CIA-28 | BRL-CAD: 03brlcad * r36040 10/brlcad/trunk/BUGS: the new mged -a attach option at least provides a work-around to the python script prompting on windows. remove until we know more |
| 10:52.27 | CIA-28 | BRL-CAD: 03erikgreenwald * r36041 10/brlcad/trunk/src/adrt/ (11 files in 3 dirs): migrate isst load_g in, and pull in tie.h as needed. |
| 10:53.09 | CIA-28 | BRL-CAD: 03brlcad * r36042 10/brlcad/trunk/ (BUGS TODO): technically, nmg raytracing so slowly was confirmed to not be a bug, just a really inefficient implementation. remove from bugs and make it a todo item to fix it. |
| 10:56.33 | CIA-28 | BRL-CAD: 03brlcad * r36043 10/brlcad/trunk/BUGS: bob fixed the problem with contextualized objects not working right with various mged commands and rt. |
| 11:05.35 | CIA-28 | BRL-CAD: 03brlcad * r36044 10/brlcad/trunk/BUGS: malloc and friends no longer take an unsigned int, now taking a size_t. soon most of the rest of librt will be following suit. |
| 11:16.03 | CIA-28 | BRL-CAD: 03erikgreenwald * r36045 10/brlcad/trunk/src/adrt/ (6 files in 2 dirs): shift file loading into the library |
| 11:25.44 | CIA-28 | BRL-CAD: 03brlcad * r36046 10/brlcad/trunk/TODO: |
| 11:25.44 | CIA-28 | BRL-CAD: whoosh. massive update to remove items that are either now complete or no |
| 11:25.44 | CIA-28 | BRL-CAD: longer relevant. in brief, includes pnts input, nurb bb, humans, adrt |
| 11:25.44 | CIA-28 | BRL-CAD: shotlines, primitive-to-breps, news2tracker, obj-g, de-mysql, isst callbacks, |
| 11:25.44 | CIA-28 | BRL-CAD: bot scripts, geometry syndicator, rtgl, and coils. |
| 11:31.48 | CIA-28 | BRL-CAD: 03brlcad * r36047 10/brlcad/trunk/TODO: promote/demote next iteration items. some missing, but accounting for mine (refactoring librt, warnings, nmg->brep, and annotations) |
| 11:32.52 | ``Erik | ponders renaming src/adrt to src/libadrt, librender.so to libadrt.so, and moving adrt_master and adrt_slave into src/remrt |
| 11:51.21 | CIA-28 | BRL-CAD: 03brlcad * r36048 10/brlcad/trunk/TODO: |
| 11:51.21 | CIA-28 | BRL-CAD: add a slew of missing todo items, pulled from recent discussions and tracker |
| 11:51.21 | CIA-28 | BRL-CAD: items. included is 64-bit port to windows, finishing obj-g, gui for repairing |
| 11:51.21 | CIA-28 | BRL-CAD: bot normals, machine independent dbupgrade, converting manpages to docbook, |
| 11:51.21 | CIA-28 | BRL-CAD: searchable help gui in mged, support to mv for rename mapping files, options to |
| 11:51.24 | CIA-28 | BRL-CAD: bot_dump for normals, push/xpush stopping at the region level, consolidating |
| 11:51.26 | CIA-28 | BRL-CAD: them, working on rtarea, and sca with three args. |
| 11:51.53 | brlcad | and with that, I think the docs are up-to-date again |
| 11:53.51 | brlcad | don't like the mixing with remrt part .. that's a specific tool with a legacy of its own |
| 11:54.16 | ``Erik | my intent is to smear adrt into the fabric of BRL-CAD |
| 11:54.26 | brlcad | src/libadrt and src/adrt wouldn't be too bad, but .. seems to be little gain |
| 11:55.06 | ``Erik | there is no technical gain, it's a matter of realigning language so there can be more integration down the road |
| 11:55.31 | ``Erik | I want it to become a part of BRL-CAD instead of a project that just happens to live in BRL-CAD |
