--- Log opened Sun Sep 11 00:00:05 2011 | ||
-!- diane [~diane@cpe-69-205-70-55.rochester.res.rr.com] has quit [Ping timeout: 276 seconds] | 01:17 | |
-!- diane [~diane@cpe-69-205-70-55.rochester.res.rr.com] has joined ##hplusroadmap | 01:20 | |
-!- diane [~diane@cpe-69-205-70-55.rochester.res.rr.com] has quit [Ping timeout: 258 seconds] | 01:25 | |
-!- Jaakko94 [~Jaakko@host81-157-105-60.range81-157.btcentralplus.com] has joined ##hplusroadmap | 02:23 | |
-!- diane [~diane@cpe-69-205-70-55.rochester.res.rr.com] has joined ##hplusroadmap | 03:36 | |
-!- uniqanomaly [~ua@dynamic-78-8-89-177.ssp.dialog.net.pl] has joined ##hplusroadmap | 05:16 | |
-!- uniqanomaly [~ua@dynamic-78-8-89-177.ssp.dialog.net.pl] has quit [Ping timeout: 252 seconds] | 05:49 | |
-!- uniqanomaly [~ua@dynamic-78-8-90-159.ssp.dialog.net.pl] has joined ##hplusroadmap | 06:01 | |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has joined ##hplusroadmap | 07:37 | |
-!- augur [~augur@208.58.5.87] has joined ##hplusroadmap | 07:49 | |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has quit [Ping timeout: 260 seconds] | 07:59 | |
-!- uniqanomaly_ [~ua@dynamic-78-8-80-47.ssp.dialog.net.pl] has joined ##hplusroadmap | 08:06 | |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has joined ##hplusroadmap | 08:07 | |
-!- uniqanomaly [~ua@dynamic-78-8-90-159.ssp.dialog.net.pl] has quit [Ping timeout: 260 seconds] | 08:08 | |
-!- uniqanomaly__ [~ua@dynamic-78-8-86-203.ssp.dialog.net.pl] has joined ##hplusroadmap | 08:33 | |
-!- uniqanomaly_ [~ua@dynamic-78-8-80-47.ssp.dialog.net.pl] has quit [Ping timeout: 276 seconds] | 08:34 | |
-!- uniqanomaly_ [~ua@dynamic-78-8-212-64.ssp.dialog.net.pl] has joined ##hplusroadmap | 08:47 | |
-!- uniqanomaly__ [~ua@dynamic-78-8-86-203.ssp.dialog.net.pl] has quit [Ping timeout: 276 seconds] | 08:49 | |
-!- delinquentme [~delinquen@c-76-125-242-200.hsd1.pa.comcast.net] has joined ##hplusroadmap | 09:13 | |
-!- delinquentme [~delinquen@c-76-125-242-200.hsd1.pa.comcast.net] has quit [Remote host closed the connection] | 09:33 | |
-!- delinquentme [~delinquen@c-76-125-242-200.hsd1.pa.comcast.net] has joined ##hplusroadmap | 09:44 | |
-!- delinquentme [~delinquen@c-76-125-242-200.hsd1.pa.comcast.net] has quit [Remote host closed the connection] | 09:52 | |
* mayko is away: make this viral woooo: http://topologicoceans.wordpress.com/2011/03/15/diy-spectro-ii/ | 11:17 | |
-!- uniqanomaly_ [~ua@dynamic-78-8-212-64.ssp.dialog.net.pl] has quit [Quit: uniqanomaly_] | 11:36 | |
-!- mayko [~mayko@71-22-217-151.gar.clearwire-wmx.net] has quit [Ping timeout: 258 seconds] | 11:41 | |
-!- mayko [~mayko@71-22-217-151.gar.clearwire-wmx.net] has joined ##hplusroadmap | 11:54 | |
-!- diane [~diane@cpe-69-205-70-55.rochester.res.rr.com] has quit [Ping timeout: 250 seconds] | 12:00 | |
-!- delinquentme [~delinquen@c-76-125-242-200.hsd1.pa.comcast.net] has joined ##hplusroadmap | 12:06 | |
delinquentme | asdf | 12:06 |
---|---|---|
delinquentme | http://www.nature.com/nm/journal/v15/n11/full/nm.2032.html | 12:07 |
delinquentme | ^^ got a link to that one kanzure ? | 12:07 |
-!- gedankenstuecke [~bastian@phylomemetic-tree.de] has quit [Ping timeout: 276 seconds] | 12:22 | |
delinquentme | apparently these guys got aquired @ 600 million http://www.caliperls.com/tech/optical-imaging/ | 12:44 |
delinquentme | as of yesterday :D | 12:44 |
-!- strages [~makers@c-71-207-181-192.hsd1.al.comcast.net] has quit [Read error: Operation timed out] | 12:49 | |
-!- strages [~makers@c-71-207-222-170.hsd1.al.comcast.net] has joined ##hplusroadmap | 12:59 | |
