02:20.41 | CIA-93 | BRL-CAD: 03brlcad * r39801 10/brlcad/trunk/src/librt/primitives/nmg/nmg_tri.c: only bomb if we don't already have a bottom point. |
02:34.21 | CIA-93 | BRL-CAD: 03brlcad * r39802 10/brlcad/trunk/src/librt/primitives/nmg/nmg_tri.c: differentiate the two bombs |
02:36.51 | starseeker | Interesting data point - time to build opennurbs with CMake: 50 seconds. Time with autotools: 3 minutes, 20 seconds |
02:39.14 | Ralith | O.o |
02:39.16 | Ralith | woah |
02:39.36 | Ralith | related: has anybody built BRL-CAD with clang? |
02:40.07 | starseeker | Ralith: I have, almost |
02:40.28 | starseeker | succeeded with a little hybrid action - no major performance differences here |
02:41.07 | Ralith | what about buildtime? |
02:41.32 | starseeker | well, the latest experiment suggests that at least on my machine autotools is dominating in both cases |
02:41.46 | Ralith | that's unexpected |
02:41.51 | Ralith | any idea why? |
02:42.18 | starseeker | running libtool does entail some overhead... could be my particular system also represents a bad case |
02:42.48 | Ralith | well, presumably cmake does something equivalent to libtool |
02:43.08 | starseeker | may take a stab at getting at least a minimal cmake build of BRL-CAD going, just to test this further... |
02:44.20 | starseeker | leave off the fancy stuff and just try to get the pre-existing CMakeList.txt files going |
02:45.14 | starseeker | thanks d_rossberg for his initial work on this - made opennurbs very simple to split into a stand-alone |
03:12.54 | CIA-93 | BRL-CAD: 03brlcad * r39803 10/brlcad/trunk/NEWS: (log message trimmed) |
03:12.54 | CIA-93 | BRL-CAD: john anderson fixed crashes of codes calling into pipe tessellation (e.g. ev |
03:12.54 | CIA-93 | BRL-CAD: command in mged) where a pipe with a bend radius set to exactly one-half of the |
03:12.54 | CIA-93 | BRL-CAD: outside diameter causes a problem. john said that tessellate_pipe_bend() was |
03:12.54 | CIA-93 | BRL-CAD: not able to handle cases where the bend radius equates the outer radius of the |
03:12.55 | CIA-93 | BRL-CAD: pipe because it would cause degenerate faces to be generated.The fix was to add |
03:12.56 | CIA-93 | BRL-CAD: a vertex tree and a vertex array to help in identifying and skiping the |
13:09.48 | *** join/#brlcad Stattrav (~Stattrav@117.192.128.94) |
13:11.40 | *** join/#brlcad parigaudi (~quassel@pd95b7f5e.dip0.t-ipconnect.de) |
13:16.23 | *** join/#brlcad csanyipal (~csanyipal@125-164-85-95.dynamic.stcable.net) |
13:16.29 | csanyipal | Hi, |
13:19.12 | csanyipal | when one in Mged run the 'size 1000' command, whether is then the size of the modellspace determined as a cube with dimensions of 1000x1000x1000 units? |
13:39.21 | brlcad | basically yes |
13:39.33 | csanyipal | thanks! |
13:39.36 | brlcad | the cube is clamped to the width of your window though |
13:39.48 | csanyipal | OK |
13:39.53 | csanyipal | that is OK |
13:40.33 | brlcad | so a long fat window 1000 units across might only show 100 units vertically, and a really tall window might be 10000 vertically (but still 100 units horizontally) |
13:40.41 | csanyipal | so the make command make the new shape in to this modelspace with dimensions of that space. |
13:40.43 | brlcad | bah, 1000 units horizontally |
13:41.05 | brlcad | yes |
13:41.53 | csanyipal | say, 'size 100' 'make sphare.1.s sph' give to me a sphare with radius of 50, yes? |
13:41.54 | brlcad | word of caution, though, that not all primitives fill that modelspace via 'make' |
13:42.00 | brlcad | some are half the space |
13:42.01 | csanyipal | not? |
13:42.21 | csanyipal | and how can I know which one? |
