| 00:01.46 | brlcad | Maloeran: I'll be rather interested in about a week or two to get some data from you |
| 00:02.20 | brlcad | performance numbers |
| 00:03.05 | Twingy | vroom! |
| 00:03.05 | Maloeran | Sure. I'll have volume tracing ( ray bundles or whatever ) done by then, though other optimisations will be missing |
| 00:03.07 | brlcad | have a presentation to give and you're work will be a part of it |
| 00:03.23 | Maloeran | Oh? A presentation on BRL-CAD I assume? |
| 00:03.28 | Twingy | bzfrag |
| 00:03.40 | Twingy | ^-- chinese version |
| 00:03.42 | brlcad | on ray-tracing and brl-cad |
| 00:03.45 | dtidrow_work | lol |
| 00:04.42 | Maloeran | Neat, I'm hoping the new code will actually perform faster than the prototype by then. By the way, if you have any interest in the code, I'm sure it could be arranged |
| 00:05.30 | brlcad | basically how you compare will be the hilight, your approach (non-technical) and how it might be of interest to the audience for V/L |
| 00:06.49 | brlcad | i won't have time to look at the code for this, but it would be interesting to demo it on your behalf |
| 00:07.30 | Maloeran | Okay. I'll try to have it threaded, that would... help for a demo |
| 00:07.32 | Twingy | will there be donuts? |
| 00:07.46 | brlcad | there will be lots of lights |
| 00:07.55 | Twingy | pretty girls? |
| 00:08.02 | brlcad | pretty sure |
| 00:08.06 | Maloeran | Or failing that, there's the old prototype for use |
| 00:08.25 | Twingy | will you have Mr. T giving the introduction? |
| 00:08.25 | brlcad | Twingy: you hear the news about someone? |
| 00:08.28 | Maloeran | May I ask who will attend the presentation? |
| 00:08.34 | brlcad | nope, not going |
| 00:08.41 | Twingy | aww |
| 00:08.46 | brlcad | just me |
| 00:09.14 | Twingy | brlcad, unless I know who that somone is I know not the news you speak of |
| 00:09.36 | brlcad | you know who that someone is |
| 00:09.50 | Twingy | I know not the news |
| 00:15.45 | *** join/#brlcad b0ef (n=b0ef@062016141085.customer.alfanett.no) |
| 00:25.55 | Twingy | little caesars, the ramen noodles of pizza |
| 01:52.10 | dtidrow | lol |
| 01:57.09 | brlcad | mm.. ramen |
| 01:58.35 | dtidrow | btw, the owner of Little Ceasers also owns the Tigers (go Tigers!!!) |
| 01:59.28 | *** join/#brlcad alvin_lee (n=iuri@201-34-233-91.bsace704.dsl.brasiltelecom.net.br) |
| 02:00.33 | alvin_lee | the program used to create the models is the mged? |
| 02:02.34 | Twingy | yes |
| 02:02.52 | Twingy | there is a manual with tutorials on the site |
| 02:03.27 | brlcad | dtidrow: heh, curious |
| 02:03.38 | brlcad | do they serve the pizza at the stadium? :) |
| 02:03.55 | dtidrow | probably - haven't been to the new one |
| 02:03.59 | brlcad | mm.. "the mged" |
| 02:04.25 | brlcad | that gives almost a sound of respectability to it |
| 02:04.44 | Twingy | mjed |
| 02:04.48 | brlcad | actually it fits with the original expanded name :) |
| 02:04.57 | brlcad | the multi-device... |
| 02:06.44 | dtidrow | your car doesn't have a cassette player? |
| 02:06.51 | brlcad | nope |
| 02:07.01 | brlcad | CD player, no input jack |
| 02:07.18 | dtidrow | new ones do |
| 02:07.32 | brlcad | yeah, well mine's not new :) |
| 02:07.44 | dtidrow | my wife's is a n '06, and it has an input jack |
| 02:07.47 | brlcad | and the connectors have been prevalent for more than a dcade |
| 02:07.56 | brlcad | s/dcade/decade even/ |
| 02:08.31 | brlcad | as have the proliferation of little audio devices |
| 02:09.18 | brlcad | hmm.. monster's looks good |
| 02:09.21 | brlcad | http://www.monstercable.com/computer/productPageComputer.asp?pin=2084&LastPage=Apple%20Products |
| 02:09.56 | dtidrow | yeah, you would think that they would have started to do that ever since portable cd players came out |
| 02:10.48 | brlcad | yep |
| 02:11.56 | dtidrow | oh well, glad I got the cd/cassette combo with my car |
| 02:12.09 | brlcad | :) |
| 02:12.33 | dtidrow | I just stuff a cassette adapter into the slot and plug in the music |
| 02:12.39 | dtidrow | and the XM radio |
| 02:13.38 | brlcad | unless it's a road trip |
| 02:13.56 | brlcad | which is why this is all the sudden of interest, trip up to boston tomorrow |
| 02:14.20 | dtidrow | ah |
| 02:16.45 | Maloeran | Mmhm. mcarp didn't quite have a great time with mged, I guess he's more into artistic rather than engineering modelling. Unfortunate |