-!- delinquentme [~delinquen@c-76-125-242-200.hsd1.pa.comcast.net] has quit [Remote host closed the connection] | 13:17 | |
-!- SDr [SDr@unaffiliated/sdr] has joined ##hplusroadmap | 13:26 | |
kanzure | foo | 13:29 |
-!- uniqanomaly [~ua@dynamic-78-8-212-64.ssp.dialog.net.pl] has joined ##hplusroadmap | 13:32 | |
kanzure | http://topologicoceans.wordpress.com/2011/03/15/diy-spectro-ii/ | 13:32 |
-!- CryptoQuick [~CryptoQui@c-174-51-232-237.hsd1.co.comcast.net] has joined ##hplusroadmap | 13:38 | |
CryptoQuick | kanzure: hey bud, you there? :) | 13:38 |
CryptoQuick | I need yo helpz | 13:38 |
CryptoQuick | I want to make a web-based CAD, with parametric geometry and basic CSG ops using WebGL. I'm cool if that's done in another language, I can always work through some python and stuff. I'm just concerned about turning parametric solids into mesh geometry. Is there anything out there already that can help? | 13:39 |
* mayko is back (gone 02:13:54) | 14:08 | |
kanzure | okay | 14:15 |
kanzure | CryptoQuick: parametric->mesh is easy | 14:15 |
kanzure | the parametric part is the hard part.. would you be interested in helping me implement it? | 14:16 |
CryptoQuick | kanzure: yeah? | 14:16 |
-!- eudoxia [~eudoxia@r190-135-73-58.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap | 14:16 | |
CryptoQuick | sure, actually | 14:16 |
CryptoQuick | if it's in JS, of course | 14:16 |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has quit [Ping timeout: 258 seconds] | 14:16 | |
kanzure | would you be ok i the webgl/javascript part interfaces with a server where the CAD kernel sits? | 14:16 |
kanzure | i see.. | 14:16 |
kanzure | well that makes it a little harder | 14:16 |
CryptoQuick | hmm | 14:16 |
CryptoQuick | I mean, yeah, you could use PyOCC | 14:16 |
kanzure | hmm no | 14:17 |
kanzure | i was rewriting a CAD kernel in python a while back | 14:17 |
kanzure | because opencascade is so awful | 14:17 |
kanzure | http://diyhpl.us/cgit/lolcad/plain/esolid/esolid.py | 14:17 |
CryptoQuick | buh! | 14:17 |
kanzure | then i started a wrapper that would wrap both esolid and pythonocc | 14:17 |
CryptoQuick | kanzure: it's so ... big! I'm not sure I'll be able to handle it all~ | 14:17 |
kanzure | i'd be ok with writing a cad kernel in vanilla python or vanilla javascript though | 14:17 |
kanzure | yeah... | 14:17 |
kanzure | just let the tip in | 14:17 |
CryptoQuick | just the tip? ok, well, as long as it's only a little bit | 14:18 |
kanzure | anyway, imho this is the remaining task to be done | 14:18 |
kanzure | i have been working off of a set of papers | 14:18 |
CryptoQuick | I'm currently making a boilerplate Three.js Canvas renderer script, very simple stuffs | 14:19 |
CryptoQuick | and that's always a good thing | 14:19 |
kanzure | meh that's not a chokepoint dude | 14:19 |
CryptoQuick | ofc not | 14:19 |
CryptoQuick | just having a fun | 14:19 |
kanzure | http://webgl.pythonocc.org/ is also something but i think it was just a renderer | 14:19 |
kanzure | let me upload the paper or find it.. | 14:19 |
CryptoQuick | yeah | 14:19 |
CryptoQuick | they use Three.js | 14:19 |
CryptoQuick | but I think they're using the WebGL renderer | 14:19 |
kanzure | this is the software i've been basing some of my wokr on: | 14:20 |
kanzure | http://gamma.cs.unc.edu/ESOLID/ | 14:20 |
kanzure | at one point they hhooked it up to BRLCAD but deleted their work? they are pretty awful | 14:20 |
CryptoQuick | hmm | 14:21 |
kanzure | here's the guy's thesis: | 14:21 |
kanzure | https://wwwx.cs.unc.edu/~geom/papers/documents/dissertations/keyser00.pdf | 14:21 |
kanzure | here's the source code: | 14:21 |
