| 00:30.28 | jano | gnop |
| 03:21.39 | PrezKennedy | ding |
| 03:47.06 | jano | gnod |
| 03:47.40 | danfalck | ok....tell me about brlcad |
| 03:51.46 | jano | well.. he's male, late twenties, likes to program |
| 03:54.50 | danfalck | do you use the brlcad application? |
| 04:00.30 | jano | sorry, I haven't used it in a long time |
| 04:00.42 | jano | it's essentially a cad program |
| 04:00.48 | jano | that mainly uses CSG for modelling |
| 04:02.31 | danfalck | are there a lot of programmers working on it? |
| 04:06.19 | danfalck | is it a project with a lot of potential? |
| 04:06.39 | danfalck | I'm interested in o/s cad/cam |
| 04:06.54 | jano | it's used by the government if that means anything to you |
| 04:07.04 | danfalck | I do tooling design/drafting and some CNC programming |
| 04:07.15 | jano | best would be to talk to brlcad when he wakes up |
| 04:07.21 | danfalck | ok. thanks |
| 04:07.25 | jano | hmm |
| 04:07.44 | jano | i dunno if he's awake |
| 04:07.55 | brlcad | i'm here.. gimmie a secd |
| 04:08.00 | jano | :) |
| 04:08.04 | danfalck | hi |
| 04:08.11 | danfalck | thanks jano |
| 04:31.23 | brlcad | sorry about that danfalck .. was away in the middle of something |
| 04:31.27 | brlcad | what would you like to know? |
| 04:32.33 | brlcad | BRL-CAD is a solid modeling system |
| 04:33.16 | brlcad | includes a csg-based modeler, several raytracers, lots of geometry import/export controls, some animation and procedural geometry support |
| 04:33.38 | brlcad | for developers it gets even better, with a couple dozen isolated libraries |
| 04:33.48 | danfalck | you invited us over from the #emc irc channel. |
| 04:33.59 | danfalck | I'm into drafting/tool design and CNC programming |
| 04:34.12 | brlcad | ahh, I thought your name was familiar |
| 04:34.26 | danfalck | I'm interested in open source cad/cam |
| 04:34.28 | brlcad | i think narnia mentioned you too |
| 04:34.59 | danfalck | would brlcad be useful, at some point, in outputting g-code? |
| 04:35.15 | brlcad | at some point, yes -- definitely |
| 04:35.28 | danfalck | ok, good |
| 04:35.34 | brlcad | in fact I believe that's specifically one of narnia's goals to get working |
| 04:35.56 | danfalck | what have you been able to do with brlcad so far? |
| 04:36.19 | brlcad | brl-cad is used extensively by the gov't for ballistics and vulnerability analyses |
| 04:36.52 | brlcad | e.g. there's a complete brl-cad model of pretty much any tank in existance |
| 04:37.04 | danfalck | ok. do you have it working on your computer? |
| 04:37.07 | brlcad | down to the wires, fuel lines, etc |
| 04:37.35 | brlcad | heh, yes, it works on my computer |
| 04:37.39 | brlcad | and lots of others ;) |
| 04:38.01 | danfalck | ok. I've been looking at the web site and just started skimming the mged tutorial |
| 04:38.02 | brlcad | it's pretty cross-platform -- weakest link is probably windows |
| 04:38.15 | danfalck | I don't need Micro$oft |
| 04:38.21 | brlcad | mged is one of 400 tools, albeit the one that gets most of the attention |
| 04:38.35 | brlcad | i'm not that fond of it myself, but it is in widespread use in the gov't |
| 04:39.44 | danfalck | are you a draftsman or computer programmer? |
| 04:40.20 | brlcad | my primary hat is that of a computer scientist |
| 04:40.46 | danfalck | have you been working with it for a long time? |
| 04:41.00 | brlcad | software design, graphics, AI |
| 04:42.13 | danfalck | sorry for all the questions. I'm trying to get a feel for what's happening with the project now. |
| 04:43.39 | danfalck | I have looked at it in the past but always avoided it because of the non-gpl license. But that's changed |
| 04:45.07 | brlcad | no problem |