| 02:18.27 | brlcad | not too surprising |
| 02:18.38 | brlcad | it's the least forgiving and most painful for new users |
| 02:18.48 | dtidrow | heh |
| 02:19.17 | brlcad | it starts out horrible and then slowly gets (much) better, but that's not an easy path for the nimble |
| 02:19.56 | brlcad | which is really a shame, because most of the underlying power is considerably better (or at least more extensive) than mged itself |
| 02:20.18 | Maloeran | I haven't really tried it ; from what I know, he found it bothersome to use commands for about everything |
| 02:20.48 | Maloeran | Perhaps there's room for improvement in the user interface then? |
| 02:20.51 | brlcad | that's actually one of the things I DO like most about it, the command line interface |
| 02:21.00 | brlcad | heh, room for improvement |
| 02:21.04 | brlcad | that's an understatement |
| 02:21.08 | Maloeran | :} |
| 02:21.13 | brlcad | it's the foundation of the new modeler environment |
| 02:25.20 | Maloeran | That conference room is a terrible raytracing benchmark, it's all cube-based and axis-aligned, large triangles to quickly fill up |
| 02:26.08 | brlcad | it's a semi-realistic environment, though as it's a "real place" |
| 02:26.18 | brlcad | as opposed to some object in a void or in a box |
| 02:26.48 | Maloeran | At least rotate the room by 30 degrees on two axis then |
| 02:27.25 | brlcad | a particular view at a teapot sitting inside the powerplant model sitting on some terrain would make for an interesting model ;) |
| 02:28.55 | Maloeran | Eheh. The two models I mentionned above are very difficult for acceleration structures, hence why I would like to see kd-trees on these |
| 02:29.32 | Maloeran | truck_bots are long thin triangles that are impossible to split up for binary trees, the frigate has ropes going everywhere |
| 02:30.02 | Maloeran | truck_bots has* long |
| 02:30.52 | brlcad | oh speaking of mcarp, another idea came to mind if he's entirely put off on using mged |
| 02:31.09 | Maloeran | Can you parse .ASE ? Model is 600mb of that text format |
| 02:31.10 | *** part/#brlcad alvin_lee (n=iuri@201-34-233-91.bsace704.dsl.brasiltelecom.net.br) |
| 02:31.16 | Maloeran | What is it? |
| 02:31.42 | brlcad | he could use something like povray, so long as it was still primitives and csg |
| 02:32.10 | brlcad | or another cad package that could dump out IGES with nurbs objects perhaps "maybe" |
| 02:32.11 | Maloeran | I'll pass on the message |
| 02:32.18 | brlcad | but povray would be easy |
| 02:32.33 | brlcad | there are gui modelers for it too, various quality |
| 02:34.03 | Maloeran | I don't think he minds CSG, but the tools he uses couldn't natively export it and ends up converting to triangles |
| 02:34.41 | brlcad | right, and that's a problem for the specific reasons I wanted it |
| 02:35.13 | brlcad | (matching a historical reference) |
| 02:36.14 | brlcad | i really should get started on that geometry website soon |
| 02:36.45 | Maloeran | A repository of geometry? |
| 02:36.52 | brlcad | i'm curious how well it would take off with the right interface, whether the community really would jump in and start bringing the models together |
| 02:37.08 | brlcad | yeah |
| 02:37.21 | Maloeran | CSG or about anything? |
| 02:37.25 | brlcad | anything |
| 02:37.49 | Maloeran | Sounds good |
| 02:39.05 | brlcad | basically a database of models, with details about the content, geometry type characteristics (solid, csg, brep, hybrid, etc), statistics (model size, number of objects, polys, etc) |
| 02:40.15 | brlcad | automatically convert models to various formats, keeping track of versions and validity of the information like whether this particular triangle teapot is the "original" or whether this brep one over here is, etc |
| 02:40.40 | brlcad | rather straightforward to do, more just a matter of grunt work to set up the infrastructure |
| 02:40.54 | brlcad | and the design and appeal/usability of the site itself |
| 02:41.10 | Maloeran | I'll adopt the BRL-CAD conversion tools one day, I went the pseudo-lazy way and wrote an ASE parser ( just text ) |
| 02:42.04 | brlcad | or maybe you'll write a brl-cad conversion tool one day ;) |
| 02:42.20 | brlcad | like an ase-g importer or an g-ase expoerter ;) |