| 11:55.56 | brlcad | making it work with unconverted .g files does that |
| 11:56.36 | brlcad | still doesn't mean you have to kick remrt in the nads to make that happen too |
| 11:56.40 | ``Erik | I'd disagree, it still feels like a second class citizen in the tree to me |
| 11:57.01 | ``Erik | remrt needs kicked in the nads anyways, but for other reasons O:-) I'll not touch that dir |
| 11:57.25 | ``Erik | find some pants and heads to work, maybe in that order O.o |
| 11:57.45 | brlcad | disagree? it still doesn't work with *unconverted* .g files afair, no? |
| 11:58.02 | brlcad | only botsies |
| 11:59.21 | brlcad | breaking out a libadrt that librt uses for bot ray-tracing would be another useful integration path (i.e., with technical gain) |
| 11:59.54 | brlcad | then adrt dir could just become an isst dir or an adrt dir with isst and rise tools but no libs |
| 12:22.18 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-42.sbndin.btas.verizon.net) |
| 12:25.47 | *** join/#brlcad parigaudi (n=quassel@pd95b7f5e.dip0.t-ipconnect.de) |
| 12:32.51 | *** join/#brlcad d_rossberg (n=rossberg@bz.bzflag.bz) |
| 12:49.50 | ``Erik | it works fine with unconverted files, it does the nmg tesselate on teh fly |
| 12:49.58 | ``Erik | that's why it takes 4 hours to load up a BIG file |
| 12:50.09 | ``Erik | that model I demo'd wasn't 100% bot |
| 12:51.10 | ``Erik | my usual test case is the ktank straight from the repo, no preprocessing :D |
| 12:52.59 | ``Erik | and libtie would be the minimal set to replace bot shootray |
| 13:23.45 | *** join/#brlcad Elrohir (n=kvirc@p5B14EF3B.dip.t-dialin.net) |
| 14:19.38 | brlcad | heh, that's not working with "unconverted" exactly .. just because it converts them on the "fly" (for extended definitions of flying) |
| 14:20.39 | brlcad | having it work as-is (sure, reduced speeds for non-tessellated) with just an option to tessellate or not |
| 14:21.23 | brlcad | making librt api additions to support the way it bundles shots, hook in libtie/libadrt into librt for bots, then hooking in librt to isst |
| 14:21.34 | brlcad | THAT would be sweet integration blissness |
| 14:35.46 | ``Erik | yes, replacing bot stuff with tie stuff is a long term goal :) the "smearing it in to the fabric" is a step in that direction |
| 14:36.00 | ``Erik | when I'm ready for that leap, I'll probably branch |
| 14:36.15 | ``Erik | as it changes big things in the critical path for other projects |
| 14:53.34 | *** join/#brlcad samrose (n=samrose@adsl-99-147-180-206.dsl.lgtpmi.sbcglobal.net) |
| 15:28.14 | *** join/#brlcad jdoliner (n=jdoliner@c-67-173-0-29.hsd1.il.comcast.net) |
| 15:39.54 | CIA-28 | BRL-CAD: 03starseeker * r36049 10/brlcad/trunk/src/proc-db/csgbrep.cpp: Add in test pipe example. |
| 16:52.44 | CIA-28 | BRL-CAD: 03brlcad * r36051 10/brlcad/trunk/src/mged/attach.c: allow 'attach nu' to work, just does nothing. |
| 17:06.01 | CIA-28 | BRL-CAD: 03starseeker * r36052 10/brlcad/trunk/src/librt/primitives/sketch/sketch_brep.cpp: Remove the sketch debugging printout. |
| 17:06.58 | brlcad | and so it begins. |
| 17:12.37 | CIA-28 | BRL-CAD: 03brlcad * r36053 10/brlcad/branches/rel8/: (log message trimmed) |