| 04:45.40 | brlcad | been working with brl-cad for about 7 years, computer programming for about 20 |
| 04:45.49 | danfalck | great |
| 04:46.28 | brlcad | most of the other/older brl-cad devs are at earshot too |
| 04:47.45 | danfalck | I see lots of references to CSG. Can it do other things like draw profiles and extrude too? |
| 04:49.02 | brlcad | yes |
| 04:49.07 | danfalck | great |
| 04:49.12 | brlcad | there are various "primitive shapes" in brlcad |
| 04:49.31 | brlcad | one of them is a "sketch" .. which is basically a common 2d drawing |
| 04:49.37 | danfalck | good |
| 04:49.40 | brlcad | that may be extruded to become a solid model |
| 04:50.07 | danfalck | I'm thinking in terms of drawing pockets to be milled out |
| 04:50.15 | danfalck | or engraving on a surface |
| 04:50.15 | brlcad | there are also extruded bitmaps and a couple other similar 2d constructs |
| 04:51.28 | danfalck | or maybe even turned applications |
| 04:51.47 | danfalck | I've always got CNC on the brain |
| 04:52.51 | danfalck | what's the base machine right now for running the application on Linux? |
| 04:52.54 | brlcad | I'll need to read up more on g-code and cnc formats as that's admitedly outside of my field of experience to date |
| 04:53.08 | brlcad | base machine? |
| 04:53.30 | brlcad | if it doesn't work, it's a bug or limitation of the new build system |
| 04:53.33 | danfalck | I have a 500 mhz AMD box here |
| 04:53.50 | danfalck | lowest class of computer... |
| 04:54.19 | danfalck | or I should say, minimum requirements |
| 04:54.50 | brlcad | there's not really minimum requirements, unless you're familiar with old machines like a gould 9000 or a vax 11/780 |
| 04:55.09 | danfalck | ok. |
| 04:55.26 | brlcad | it'll just be a function of your model size |
| 04:55.37 | danfalck | thanks |
| 04:55.52 | danfalck | G-code isn't really complicated |
| 04:55.58 | brlcad | that's source code though -- there are implicit limitations on the posted binaries |
| 04:56.04 | danfalck | it's just lines and arcs |
| 04:57.15 | brlcad | e.g. the binaries are linked against opengl .. so massaging (or recompiling) is needed for non-opengl |
| 04:57.28 | brlcad | lines and arcs are pretty simple |
| 04:57.59 | brlcad | could probably make a sketch -> g-code converter easily enough |
| 04:58.12 | danfalck | I'm learning python right now. Using simple drawing programs to study |
| 04:58.14 | brlcad | but more useful would probably be an arbitrary plane cross section to g-code |
| 05:01.26 | brlcad | what kind of views are usually significant for cnc machines? |
| 05:01.53 | brlcad | does it genearlly work on fixed orthogonal planar projection outlines? |
| 05:02.06 | brlcad | or are they arbitrary? |
| 05:02.25 | danfalck | It depends on the work you want to do. Basic milling can be done while looking at 2D view |
| 05:02.55 | danfalck | but it's nice to have ortho view to see where the end mill(tool bit) is plunging or going back up |
| 05:03.14 | brlcad | 2d milling is assumedly from a model that was modeled in 2d usually no? |
| 05:03.23 | danfalck | yes |
| 05:03.30 | brlcad | on the same plane even |
| 05:03.36 | danfalck | yes |
| 05:03.56 | brlcad | what about 3d milling? |
| 05:04.00 | danfalck | but if you are doing multiple depth passes, it's nice to see where the tool is going |
| 05:04.16 | danfalck | for surfacing (3d milling) ortho views are a must |
| 05:04.34 | danfalck | for lathe work 2d is just fine |
| 05:04.40 | brlcad | hmm.. multiple depths .. how's that generally work in the modeling software? |
| 05:05.03 | danfalck | well in commercial cam software, you start with a profile |
| 05:05.17 | danfalck | select the direction of the tool path |
