| 14:26.22 | starseeker | mpictor: what do you think - would it be worthwhile to have someone take a run at multithreading issues? |
| 19:57.46 | *** join/#stepcode starseek1r (~starseeke@66-118-151-70.static.sagonet.net) |
| 20:45.50 | mpictor | starseeker: oops, didn't see your q until now |
| 21:10.06 | mpictor | looks like spammers are playing tug of war with the wiki main page :( |
| 21:41.52 | starseeker | mpictor: just curious - do you use stepcode in your "real" job or is it just a hobby? |
| 21:46.03 | starseeker | hates spammers |
| 21:46.36 | mpictor | heh I wish my real job was related to this! |
| 21:46.39 | mpictor | just a hobby |
| 21:46.49 | starseeker | bows in geek awe |
| 21:47.12 | mpictor | matter of fact, I'm looking for a new job - my first where I'll be a programmer |
| 21:47.16 | mpictor | heh |
| 21:48.13 | starseeker | did you see the inquriy about threading? |
| 21:48.33 | mpictor | what about you? contributing to BRL-CAD as a hobby, or as part of your job? |
| 21:48.44 | mpictor | I replied on the list |
| 21:48.45 | starseeker | both |
| 21:48.48 | starseeker | ah, k |
| 21:48.50 | mpictor | cool |
| 21:50.51 | mpictor | starseeker: I also asked on the list about compilers. I'd appreciate your thoughts on that if you have a chance |
| 21:51.15 | starseeker | mpictor: can do - should be able to reply tonight sometime |
| 21:51.22 | mpictor | cool |
| 21:52.07 | mpictor | by the way, don't be afraid to speak up if you have any feedback on the direction STEPcode is going or on features that are most important to you! |
| 21:52.29 | starseeker | mpictor: so far, the directions and features are looking pretty awesome |
| 21:52.35 | mpictor | good |
| 21:53.00 | starseeker | the "break it into smaller pieces so it can actually be debugged" bit is seriously cool |
| 21:53.54 | mpictor | I'm glad I got that working without writing an entire new parser from scratch! |
| 21:54.11 | starseeker | indeed - I half expected that would be necessary |
| 21:54.20 | mpictor | and I think I'm close to fixing the bug |
| 21:55.30 | mpictor | then I hope to merge that branch, fix the bug(s) in my inverse attribute branch, and then it will be possible to check what instances reference another |
| 21:55.50 | mpictor | well, when using the lazy loading code |
| 21:56.15 | starseeker | nods. Is the next stage then to try and get the lazy loading code into a state where it can become "the" code? |
| 21:56.35 | mpictor | yes |
| 21:57.14 | mpictor | actually, I think I'm going to do some refactoring first - fix the things that annoy me the most |
| 21:57.28 | starseeker | grins |
| 21:57.41 | starseeker | one of the joys of doing open source projects |
| 21:57.56 | starseeker | squash the annoyances! |
| 21:58.13 | mpictor | maybe push the average function length in exp2cxx farther from 1000 and closer to 50-100 |
| 21:58.18 | mpictor | yeah |
| 21:58.23 | starseeker | heh |
| 21:58.49 | starseeker | yeah, that sort of function length is one of those "this code is from academia" flags |
| 21:59.08 | mpictor | so I'm not the only one who thinks that way, eh? |
| 21:59.27 | mpictor | you should see the FBICS code x-( |
| 21:59.37 | mpictor | it's a CAM system developed at NIST |
| 21:59.55 | starseeker | hmm, haven't heard of that before |
| 22:00.17 | mpictor | I think it has 1 file per executable, and at least one is around 40k lines |
| 22:01.13 | mpictor | the author - same guy who wrote rs274-ngc - doesn't want it published as is, because it is missing a graphics system and kernel |
| 22:01.24 | starseeker | heh |
| 22:01.30 | starseeker | details, details... |
| 22:01.31 | mpictor | it used parasolid and HOOPS IIRC |
| 22:01.51 | mpictor | oh, and STEPTools to generate STEP code |
| 22:02.02 | starseeker | smacks forehead |
| 22:02.08 | mpictor | that's the reason I first dug into SCL |
| 22:02.17 | starseeker | the *one* place you'd think they'd use SCL |
| 22:02.51 | mpictor | I asked, and he said he'd been involved in the very early development at NIST but wasn't aware that it had gone anywhere |
| 22:03.30 | mpictor | Tom Kramer... couln't think of his name at first |
| 22:05.28 | starseeker | is he game longer term to publishing it if he can replace those bits with stepcode/opencascade/brlcad/openscenegraph/whatever? |
| 22:08.19 | mpictor | he gave me the code |
| 22:08.25 | mpictor | reads ancient email |
| 22:08.57 | mpictor | "OK, here is a zip file with FBICS in it. It is for your use only," |
| 22:08.57 | mpictor | not for putting on the web; it's too messy for that, and the |
| 22:08.57 | mpictor | NIST disclaimer needs to be inserted in many dozens of |
| 22:09.08 | mpictor | files." |
| 22:09.18 | starseeker | nods |
| 22:09.30 | starseeker | so you'd have to feed it back to him cleaned up to get his OK probably |
| 22:09.42 | mpictor | I thought he'd asked that it not be published until it worked |
| 22:09.52 | mpictor | maybe |
| 22:10.11 | mpictor | he provided the disclaimer that needs to be inserted |
| 22:10.42 | starseeker | but the "messy" part would have to go too, which is too vague |
| 22:10.47 | mpictor | in adition, I just saw an email from someone else at NIST - I believe his boss - that makes it clear that their work is public domain |
| 22:11.08 | mpictor | split into smaller files, run through astyle, ... |
| 22:11.15 | mpictor | I'll email him about it |
| 22:11.25 | starseeker | cool |
| 22:55.33 | mpictor | not sure how long it will take for him to reply |
| 22:55.53 | mpictor | if I don't say anything for a week or so, will you remind me? |
| 22:59.00 | mpictor | IIRC, FBICS does collision detection by creating a cube at the point in question, subtracting from the object, and checking whether the object's mass changes |
| 23:01.46 | mpictor | oh, it also uses ap224, which ap238/iso14649 are based upon |