kanzure | http://research.cs.tamu.edu/keyser/geom/esolid/releases/version_0.3/ESOLID-0.3.tar.gz | 14:21 |
CryptoQuick | well, it's nice to have CSG and stuff, but what I really want is to perform basic CSG ops on arbitrary 2D vector drawings, in respect to other parametric solids, just like SolidWorks | 14:21 |
kanzure | in lolcad.git you can see my cython wrapper.. (esolid.py is my own implementation) | 14:22 |
kanzure | yeah this is the implementation for operations like union and difference and intersection | 14:22 |
CryptoQuick | intersection not so super intuitive-- is that ever used in SolidWorks? | 14:23 |
kanzure | intersection is like "take the common area between two somewhat overlapping cubes" | 14:23 |
CryptoQuick | I've worked with stuff like OpenSCAD, and I've done pretty cool things with it, mind you | 14:23 |
CryptoQuick | and yeah, I know :) | 14:23 |
kanzure | openscad sucks for a few reasons: | 14:23 |
kanzure | 1) custom language.. yawn | 14:23 |
CryptoQuick | it's easy syntax though | 14:24 |
kanzure | 2) all of their boolean operations are just based on hiding/showing shit in opengl buffers | 14:24 |
kanzure | well, they have some python wrappers now | 14:24 |
kanzure | pyopenscad, openscadpy, cadmium, etc. | 14:24 |
CryptoQuick | mmh | 14:24 |
CryptoQuick | their program is very hardware heavy | 14:24 |
kanzure | cloudscad did some javascript wrapper i think.. it involved bouncing shit back to a server | 14:24 |
CryptoQuick | yeah... | 14:24 |
kanzure | someone else was making a webgl/javascript library that talked to opencascade (pythonocc?) on a server | 14:25 |
kanzure | anyway, i'm totally down for writing a javascript library (or any library) | 14:25 |
CryptoQuick | oh sweet | 14:25 |
kanzure | once it's in python or javascript and if it's written well enough then it's not that hard to translate it into whatever.. or just stay with it | 14:25 |
CryptoQuick | yeah, definitely | 14:25 |
kanzure | with the help of v8 any js library will work fine with python anyway | 14:25 |
CryptoQuick | be faster than python ;D | 14:26 |
CryptoQuick | V8 changed everything | 14:26 |
kanzure | v8 changed my diapers | 14:26 |
CryptoQuick | yeah, V8 is good | 14:27 |
-!- uniqanomaly [~ua@dynamic-78-8-212-64.ssp.dialog.net.pl] has quit [Remote host closed the connection] | 14:27 | |
CryptoQuick | javascript: | 14:28 |
CryptoQuick | http://images.wikia.com/labyrinth/images/a/aa/Jareth.jpg | 14:28 |
CryptoQuick | python: | 14:28 |
CryptoQuick | http://userserve-ak.last.fm/serve/_/545533/Cyndi+Lauper+clothes+pegs.jpg | 14:28 |
kanzure | going from parametric CAD to STL isn't a big deal.. it's just a tessellation algorithm | 14:29 |
CryptoQuick | oh nice | 14:29 |
CryptoQuick | that makes sense | 14:29 |
kanzure | in lolcad i used wx and opengl to display the objects as you updated them in your interpreter.. | 14:29 |
kanzure | i just used the default sgi-provided tessellator in the glunurbs part of their lirbary | 14:30 |
kanzure | library | 14:30 |
CryptoQuick | buh | 14:30 |
CryptoQuick | SGI | 14:30 |
kanzure | apparently it has some issues on mesa? fenn's netbook displayed circles as ovals.. or something | 14:30 |
CryptoQuick | best use canvas, it's very ... compatible. | 14:31 |
kanzure | http://diyhpl.us/~bryan/screenshots/2010-08-24-1610-runten-pynurbs.png | 14:31 |
kanzure | http://diyhpl.us/~bryan/screenshots/2010-08-19-python-opengl-opennurbs.png | 14:31 |
CryptoQuick | nurbular | 14:31 |
CryptoQuick | osnap | 14:31 |
kanzure | http://heybryan.org/shots/2011-06-29-lolcad-GL_MAP1_VERTEX_3.png | 14:31 |
kanzure | http://heybryan.org/shots/2011-06-23-lolcad.png | 14:31 |