| 05:05.30 | danfalck | by clicking on the profile geometry on the screen |
| 05:05.52 | danfalck | and changing the direction of arrows on the profile path |
| 05:05.54 | brlcad | and it output's "slices" to mill |
| 05:06.10 | brlcad | s/slices/depths/ |
| 05:06.13 | danfalck | then you fill in a dialog box telling the app depth, etc |
| 05:07.19 | danfalck | so simple profiles, with dialog boxes, then an editor popping up with G-code for the user to examine |
| 05:07.55 | danfalck | It could really be done with a fairly simple program, which I would like to learn how to program |
| 05:08.11 | danfalck | so, I'm playing with some python cad apps |
| 05:08.24 | brlcad | say I have a circle 2d model that I want for 3d milling? |
| 05:08.36 | brlcad | like I want to mill a half sphere |
| 05:08.41 | brlcad | how would that work? |
| 05:08.58 | danfalck | well then I guess it gets more complicated |
| 05:09.29 | danfalck | I have seen "waterline" machining strategy where the tool goes around the part in the same plane |
| 05:09.39 | danfalck | then moves down to the next depth |
| 05:09.55 | brlcad | i can imagine how the machine itself does it :) |
| 05:09.55 | danfalck | and does the same thing again at the larger diameter |
| 05:10.12 | danfalck | there are zigzap strategies |
| 05:10.14 | brlcad | how do the commercial modelers do that? :) |
| 05:10.41 | danfalck | surfaces are selected |
| 05:11.02 | danfalck | the app projects some sort of mesh over the surface |
| 05:11.20 | danfalck | that is supposed to project the toolpath |
| 05:11.38 | danfalck | without gouging into the part (tool tips have to be considered) |
| 05:12.14 | danfalck | I've done guitar necks using Smartcam a few years ago |
| 05:12.33 | danfalck | and we used a strategy that ran the cutter along the length of the neck |
| 05:12.50 | danfalck | moving to a different Z depth every pass |
| 05:13.03 | danfalck | we used a ball end mill |
| 05:13.04 | brlcad | right.. but how did you tell the software to do that? |
| 05:13.15 | brlcad | i presume you modeled the neck in 3d |
| 05:13.25 | danfalck | yes, with ACAD, Rhino |
| 05:13.44 | brlcad | ACAD or autocad? |
| 05:13.52 | danfalck | Autocad |
| 05:13.54 | brlcad | ahh |
| 05:14.09 | brlcad | (there's another cad package called acad) |
| 05:14.12 | danfalck | oh |
| 05:14.29 | danfalck | Now I work in a shop that does mostly turning |
| 05:14.39 | danfalck | and it's a whole different ball game |
| 05:14.53 | danfalck | I program by hand... |
| 05:15.24 | brlcad | so now once you have the model imported from your cad package, how did you go about telling the software what you wanted to do? |
| 05:15.38 | brlcad | milling-wise |
| 05:15.49 | danfalck | the model imported in as a lofted surface |
| 05:16.07 | brlcad | ah, so that gave you a starting plane |
| 05:16.20 | danfalck | so once it's in the CAM app, you end up selecting all the lofted arcs in sequence |
| 05:16.34 | danfalck | then the app has an idea where to start from |
| 05:16.52 | brlcad | were they linear lofts? |
| 05:17.05 | brlcad | a guitar neck isn't linearly lofted in any dimension |
| 05:17.27 | danfalck | well think of it as a series of arc in the YZ plane |
| 05:17.45 | danfalck | sorry I'm not better at explaining it |
| 05:18.04 | brlcad | better yet.. what was your model lofted from? |
| 05:18.18 | brlcad | from the base/top of the neck? from the flat top surface? |
| 05:18.20 | danfalck | I started by digitizing an existing model |
| 05:18.46 | danfalck | layed the neck with the fretboard(flat surface) down |
| 05:19.05 | danfalck | probed over the back from one side to the other |
| 05:19.19 | danfalck | one YZ profile at a time |
| 05:19.31 | danfalck | move in X .1" then do it again |
