| 01:02.44 | *** join/#brlcad archivist (~archivist@host81-149-189-98.in-addr.btopenworld.com) |
| 06:15.57 | *** join/#brlcad ChanServ (ChanServ@services.) |
| 06:15.57 | *** mode/#brlcad [+o ChanServ] by leguin.freenode.net |
| 06:21.37 | *** join/#brlcad louipc (~louipc@archlinux/fellow/louipc) |
| 12:10.28 | *** join/#brlcad andrei (~andrei@188.25.162.226) |
| 12:34.21 | *** join/#brlcad Al_Da_Best (Al_Da_Best@cpc2-shep12-2-0-cust21.8-3.cable.virginmedia.com) |
| 14:04.53 | Notify | 03BRL-CAD:brlcad * 53812 (brlcad/trunk/src/librt/primitives/half/half.c): comment the peculiarity I had to research. ensures segment is valid, equiv to >inf || <-inf |
| 15:35.57 | *** join/#brlcad jarray52 (~bigbear@unaffiliated/jarray52) |
| 16:18.52 | *** part/#brlcad jarray52 (~bigbear@unaffiliated/jarray52) |
| 17:15.14 | *** join/#brlcad archivist (~archivist@host81-149-189-98.in-addr.btopenworld.com) |
| 17:55.07 | *** join/#brlcad jarray52 (~bigbear@209.203.86.178) |
| 17:55.27 | *** join/#brlcad jarray52 (~bigbear@unaffiliated/jarray52) |
| 17:56.58 | jarray52 | expr (2 + 2) gives 4 at the mged prompt. Similarly, expr (2 - 2) gives 0, and expr (2 / 2) gives 1. However, expr (2 * 2) states Error: unbalanced open paren in expression "(2". |
| 17:57.36 | jarray52 | How does one multiply two numbers? |
| 17:58.08 | ``Erik | globbing is turned on and * is read as a file matching argument instead of as an expression symbol, you can either turn globbing off or escape the * "expr (2 \* 2)" |
| 17:59.04 | jarray52 | will \* work even when globbing is turned off? I'm writing an mged script. |
| 18:00.33 | ``Erik | I'm not sure, one of the tcl weenies might have a good answer... might consider turning off globbing at the beginning of the script just to be sure? (I don't know if there's a scope/state issue doing that) |
| 18:01.11 | jarray52 | ``Erik: Thanks. That was helpful. |
| 18:01.14 | ``Erik | I'm sure an answer will pop up if you lurk for a while (might take a day or so, though) |
| 18:01.31 | ``Erik | or maybe look at other scripts, or just try it :D |
| 18:02.30 | jarray52 | Are example scripts available somewhere? So far I've only seen the box and logo scripts |
| 18:03.42 | ``Erik | src/brlcad/tclscripts/ would be where I'd start |
| 18:05.04 | ``Erik | mged/dbfindtree.tcl saves glob_compat_mode to a var, sets it, runs, then sets it back to var |
| 18:11.14 | jarray52 | ``Erik: Thanks again. |
| 18:11.51 | ``Erik | np, good luck! |
| 18:24.04 | *** join/#brlcad jarray52 (~bigbear@unaffiliated/jarray52) |
| 18:40.13 | brlcad | jarray52: set glob_compat_mode 0 |
| 18:40.19 | brlcad | that will turn globbing off in a script |
| 18:41.27 | jarray52 | brlcad: Is there a problem with leaving the glob_compat_mode set to 1 and using \* wherever multiplication is needed? |
| 18:42.31 | brlcad | there may be a problem depending on how the script is passed to mged, but it's more of a syntax issue |
| 18:42.56 | brlcad | it won't look like proper tcl source, especially once you start needing to also escape the other globbing characters like [] |
| 18:43.09 | brlcad | which is the common way to invoke a subexpression in tcl |
| 18:43.51 | brlcad | puts "The result is [expr 2*2]" |
| 18:45.37 | jarray52 | brlcad: That is good to know. Thanks. |
| 18:47.08 | *** join/#brlcad andrei (~andrei@188.25.162.226) |
| 19:37.07 | jarray52 | I created a part using brlcad, and I would like to create g-code for this part. I created a .dxf file using g-dxf, and I was planning on using dxf2gcode to generate gcode. However, the output of g-dxf output 3124 triangles. Will the final g-code be useable? |
| 19:37.45 | jarray52 | Does anyone have suggestions on generating g-code using brlcad? |
| 19:38.33 | brlcad | that entirely depends on dxf2gcode |
| 19:39.04 | brlcad | dxf is a polygonal format, so I'm not sure what the concern is exactly though |
| 19:41.39 | jarray52 | brlcad: I was concerned that converting to a format that uses triangles will lose information/make the information unuseable. But, if that's the only way to store things in the dxf format, then it is dxf2gcode's task to handle the generation of gcode. |
| 20:54.26 | jarray52 | brlcad: When I ran g-dxf -i -o part.dxf -p a.g part.c, I got the error message _db_walk_subtree() FAIL on ... |
| 21:05.03 | brlcad | jarray52: yeah, dxf is a rather limited format in that regard -- it has some support, but I highly double dxf2gcode would handle them any more than we would write them |
| 21:05.57 | brlcad | whether a FAIL message is significant depends on your hierarchy construction (and whether it's correct) and the object that failed |
| 21:06.22 | brlcad | best way to check is to run dxf-g and look at the result in mged, see if it's missing anything you care about |
| 21:07.23 | brlcad | if it is, you can often overcome the failure by changing the geometry that failed in minor ways (so that faces and edges aren't coinciding) |
| 21:31.22 | *** part/#brlcad jarray52 (~bigbear@unaffiliated/jarray52) |
| 22:02.10 | *** join/#brlcad jarray52 (~bigbear@unaffiliated/jarray52) |
| 22:02.51 | *** join/#brlcad cristina (~quassel@188.24.72.207) |