CryptoQuick | yeap, si teh nurbs | 14:31 |
kanzure | http://heybryan.org/shots/2011-04-27-1440-boolean-nurbs-difference.png | 14:32 |
kanzure | sorry.. just some screenshots of shit | 14:32 |
CryptoQuick | heheheh, they nice | 14:32 |
CryptoQuick | wanna see some of my work? | 14:32 |
kanzure | sure why not.. | 14:32 |
CryptoQuick | http://trozki.org/demos/nanoblok/nb07/ | 14:32 |
CryptoQuick | http://editor.nanoblok.com/ | 14:32 |
CryptoQuick | press E key in either to cycle through pre-built models | 14:32 |
CryptoQuick | color cube broke in the trozki version, but the trozki version has a working Z-rendering algorithm | 14:33 |
CryptoQuick | well, relatively working ... | 14:33 |
kanzure | didn't i see this before? | 14:34 |
CryptoQuick | also, this in 3D | 14:34 |
CryptoQuick | http://trozki.org/demos/nanoblok/nb09/ | 14:34 |
CryptoQuick | maybe, been a while dude :) | 14:34 |
kanzure | some company stole this and got a lot of publicity on HN | 14:34 |
CryptoQuick | also, a game I never finished ... | 14:34 |
CryptoQuick | http://trozki.org/demos/apsis/ | 14:34 |
CryptoQuick | ? | 14:34 |
kanzure | hmm what was this called.. | 14:35 |
CryptoQuick | oh yeah, also | 14:35 |
CryptoQuick | http://www.infocraft.org/three/ | 14:35 |
CryptoQuick | the conversion was all done by hand though, don't get excited :P | 14:35 |
kanzure | bah | 14:35 |
CryptoQuick | it was just a thing to demo a part, shut up :P | 14:35 |
kanzure | about that last one.. | 14:36 |
kanzure | i wrote the visualizer for http://diyhpl.us/~bryan/papers2/Automated%20Synthesis%20and%20Optimization%20of%20Gear%20Train%20Topologies%20-%20Albert%20Swantner%20-%202009.pdf | 14:36 |
kanzure | hrmm i'm trying to fidn this thing | 14:39 |
kanzure | it's not thingiview.. | 14:39 |
kanzure | here we go | 14:39 |
kanzure | CryptoQuick: http://www.3dtin.com/ | 14:39 |
CryptoQuick | yeah | 14:39 |
CryptoQuick | seen it :) | 14:39 |
CryptoQuick | very nice stuff | 14:39 |
kanzure | did they use your code? | 14:40 |
kanzure | or did you use theirs? | 14:40 |
CryptoQuick | nah | 14:40 |
CryptoQuick | neither | 14:40 |
kanzure | hmmm | 14:40 |
CryptoQuick | it's more like a Mr. Doob demo | 14:40 |
CryptoQuick | http://mrdoob.com/projects/voxels/ | 14:40 |
CryptoQuick | they work differently :) | 14:41 |
CryptoQuick | very differently, actually, considering how my renderer is fixed isometric, and uses a lot of shortcuts because of that. | 14:41 |
CryptoQuick | it's pretty unique, I'd like to think... | 14:41 |
CryptoQuick | well, kanz, I'd be cool with a python kernel, I mean, rewriting everything you have in JavaScript is not a productive activity | 14:42 |
CryptoQuick | well, not terribly productive, you know | 14:43 |
CryptoQuick | it's not conducive to finishing~ | 14:43 |
CryptoQuick | I fail at finishing :( | 14:43 |
kanzure | so i have a few gripes about CAD in general, | 14:43 |
CryptoQuick | RELEASE | 14:43 |
CryptoQuick | yeah? | 14:43 |
kanzure | first is that people say it's too hard for 1 person to do | 14:43 |
CryptoQuick | :T | 14:43 |
kanzure | the second is that all of the implementations that i have seen-- nobody understands | 14:43 |
CryptoQuick | mmh | 14:43 |
CryptoQuick | they're 'I, Pencils' | 14:43 |
kanzure | so the problem is likely that nobody has simplified the math or an understanding of how to make this sort of software | 14:44 |
CryptoQuick | quite | 14:44 |
kanzure | my "terrible scheme" was to write a very simple python library | 14:44 |
CryptoQuick | sketchup maybe comes close? but not really | 14:44 |