| 05:19.43 | brlcad | so you modeled the parabolic curve for one slice? |
| 05:20.26 | danfalck | I actually had to probe Jimmy Pages guitar neck and it was worn in spots |
| 05:20.39 | danfalck | so we did the whole back of the neck |
| 05:20.48 | danfalck | we didn't do the headstock |
| 05:21.00 | brlcad | i'm asking, though .. how did you "do" that back of the neck |
| 05:21.23 | brlcad | you have basically a long rectangle for starters |
| 05:21.29 | danfalck | ok I see what you mean, I think |
| 05:22.01 | danfalck | I ran the probe across the neck from side to side generating a roughly parabolic shape |
| 05:22.26 | danfalck | probed slices .1" apart from each other. |
| 05:22.32 | brlcad | no worries.. I can wander on over to a milling shop if need be :) |
| 05:23.00 | danfalck | forming a series of profiles that remind one of the skeleton of a snake :) |
| 05:23.02 | brlcad | or just talk to one of my friends that work for some of the cam companies |
| 05:23.31 | danfalck | basic milling is a lot easier to deal with |
| 05:23.36 | brlcad | okay, so sliced perpendicularly down the neck |
| 05:23.42 | danfalck | yes |
| 05:23.57 | danfalck | once the data is in the modelling/cad program |
| 05:24.03 | brlcad | so you end up with a slew of parabolicish curves |
| 05:24.09 | danfalck | yes |
| 05:24.11 | brlcad | that end up forming the surface mesh |
| 05:24.38 | brlcad | so when you brought that into the cam software |
| 05:24.52 | danfalck | yes. But I didn't need to massage the surface much since I took so many readings |
| 05:25.03 | brlcad | you told it to start with the long rectangular base as the starting face .. i.e. "the bottom" |
| 05:25.16 | danfalck | yes |
| 05:25.35 | danfalck | tell the cam app to start machining from one end or the other |
| 05:25.49 | danfalck | give it the diameter of the ball end mill |
| 05:26.04 | brlcad | then you selected each arc from smallest to large/widest for milling "down" the neck |
| 05:26.44 | danfalck | With Smartcam, it could either loft the same way that I probed the part or it could generate lines perpendicular |
| 05:26.54 | danfalck | to the original ones |
| 05:27.19 | danfalck | I ended up machine it both ways to see which worked the best for finishing |
| 05:27.27 | brlcad | so the tool bit is actually pointing down the guitar neck (parallel to the rectangular base) as opposed to pointing towards the front of the guitar |
| 05:27.47 | danfalck | it was pointing toward the fron of the guitar |
| 05:28.04 | danfalck | as if the guitar was laying on it's face |
| 05:28.09 | brlcad | okay |
| 05:28.30 | danfalck | we used a giant Komo router to do the first 50 necks for production |
| 05:28.30 | brlcad | hm.. then the software must of had some sort of path interpolation |
| 05:28.35 | danfalck | yes |
| 05:28.45 | brlcad | to tell it how to mill smoothly across the slew of slice surfaces |
| 05:29.01 | danfalck | yes it was really slick |
| 05:29.16 | danfalck | multi thousand dollar package |
| 05:29.37 | danfalck | it might have been 10k back then |
| 05:29.51 | danfalck | Luckily I didn't have to buy it myself |
| 05:30.19 | danfalck | I have a small garage shop for a hobby |
| 05:30.44 | danfalck | and use Autocad , Rhino and a small cam package named Vector |
| 05:30.57 | danfalck | I just want to do some small hobby stuff |
| 05:31.03 | jano | ooo i remember rhino |
| 05:31.11 | jano | what happened to them? they still around? |
| 05:31.13 | danfalck | I used to do lots of engraving with Vector |
| 05:31.27 | danfalck | rhino is still very popular |
| 05:31.30 | jano | hmm |
| 05:31.42 | jano | for what.. cad stuff or ? |
| 05:31.46 | danfalck | yes |
| 05:31.48 | jano | ah |