| 02:42.48 | Maloeran | If you care about ASE, sure :) |
| 02:44.20 | Twingy | not to be confused with ASS |
| 02:44.22 | brlcad | brl-cad already has more importers and exporters than just about any other open source modeling system (though blender is close, but they don't do any of the hard ones) |
| 02:45.12 | Maloeran | Blender's exporters and importers were terribly broken for me |
| 02:45.18 | brlcad | it wouldn't take much to make a fairly universal converter library and app (ala image magic's "convert") |
| 02:45.22 | Maloeran | So were lib3ds, libase, and a couple other things |
| 02:46.15 | Maloeran | I don't easily trust any other code than my own now, due to such... historical reasons. I just write what I need |
| 02:46.42 | brlcad | step is the format I really want to see implemented |
| 02:47.42 | brlcad | it's a horribly complex format, but adopted by every one of the major commercial systems |
| 02:48.02 | brlcad | and is about as faithfully representing as you can get |
| 02:48.39 | brlcad | since it's pretty much the union of all possible CAD/CAM/modeling features you might have wanted in your own format |
| 02:49.50 | Maloeran | Interesting, never heard of it |
| 02:51.07 | brlcad | if you're not in CAD/CAM or otherwise spending $20k on a CAD system, you probably wouldn't |
| 02:51.23 | brlcad | it's the successor to IGES |
| 02:51.44 | Twingy | then there was GCAM |
| 02:51.47 | brlcad | which was the definitive CAD format before it |
| 02:51.52 | brlcad | heh |
| 02:52.36 | Twingy | an italian grad student donated the other day |
| 02:53.15 | brlcad | Maloeran: other than .g, have you dealt with *any* solid modeling or CAD format that you know of? :) |
| 02:53.46 | Maloeran | I used Autocad 13 and 14 for years younger, helping my father's engineering work, if that counts for anything |
| 02:53.48 | brlcad | probably mostly dealt with general modeling formats like stl, dxf, vrml, etc |
| 02:54.16 | Maloeran | Otherwise, never dealt with anything CAD related :) |
| 02:56.16 | Maloeran | I suppose step and Iges support about all primitives one can think of, CSG or otherwise? |
| 02:56.33 | brlcad | pretty much |
| 02:56.57 | brlcad | as well as things like constraints, parametrics |
| 02:58.07 | brlcad | construction history, tolerance limits (per part or global), validation and verification |
| 02:59.55 | brlcad | formats like iges, step, jt, parasolid's x_t/x_b, vda, brl-cad's format (to a lesser extent for the advanced features) |
| 03:00.57 | brlcad | other details the geometry file might contain is whether the model is actually just a wireframe, or it's polygons, or it's a solid, or just a surface model, is there connectivity, etc |
| 03:01.18 | brlcad | dxf probably belongs in that mix too |
| 03:01.44 | Maloeran | Quite nice, sounds like troublesome to fully parse too |
| 03:02.03 | brlcad | depends on the format and the specification |
| 03:02.22 | brlcad | parsing is often the easiest part |
| 03:03.01 | brlcad | it's harder to actually convert.. especially when their representation is something you don't support, or the model doesn't have something you need |
| 03:03.28 | Maloeran | Yes that's what I meant, handling all the possible kinds of geometry representations |
| 03:03.39 | Maloeran | representation* |
| 03:03.54 | brlcad | like reading in nurbs if you only supported triangles in your format... or only supporting solid models, yet the model you're converting has degenericies |
| 03:04.00 | brlcad | yeah |
| 03:04.52 | brlcad | that's where brl-cad does pretty good as we intrinsicly support just almost everything commonly in use |
| 03:05.55 | brlcad | we lack superquadrics and have poor support for 2D primitives and don't really do trimmed nurbs (but do do untrimmed ones) |
| 03:06.20 | brlcad | at least not yet |
| 03:06.56 | Twingy | heh, you said do do |
| 03:07.37 | brlcad | re re me me so so la la si si do do |
| 03:08.16 | Maloeran | Do ré mi fa sol la si do in french, I never heard that in English |
| 03:08.23 | Maloeran | Ahah |
| 03:08.44 | brlcad | Maloeran: it's the same |
| 03:10.51 | brlcad | i don't think that'd be very comfortable to ride on |
| 03:11.03 | Twingy | you can keep it in my garage if you need to |
| 03:11.04 | brlcad | or go very fast |
| 03:11.33 | Twingy | if one of the strings break I have some 10AWG wire left over from solar installation |