kanzure | and that by making the library simple enough, the actual codebase would be only maybe 1000 lines | 14:44 |
kanzure | esolid's codebase is about 20k lines | 14:44 |
CryptoQuick | ok :) | 14:44 |
kanzure | opencascade is at least 100k lines | 14:44 |
CryptoQuick | onoz | 14:45 |
kanzure | i dunno if you've ever seen the nurbs code in opencascade but it's not pretty | 14:45 |
kanzure | it's a mix of english, russian and french | 14:45 |
CryptoQuick | no nurbz maaaan | 14:45 |
kanzure | from the early 90s | 14:45 |
CryptoQuick | buh! | 14:45 |
CryptoQuick | that was back when nurbz was the it girl, i'm sure | 14:45 |
kanzure | yesterday i was working on lolcad and was going to package up a wrapper on top of it and just use esolid.so as a drop-in library for now | 14:46 |
kanzure | and slowly replace it with my slightly-working-python-rewrite | 14:46 |
kanzure | but that doesn't help towards what yuo want (a complete implementation) | 14:46 |
kanzure | *you | 14:46 |
CryptoQuick | well, wait, no | 14:47 |
kanzure | the problem i had with my complete-python-rewrite is that i didn't know how to test it | 14:47 |
kanzure | you can't just write thousands of lines of python and hope that the CAD works in the end... | 14:47 |
CryptoQuick | what I need is a library with the basic functionality of a parametric modeling program would occur, the heavy duty math, tesselation, all that, could be used with-- I'd make model data, it would process that data, and I would get mesh data back. | 14:48 |
kanzure | yep | 14:48 |
kanzure | well it would include a modeling API | 14:48 |
CryptoQuick | maybe, but I want to do the modeling in JavaScript really | 14:48 |
kanzure | yeah you'd do that | 14:48 |
CryptoQuick | it should really be interface agnostic | 14:49 |
kanzure | var sphere = new Sphere(); sphere.union(someotherobject) | 14:49 |
CryptoQuick | sure sure | 14:49 |
kanzure | can't be stl input | 14:49 |
CryptoQuick | nah, all solids | 14:49 |
CryptoQuick | well, wait-- | 14:49 |
CryptoQuick | what about DXF input? | 14:49 |
kanzure | i haven't worked with that file format much but there's many versions iirc.. like a 3d dxf format even O_o | 14:50 |
CryptoQuick | I mean, SolidWorks, which I find to be very intuitive, lets you draw the equivalent of DXF at a certain orientation in respect to another object, then that is applied (union, diff) on that object | 14:50 |
kanzure | ok so you want to extrude sketches | 14:51 |
CryptoQuick | quite | 14:51 |
kanzure | i have extruding working in the python version.. | 14:51 |
CryptoQuick | osweet | 14:51 |
CryptoQuick | well, let's just use the python kernel for now | 14:51 |
kanzure | well "working" to the extend that i am able to extrude a line into a nurbs object | 14:51 |
CryptoQuick | XHR2 is fine | 14:51 |
CryptoQuick | do we really need nurbs yo | 14:51 |
CryptoQuick | I guess nurbs is helpful for spheres... | 14:52 |
CryptoQuick | those are a little hard to do in solidworks, but only because it's hard to position the drawings. theoretically you could just draw an arc and revolve it around an axis. | 14:52 |
kanzure | what? solidworks isn't using nurbs? | 14:53 |
CryptoQuick | no idea what it does inside the engine... | 14:53 |
kanzure | what sort of mathematically defined surface are they using under the hood | 14:53 |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has joined ##hplusroadmap | 14:53 | |
kanzure | i bet it's nurbs man ;) | 14:53 |
CryptoQuick | o. | 14:53 |
kanzure | most users don't have to care about it 100% of the time | 14:53 |
CryptoQuick | so, parametric modeling and nurbs go hand-in-hand | 14:53 |