| 05:32.00 | danfalck | jewelry industry uses it a lot |
| 05:32.06 | jano | there's this other program out now that seems to have amazing results for other 3d |
| 05:33.02 | danfalck | I hear that a lot of people use Solidworks |
| 05:33.16 | danfalck | We just purchase ProE at work. Haven't used it yet |
| 05:33.25 | jano | it was a program used by the dudes that did LoTR |
| 05:33.43 | danfalck | But, I'm more interested in open source stuff because I like to tinker |
| 05:33.47 | jano | yah |
| 05:34.55 | danfalck | brb |
| 05:37.42 | danfalck | ok |
| 05:42.18 | danfalck | I need to go. I'll be monitoring this channel from now on. Good talking with you guys. |
| 06:14.02 | brlcad | same to you |
| 06:14.20 | brlcad | yeah, most commercial cad packages are several thousand dollars |
| 06:14.41 | brlcad | heck even brl-cad has had millions of dollars invested in it |
| 06:18.00 | brlcad | that's what's made brl-cad's open sourcing so mutually hopeful |
| 10:06.40 | *** join/#brlcad Pimpinella (~frank@p50821131.dip0.t-ipconnect.de) | |
| 15:56.40 | danfalck | morning |
| 16:29.51 | brlcad | mornin |
| 16:40.44 | danfalck | can brlcad do any work on meshes? |
| 16:40.54 | danfalck | or export them? |
| 16:41.52 | brlcad | yes |
| 16:42.45 | brlcad | brl-cad deals with triangles via BoT (Bag of Triangles) and NMG (generic non-manifold geometry) primitives |
| 16:43.43 | danfalck | it looks like I need to install brlcad and explore it |
| 16:44.45 | brlcad | by design, brl-cad prefers non-polygonalized geometry but there is a conversion path to triangles for any object/primitive |
| 16:47.07 | danfalck | so if I wanted to get started, I should just grab the *tar file and start reading the mged (sp?) manual? |
| 16:47.39 | brlcad | that would be reasonable :) |
| 16:48.23 | danfalck | ok. I will do it. |
| 16:49.02 | danfalck | I eventually want to set up a manufacturing system using o/s cad linked to EMC on my machines |
| 16:49.09 | danfalck | in my garage |
| 16:49.58 | danfalck | I own some cam stuff, but I have totally lost my appetite for closed source software |
| 16:56.22 | brlcad | brl-cad's always been written in as "open" a development manner as was possible |
| 16:57.04 | brlcad | it was only through significant persistance and a bit of luck that it was able to go open-source |
| 16:58.20 | brlcad | but it does make complete sense for brl-cad's specific situation and history |
| 16:58.21 | jano | does BoT come with a bonus set of monkeys in a barrel? |
| 17:00.09 | jano | the sound you just heard was my joke doing mach 7 over your head |
| 17:00.15 | jano | :D |
| 17:00.42 | jano | Bag of Triangles, Barrel of Monkeys |
| 17:02.29 | brlcad | dunno if he'd actually eat those |
| 17:02.39 | brlcad | prolly |
| 17:03.51 | jano | has it grown anymore? |
| 17:04.30 | danfalck | so you mentioned that a lot of money went into the development of brlcad |
| 17:04.34 | danfalck | millions? |
| 17:14.44 | brlcad | consider that the package has been around for 20 years with constant funding, support, staff, training, improvements, etc |
| 17:15.29 | brlcad | that doesn't necessarily mean that it's commercially viable or great for the general public's needs -- it was designed to efficiently perform a very specific set of tasks |
| 17:15.40 | brlcad | and it does that _very_ well |
| 17:16.48 | brlcad | but things like usability and user interface enhancements is something that gets handled as an afterthought as it's never funded |
| 17:28.10 | danfalck | thanks for the info. I will install tonight. I'm going out for a walk. |
| 17:28.36 | brlcad | have a good one |
| 19:09.08 | *** join/#brlcad narnia (~terrylr@johann.blauedonau.com) | |
| 19:25.01 | PrezKennedy | ding |