13:42.58 | brlcad | for the basic primitives, you probably won't encounter it |
13:43.36 | csanyipal | OK, thanks! |
13:43.44 | brlcad | unintentional inconsistent behavior, if you run into it -- feel free to report it so they can be adjusted |
13:44.21 | brlcad | I just remember running into the problem a few years back when I made the primitives chart: http://brlcad.org/tmp/primitives/Primitives3.png |
13:44.31 | csanyipal | I'll do that! I must go now. Thanks! |
13:45.03 | brlcad | cya! |
13:45.14 | csanyipal | I'll come back after a three hours. |
13:45.19 | csanyipal | cya! :) |
16:09.23 | *** join/#brlcad Stattrav (~Stattrav@117.192.145.72) |
17:31.22 | *** join/#brlcad mafm (~mafm@83.54.182.208) |
17:39.08 | *** join/#brlcad mafm_ (~mafm@193.153.199.176) |
18:07.12 | *** join/#brlcad mafm (~mafm@83.38.34.51) |
19:13.23 | *** join/#brlcad csanyipal (~csanyipal@125-164-85-95.dynamic.stcable.net) |
19:13.33 | csanyipal | Hi, |
19:15.29 | brlcad | howdy |
19:23.30 | csanyipal | for me is in the BRL-CAD system the more difficult part how to compose from primitives a model; how to know where must go primitives in the 3D space to get a model. |
19:23.53 | csanyipal | is there a guide for this, or howto? |
19:25.08 | csanyipal | using mged certainly.. |
19:25.29 | brlcad | csanyipal: yes, our "principles of effective modeling" document covers more of the rationale and some modeling best practices |
19:25.39 | brlcad | a lot of it comes with experience and training |
19:26.09 | brlcad | the introduction to mged tries to cover some of the basics of picking the right primitives through tutorials |
19:26.23 | brlcad | http://brlcad.org/wiki/Documentation <-- links 2 and 3 |
19:27.23 | brlcad | usually the best summary is to think of the "overall" shape of an object, find a primitive that best fits that volume or shape, and then add/subtract primitives to fill in detail and capture more volume |
19:28.03 | csanyipal | brlcad: thanks! |
19:28.12 | brlcad | that sort of top-down approach is at least the dominant approach used for analysis purposes so that you can model to a maximum level of detail in a constrained amount of time |
19:31.16 | csanyipal | I understand that top-down approach and can imagine primitives that best fit the volume and shape of a model. |
19:34.33 | csanyipal | still i think that that when one have an idea of such primitives for constructing a model, the more difficult job is to locate in 3D space the exact locations of primitives to get the right model. |
19:38.48 | csanyipal | it's like when a child play with building blocks except that that here we can locate a primitives (building blocks) one inside of another. |
19:39.38 | csanyipal | and when this part is finished, than we can add/substract primitives to get the final model. |
19:40.20 | brlcad | nods |
19:40.43 | brlcad | even more advanced, using intersections to obtain some hybrid shape that is no longer just stuff added or removed, but represents that shared space |
19:43.43 | csanyipal | can you give for this an example? I don't quite understand your last sentence. (my poor English, sorry) |
19:44.34 | csanyipal | I know what is intersection (I think) but mayhap I misunderstand here something.. |
19:46.51 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) |
19:47.11 | csanyipal | ah OK, I remember now: say we have two sphere that has a common volumes and intersection is just that common volume, without other parts of spheres. |
20:29.10 | csanyipal | I'm being reading document 'Principles of Effective Modeling'. Cya! |
20:31.25 | *** join/#brlcad mafm (~mafm@100.Red-88-18-68.staticIP.rima-tde.net) |
20:40.10 | *** join/#brlcad Ralith (~ralith@216.162.199.202) |