kanzure | since they just use primitives and sketches-turned-into-bezier-curves | 14:54 |
CryptoQuick | hm | 14:54 |
CryptoQuick | kay | 14:54 |
CryptoQuick | so, how to get this started ... | 14:55 |
CryptoQuick | I think what I'll do is try to get a server working with your kernel ? | 14:55 |
CryptoQuick | I'm thinking something simple, a microframework like Flask | 14:55 |
CryptoQuick | I love that one | 14:55 |
CryptoQuick | very easy! | 14:55 |
kanzure | i would suggest you read thually | 14:56 |
kanzure | fdfladjfkladf | 14:56 |
kanzure | fucking irssi | 14:56 |
kanzure | i would suggest you read the paper actually | 14:56 |
kanzure | https://wwwx.cs.unc.edu/~geom/papers/documents/dissertations/keyser00.pdf | 14:56 |
kanzure | there are also shorter review papers of this same work out there.. | 14:56 |
kanzure | shorter paper: http://gamma.cs.unc.edu/ESOLID/keyser02.pdf | 14:57 |
kanzure | then look over the source code they published.. | 14:57 |
kanzure | http://research.cs.tamu.edu/keyser/geom/esolid/releases/version_0.3/ESOLID-0.3.tar.gz | 14:57 |
kanzure | unfortunately, it has no license | 14:57 |
kanzure | i asked john keyser about the licensing about six months ago but he hasn't replied yet | 14:58 |
CryptoQuick | can't I just use your stuff plz? you're the expert, yo | 14:58 |
CryptoQuick | and that's why they make libraries, yus | 14:58 |
kanzure | well, it's a little bit much for me to chew on my own | 14:58 |
CryptoQuick | oh, you want me to help finish your stuff? | 14:58 |
kanzure | and it would be great to have someone else looking over my work and these papers | 14:58 |
CryptoQuick | hmm... | 14:58 |
CryptoQuick | well, show me what you have so far. is it that one big file you sent me earlier? I'm looking at your git | 14:59 |
CryptoQuick | http://diyhpl.us/cgit/lolcad/ | 14:59 |
kanzure | it's in lolcad.git somewhat | 14:59 |
kanzure | let me check what's uncommitted.. | 14:59 |
kanzure | ok it looks like my cython wrapper is in there in lolcad/esolid/wrapper/ | 15:00 |
kanzure | esolid.py is in lolcad/esolid/esolid.py | 15:00 |
kanzure | esolid.py is my python rewrite from the ground up | 15:00 |
kanzure | lolcad/esolid/wrapper/wrapper.pyx is my cython wrapper around libesolid.a | 15:00 |
CryptoQuick | jesus christ | 15:00 |
CryptoQuick | 4600 lines | 15:00 |
kanzure | yeah but i dunno if it's working ;) | 15:00 |
kanzure | that's the part about "i dunno how to test it" | 15:01 |
CryptoQuick | you ported this mostly from esolid codebase, right? | 15:01 |
kanzure | i was coding and coding and coding and i realized, i don't know if individual components will work or not | 15:01 |
kanzure | yes | 15:01 |
kanzure | so that's why i started the wrapper/wrapper.pyx code | 15:01 |
kanzure | i figured i'd wrap the library and then slowly replace it after it was working already | 15:01 |
kanzure | a lot of this seems like it could be replaced by scipy or numpy or something | 15:02 |
kanzure | like.. why make a custom rational polynomial class? | 15:02 |
kanzure | the next steps for the cython-based wrapper are to finish fleshing out the methods (easy), then write some converters to take data from libesolid and draw it on screen through opengl | 15:05 |
kanzure | then slowly-over-time replacing cython with rewritten versions that do not depend on libesolid | 15:06 |
kanzure | now.. the alternative is to re-do my custom-rewrite-from-scratch, *after* fully understanding that guy's phd thesis :P | 15:06 |
kanzure | and writing unit tests along the way. | 15:06 |
CryptoQuick | well, the imports for esolid.py are -very- lightweight | 15:06 |
CryptoQuick | this could be run on App Engine, for example | 15:06 |