| 19:53.09 | narnia | knock, knock, anyone home? |
| 19:58.23 | danfalck | hi |
| 19:58.31 | danfalck | narnia: hello |
| 20:05.55 | narnia | danfalck, hello, how goes it? |
| 20:06.30 | danfalck | ok |
| 20:06.44 | narnia | danfalck, sorry, i was searching on google for some more info on the express language. |
| 20:06.51 | danfalck | I have been quizzing brlcad about BRLCAD |
| 20:07.06 | narnia | danfalck, that is good. |
| 20:07.19 | danfalck | it sounds like a pretty heavy duty app with a lot of potential for what we would like it for |
| 20:07.26 | narnia | danfalck, so what do you think so far? |
| 20:07.53 | narnia | brb, need coffee. |
| 20:08.05 | danfalck | narnia: well, I have just scratched the surface. It is kind of intimidating |
| 20:10.22 | danfalck | narnia: but if you're into it, I am willing to try it out |
| 20:10.37 | danfalck | brb-fixing sandwich for son |
| 20:14.12 | danfalck | ok I'm back |
| 20:14.33 | narnia | danfalck, i am here. |
| 20:14.55 | danfalck | narnia: are you running brlcad now? |
| 20:15.03 | narnia | yes. |
| 20:15.20 | danfalck | what have you been able to do with it so far? |
| 20:17.22 | narnia | mainly building models of past projects. exploring alternative methods of construction. |
| 20:18.49 | danfalck | can one make a drawing w/ dimensions on it for presentation to a shop? |
| 20:19.26 | danfalck | I see lots of references to 3d models and CSG, but wonder how deep it is. |
| 20:20.21 | narnia | probably the best way would be to look at the brlcad manual and tutorial. |
| 20:20.41 | danfalck | yes |
| 20:20.51 | danfalck | mged manual |
| 20:21.01 | narnia | yes |
| 20:22.17 | danfalck | I have a Morphix based computer here, with EMC RC46 on it. I will see how much space I have left on it and install there. |
| 20:22.41 | danfalck | I've got the manual on this OS X box now |
| 20:23.46 | narnia | as in macos x? |
| 20:25.08 | danfalck | yes |
| 20:25.20 | narnia | imac or emac? |
| 20:25.23 | danfalck | cube |
| 20:25.33 | narnia | ah |
| 20:25.41 | danfalck | very quiet and small |
| 20:26.06 | danfalck | I can run a load of gnu apps on it using X11 |
| 20:26.28 | narnia | i have an imac and an emac. i would like to have a mac mini to play with. those look interesting. |
| 20:26.49 | danfalck | same here. I would like one too. |
| 20:28.12 | narnia | did brlcad happen to mention if he will be around today when you chatted with him? |
| 20:28.47 | danfalck | he was around this morning about 8:00 am pacific time |
| 20:33.55 | narnia | ah |
| 20:34.09 | narnia | sorry was dealing with the puppy. |
| 20:34.35 | narnia | i need to take him for a walk. i will be back later. |
| 20:34.46 | danfalck | ok chat with you later |
| 20:39.02 | *** join/#brlcad jnc (shadow@macco.pimpcat.org) | |
| 20:39.05 | jnc | howdy |
| 20:41.51 | danfalck | hello |
| 20:50.58 | jnc | i'm thinking about writing up an ebuild, which would allow Gentoo Linux users to install brlcad |
| 20:52.32 | danfalck | cool |
| 20:53.03 | danfalck | I'm using a Morphix/Debian based distro |
| 21:20.09 | brlcad | jnc: starseeker had put together and submitted an ebuild already |
| 21:20.34 | brlcad | not sure where or how far he got, but he did help me track down a coupld sandbox problems that were fixed |
| 21:21.39 | brlcad | jnc: http://bugs.gentoo.org/show_bug.cgi?id=77197 |
| 21:22.03 | brlcad | some things there I'd hope change, but it's a start |
| 21:25.20 | brlcad | things like running on more than x86 -- i just decommisioned by ppc gentoo box, or I would be pressing that myself |
| 21:26.53 | brlcad | a couple more concerns are fixed in the latest builds and will be in a new source/binary upload that should get posted this coming week |