| 03:12.39 | brlcad | hey, there's even a wiki page on it whowuddathunk http://en.wikipedia.org/wiki/CAD_data_exchange |
| 03:13.05 | brlcad | mm.. 10AWG |
| 03:13.28 | brlcad | sounds like it'd be useful.. FOR A LYNCHING! |
| 03:13.30 | brlcad | *ahem* |
| 03:13.42 | Twingy | I'm real curious if I can melt copper |
| 03:13.54 | brlcad | with your bare hands |
| 03:13.56 | Twingy | I almost have enough scrap to try |
| 03:14.51 | Twingy | even if I could get it hot enough I doubt I could keep more than an ounce hot enough |
| 03:19.24 | Twingy | GCAM just about hit the 9k line mark, roughly 25% the size nurbana was |
| 03:25.23 | Twingy | ok, I'm not sure how that relates to what I just said... |
| 03:29.10 | Maloeran | It probably doesn't :). Anything missing in gcam besides 3d view and GUI candy? |
| 03:31.26 | Twingy | I've got 3d view and I don't think GUI candy is appropriate for gcam |
| 03:32.25 | Twingy | there's a number of features that still remain, but it's moving at a healthy pace, just check out its bugzilla |
| 03:32.43 | Twingy | I'm averaging a feature just about every other night |
| 03:32.51 | Twingy | finishing up pocketing now |
| 16:23.34 | Maloeran | Erik, I'll be going to the conference finally ( Beatrice from Survice was sick hence the long delay in response ) |
| 17:10.45 | ``Erik | ah, cool |
| 19:45.44 | CIA-5 | BRL-CAD: 03erikgreenwald * 10brlcad/src/librt/g_metaball.c: More print capability. Fixed missing magic in st_specific. |
| 20:11.11 | Maloeran | Metaballs, now with more magic! |
| 20:57.58 | CIA-5 | BRL-CAD: 03erikgreenwald * 10brlcad/src/mged/edsol.c: some (not very stable or complete) metaball editing |
| 21:40.48 | *** join/#brlcad Twingy (n=justin@74.92.144.217) |
| 21:50.07 | ``Erik | heh |
| 21:54.56 | ``Erik | fear my magic balls. |
| 21:55.39 | *** join/#brlcad Twingy (n=justin@74.92.144.217) |
| 21:58.24 | ``Erik | heh, yeah, that's it, kill and scramble braincells, that'll help you solve a difficult problem... |
| 21:58.29 | Maloeran | The basic idea is to try to reunite a volume later on even if it appears to split up, through the mess of sectors and nodes |
| 21:59.05 | Maloeran | Oh the problem has no physical substance, banging one's head against it is purely metaphorical :) |
| 21:59.11 | ``Erik | you can't extrapolate that from the knowlege of neighbors? |
| 21:59.18 | ``Erik | BANG YOUR HEAD! METAL HEALTH WILL DRIVE YOU MAD! |
| 22:00.20 | Maloeran | I can't extrapolate anything, I must walk the quad volume and compute on the way if no sector is contained without being intersected by one of the four rays |
| 22:00.40 | Maloeran | And do it efficiently, that is |
| 22:01.25 | ``Erik | uhmmmm |
| 22:01.36 | ``Erik | uhh, crap, I left that booklet at work |
| 22:01.59 | Maloeran | Simple cases are easy, but I want something that will be as aggressive as possible, to avoid processing smaller bundles or rays as long as possible |
| 22:02.02 | ``Erik | in your rt06 poster book, look for the poster that japanese dudes did... I *THINK* what your'e talking about is what they solved? |
| 22:02.28 | Maloeran | Hum. Looking for the poster book |
| 22:02.43 | Maloeran | No one is using sectors out there, I doubt anyone solved that. It's very different with kd-trees |
| 22:02.44 | ``Erik | you remember the one I'm talking about, right? the one next to mine? |
| 22:03.13 | ``Erik | hm, *shrug* they may have information that might be useful... *shrug* research is good. |
| 22:05.52 | Maloeran | That's basic stuff, and they aren't aggressive in their traversal at all |
| 22:06.11 | Maloeran | Hum, and Reschetov did better in 2005 |
| 22:06.38 | ``Erik | reschetov does some cool stuff |
| 22:07.06 | Maloeran | Yes, I quite like the guy ; he writes papers when he has something to say... unlike Wald |
| 22:07.24 | ``Erik | what he has to say is also worth listening to, unlike many |
| 22:08.10 | Maloeran | Right. I'm really not fond of this Wald character, the impression I have is that he's more interested by attention and praise than communicating anything of value |
| 22:11.24 | ``Erik | not uncommon, unfortunately |
| 22:51.46 | *** join/#brlcad SWPadnos_ (n=Me@dsl245.esjtvtli.sover.net) |
| 23:02.46 | *** join/#brlcad SWPadnos_ (n=Me@dsl245.esjtvtli.sover.net) |