kanzure | yep | 15:07 |
kanzure | just lots of math dude | 15:07 |
kanzure | MATHS! | 15:07 |
CryptoQuick | i not so gud at mafh :T | 15:07 |
kanzure | erm.. what does that mean, in your case? | 15:08 |
kanzure | like surely you can do algebra right? | 15:08 |
CryptoQuick | oh, dude, I'm not that bad | 15:09 |
CryptoQuick | I mean, I can use 3D matrix transformations, I've done them before in my code, but usually looking up a lot of documentation on math and code online. | 15:10 |
CryptoQuick | I've never attempted quaternions, though | 15:10 |
CryptoQuick | the thing is, I'm not that formally educated | 15:10 |
CryptoQuick | I'm a terrible student | 15:10 |
kanzure | matrix transformations are the most complicated thing going on here as far as i can tell | 15:11 |
kanzure | there's some neat simplifications and tricks that the thesis has, but they are mostly understandable ("you can eliminate these terms because..") | 15:12 |
CryptoQuick | well at least they don't use lamdas in this code, those are kinda hard to decipher oftentimes | 15:13 |
CryptoQuick | it's really not all that pythonic, tbh | 15:13 |
CryptoQuick | did your test file work? | 15:13 |
kanzure | honestly it's been a while, i forget | 15:14 |
kanzure | it's not a complete implementation | 15:14 |
kanzure | it has no syntax errors | 15:14 |
CryptoQuick | :) | 15:14 |
kanzure | it's not pythonic- absolutely true. | 15:14 |
kanzure | i wanted to make sure i knew what was going on.. | 15:14 |
CryptoQuick | heh :) | 15:14 |
kanzure | but i'm not opposed to "read the papers -> figure out what has to be done -> write a simple version from scratch" | 15:15 |
kanzure | s/simple/pythonic/ | 15:15 |
CryptoQuick | heh, you use graphs in here | 15:15 |
CryptoQuick | I've never really implemented graphs before, only trees really | 15:15 |
CryptoQuick | I've used stuff like octrees and tree search in my game experience, though | 15:15 |
kanzure | i worked in a graph theory lab for a while (in a mechanical engineering department) | 15:16 |
CryptoQuick | heh, nice | 15:16 |
kanzure | http://graphsynth.com/ is some stuff i worked on | 15:16 |
CryptoQuick | oh right! | 15:16 |
CryptoQuick | I remember you postan this once o twiz | 15:16 |
CryptoQuick | so, let's get this show on the road, man. I'm going to test your library out a bit | 15:17 |
CryptoQuick | let me know if there's any other administrative things you want to do | 15:17 |
kanzure | you should try to compile esolid from source so that you can test out the cython wrapper i wrote | 15:18 |
kanzure | there's a lot of problems in the codebase that i fixed, but i haven't produced a set of patches yet | 15:18 |
kanzure | i think it's worth me making the patches so please bug me about errors that you get | 15:18 |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has quit [Ping timeout: 260 seconds] | 15:18 | |
CryptoQuick | I'd prefer to work on your py stuff instead ... ? | 15:19 |
kanzure | how do i generate a set of patches against an entire directory of a project? | 15:19 |
kanzure | yeah sure.. the cython stuff is python | 15:19 |
CryptoQuick | you talking about the stuff in /esolid/esolid.py, or /esolid/wrappar ? | 15:19 |
kanzure | the wrapper is the cython wrapper | 15:20 |
kanzure | esolid.py is a self-contained partial untested implementation of esolid | 15:20 |
CryptoQuick | so the wrapper is necessary to run the python? I can't python esolid.py ? | 15:20 |
CryptoQuick | o i c | 15:20 |
kanzure | the wrapper is separate | 15:20 |
CryptoQuick | well, it's better to work off of that, I would think ... | 15:20 |