| 17:12.37 | CIA-28 | BRL-CAD: And so begins preparations for the mighty "Release 8"... |
| 17:12.37 | CIA-28 | BRL-CAD: As implied, this version will represent a major incompatibility (i.e., v6 |
| 17:12.37 | CIA-28 | BRL-CAD: database format), signficant features, and/or other big changes. Already |
| 17:12.37 | CIA-28 | BRL-CAD: expected are changes to the database format to repair mistakes and expand |
| 17:12.38 | CIA-28 | BRL-CAD: support for more advanced persistence. Also included and probably first to get |
| 17:12.40 | CIA-28 | BRL-CAD: applied are a slew of signature changes to several of the core libraries (e.g., |
| 17:12.42 | CIA-28 | BRL-CAD: 03brlcad * r36050 10/brlcad/trunk/src/mged/mged.c: fix the new display manager attaching so it's only called once and actually works. simplify a bit. |
| 17:52.23 | CIA-28 | BRL-CAD: 03starseeker * r36054 10/brlcad/trunk/src/librt/primitives/pipe/pipe_brep.cpp: OK, let's take this a step at a time. This appears to generate valid endcap curves. |
| 17:59.28 | *** join/#brlcad mafm (n=mafm@2.Red-83-63-197.staticIP.rima-tde.net) |
| 18:32.00 | CIA-28 | BRL-CAD: 03starseeker * r36055 10/brlcad/trunk/src/librt/primitives/pipe/pipe_brep.cpp: Well, at least it generates a ruled surface now... |
| 18:36.49 | CIA-28 | BRL-CAD: 03starseeker * r36056 10/brlcad/trunk/src/librt/primitives/pipe/pipe_brep.cpp: Remove the saved_pt testing hack - generating weird stuff, but it's generating SOMETHING now so enable it. |
| 18:37.40 | *** join/#brlcad Ralith (n=ralith@69.90.48.127) |
| 19:26.49 | CIA-28 | BRL-CAD: 03starseeker * r36057 10/brlcad/trunk/src/librt/primitives/pipe/pipe_brep.cpp: Well, closer, but one of the ruled surfaces is twisted. arrrgh. |
| 20:26.51 | CIA-28 | BRL-CAD: 03starseeker * r36058 10/brlcad/trunk/src/librt/primitives/pipe/pipe_brep.cpp: Better - can handle a simple linear pipe. |
| 20:32.57 | *** join/#brlcad _sushi_ (n=_sushi_@84-72-9-202.dclient.hispeed.ch) |
| 20:33.54 | starseeker | you know, the general case of pipe will actually require the full surface eval routines - pipe can self-intersect |
| 20:34.25 | starseeker | declares victory sufficient for the time being and moves on to dsp |
| 20:59.21 | starseeker | goggles at dsp, decides that's too complex to start on a Friday afternoon, and hits hyp instead |
| 21:33.18 | CIA-28 | BRL-CAD: 03starseeker * r36059 10/brlcad/trunk/ (4 files in 3 dirs): Untested and doesn't scale the main surface to be elliptical, but commit basic hyp code. |
| 21:39.54 | *** join/#brlcad Ralith (n=ralith@69.90.48.127) |
| 21:45.17 | ``Erik | heh |
| 21:54.16 | *** join/#brlcad cpc26 (n=cpc26@72.170.156.242) |
| 22:25.43 | *** join/#brlcad Patmcc19 (n=chatzill@97-117-150-47.phnx.qwest.net) |
| 22:44.38 | *** join/#brlcad samrose (n=samrose@c-24-11-185-57.hsd1.mi.comcast.net) |
| 23:31.23 | CIA-28 | BRL-CAD: 03erikgreenwald * r36060 10/brlcad/trunk/src/adrt/ (Makefile.am master/main.c master/master.c slave/slave.c): eliminate the "main" files |
| 23:32.51 | CIA-28 | BRL-CAD: 03erikgreenwald * r36061 10/brlcad/trunk/src/adrt/slave/smain.c: disappears. |
| 23:51.09 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-42.sbndin.btas.verizon.net) |
| 23:51.20 | *** part/#brlcad jdoliner (n=jdoliner@c-67-173-0-29.hsd1.il.comcast.net) |