| 22:19.12 | jarray52 | What's the best way to go from brlcad 3D models to engineering drawings and/or file formats that can be submitted for the purpose of manufacturing physical parts? |
| 22:27.42 | brlcad | jarray52: the best way will be our step exporter, but we're not there yet -- for now, it's the complexities of performing an export conversion |
| 22:28.56 | jarray52 | brlcad: Once the step exporter is complete, we can create a step file and then submit it to a machine shop? |
| 22:29.04 | brlcad | yep |
| 22:29.11 | brlcad | technically most take stl these days too |
| 22:29.20 | brlcad | it's sort of the lowest common denominator |
| 22:29.28 | jarray52 | stl seems to be a lossy format, right? |
| 22:29.29 | brlcad | some *only* take stl |
| 22:30.01 | brlcad | of course, but that is meaningless for some purposes (like gcode generation) |
| 22:30.13 | brlcad | you just have to control the tessellation quality (with is hard, but necessary) |
| 22:31.08 | jarray52 | brlcad: Can gcode generated from stl create parts with small tolerances? Parts such as pistons. |
| 22:32.58 | brlcad | depends which machine is creating the piston |
| 22:33.15 | jarray52 | brlcad: Something like a turning center. |
| 22:33.26 | brlcad | the only way to do that well is with a CAM system that goes *directly* from the parametric form to output "smart" gcode instructions |
| 22:34.16 | brlcad | the ones that take stl will perform an automatic tool-path conversion, which is but yet another approximation on top of the stl |
| 22:34.58 | jarray52 | Are there any CAM systems that output "smart" gcode which can use brlcad models as the input? |
| 22:35.09 | jarray52 | the CAM system can be open source or commercial. |
| 22:35.11 | brlcad | a turning center is going to want precision measurements, probably drafting instructions so that they can set turning knobs to specific measurements |
| 22:35.34 | brlcad | there are lots of commercial ones, gibbscam comes to mind |
| 22:35.53 | brlcad | there are a couple toy open source projects, but nothing to the level of precision a piston would imply |
| 22:36.35 | jarray52 | what input formats does gibbscam use? |
| 22:37.18 | brlcad | no idea |
| 22:37.30 | brlcad | but they're one of the big names |
| 22:37.36 | brlcad | so it should be "most formats" |
| 22:38.24 | jarray52 | Does any CAM system take brlcad databases as an input? |
| 22:39.56 | brlcad | I have heard of a couple commercial CAD systems that implemented support for our format, but not any of the CAM systems |
| 22:40.12 | brlcad | it'd be unlikely -- best bet is going to be that step exporter |
| 22:40.40 | brlcad | perhaps g-iges, but I lack faith in it's stability (it's a dying format) |
| 22:41.19 | brlcad | jarray52: so you see our conundrum and why we need more hands ;) |
| 22:41.54 | brlcad | that's a gap I've been working on bridging for a long time |
| 22:41.56 | jarray52 | brlcad: I would like to see brlcad grow into a complete solution. |
| 22:42.05 | brlcad | ditto |
| 22:42.30 | brlcad | getting nurbs/brep support was a HUGE piece of that puzzle |
| 22:42.40 | brlcad | that lets us import and gives us foundation for export |
| 22:43.38 | brlcad | that's also the same foundation that will give us shaded opengl displays, not just wireframe, not just raytracing |
| 22:44.07 | brlcad | same foundation giving STABLE export to the polygonal exports without those FAIL messages |
| 22:44.28 | brlcad | same foundation giving explicit boundaries so we can spit out drafting documents |
| 22:45.37 | jarray52 | How far is brlcad from producing drafting documents? |
| 22:45.45 | jarray52 | labeled drafting documents? |
| 22:48.08 | brlcad | we have about ... |
| 22:48.11 | brlcad | calculates in head |
| 22:50.50 | brlcad | there's about 9 weeks of effort needed to spit out a labeled drafting document (with proper infrastructure support and some basic docs) |
| 22:51.32 | jarray52 | brlcad: Is anyone working on this? |
| 22:53.04 | brlcad | the problem is that is roughly 4 major tasks to get there (csg->brep, projected bbs, ann objects, output) |
| 22:53.29 | brlcad | two of those are being worked, so we're getting closer, but nobody specifically working towards that goal |
| 22:54.05 | jarray52 | brlcad: I was imagining something to draw lines on an rtedge drawing. Is that the wrong approach? |
| 22:54.17 | brlcad | csg->brep is top priority this year, kind of waiting to see how good a job on the hardest problem |
| 22:54.35 | brlcad | you could do that now frankly, but yeah completely different approach |
| 22:55.14 | jarray52 | brlcad: I have to checkout of this room. I will see you around though. |
| 22:55.14 | brlcad | getting that machine-precise is tricky |
| 22:55.24 | brlcad | okay, see ya |
| 22:55.28 | jarray52 | brlcad: Yes. That is the tricky part. |
| 22:55.41 | jarray52 | brlcad: Thanks for answering my questions. |
| 22:55.42 | brlcad | doing it via brep and projection is not as hard |
| 22:56.04 | jarray52 | brlcad: I see. |
| 22:56.18 | jarray52 | brlcad: Thanks. Bye. |
| 22:56.20 | brlcad | cya |
| 22:56.27 | *** part/#brlcad jarray52 (~bigbear@unaffiliated/jarray52) |
| 22:58.51 | brlcad | finishes adding some basic table formatting to help with Deuces readability |