CryptoQuick | I see | 15:20 |
kanzure | off of what? | 15:21 |
CryptoQuick | esolid.py | 15:21 |
CryptoQuick | interesting | 15:23 |
CryptoQuick | yo shell script runs | 15:23 |
CryptoQuick | teh sweetz | 15:23 |
CryptoQuick | I wonder what it all means~ | 15:23 |
kanzure | heh | 15:23 |
kanzure | dunno you'll have to check | 15:23 |
kanzure | this is a pretty big project, i often forget about huge chunks | 15:23 |
kanzure | see pm | 15:23 |
kanzure | ok thanks. cya | 15:27 |
CryptoQuick | :) | 15:27 |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has joined ##hplusroadmap | 15:33 | |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has quit [Client Quit] | 15:37 | |
kanzure | http://www.genengnews.com/gen-articles/do-it-yourself-bioengineers-bedeviled-by-society-s-paranoia/3383/ | 15:38 |
kanzure | http://arcollab.org/2011/08/31/molecular-biotechniques-workshop/ | 15:41 |
-!- Jaakko94 [~Jaakko@host81-157-105-60.range81-157.btcentralplus.com] has quit [Quit: Nettalk6 - www.ntalk.de] | 15:48 | |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has joined ##hplusroadmap | 15:55 | |
kanzure | oh god tito updated openpcr.org's template to that lotus wordpress theme.. | 16:00 |
kanzure | wonder what happened to the $512 price tag | 16:00 |
-!- augur [~augur@208.58.5.87] has quit [Remote host closed the connection] | 16:03 | |
kanzure | mac's site is up.. | 16:05 |
kanzure | http://cofactorbio.org/ | 16:05 |
jrayhawk | .com | 16:07 |
kanzure | damn it | 16:07 |
kanzure | yes that makes sense | 16:07 |
kanzure | also, singularity university now has created a 'singularity angels' group.. (angel fundin) | 16:07 |
kanzure | *funding | 16:07 |
-!- SDr [SDr@unaffiliated/sdr] has quit [] | 16:07 | |
kanzure | updated this.. http://openwetware.org/wiki/DIYbio/FAQ/News#Has_DIYbio_been_in_the_news.3F | 16:09 |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has quit [Ping timeout: 240 seconds] | 16:14 | |
-!- Tyrant [~Tyrant@c-76-21-1-190.hsd1.ca.comcast.net] has joined ##hplusroadmap | 16:19 | |
kanzure | fenn: can you get the thiel foundation application form from todd please? | 16:19 |
kanzure | i'm interested in looking at it | 16:20 |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has joined ##hplusroadmap | 16:23 | |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has quit [Client Quit] | 16:25 | |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has joined ##hplusroadmap | 16:26 | |
-!- diane [~diane@cpe-69-205-70-55.rochester.res.rr.com] has joined ##hplusroadmap | 16:29 | |
-!- augur [~augur@129.2.129.32] has joined ##hplusroadmap | 16:42 | |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has quit [Ping timeout: 252 seconds] | 16:47 | |
kanzure | https://github.com/fpletz/kernelroll | 16:58 |
-!- mayko [~mayko@71-22-217-151.gar.clearwire-wmx.net] has quit [Remote host closed the connection] | 17:01 | |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has joined ##hplusroadmap | 17:09 | |
-!- delinquentme [~delinquen@c-76-125-242-200.hsd1.pa.comcast.net] has joined ##hplusroadmap | 17:19 | |
delinquentme | Tyrant: something you needed? | 17:19 |
delinquentme | i've got skype but not installed | 17:20 |
-!- devrandom [~devrandom@gateway/tor-sasl/niftyzero1] has quit [Quit: leaving] | 17:23 | |
-!- Tyrant [~Tyrant@c-76-21-1-190.hsd1.ca.comcast.net] has quit [Ping timeout: 260 seconds] | 17:43 | |
-!- mayko [~mayko@71-22-217-151.gar.clearwire-wmx.net] has joined ##hplusroadmap | 18:29 | |
-!- chimera [~eudoxia@r190-135-73-58.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap | 18:44 | |
-!- chimera [~eudoxia@r190-135-73-58.dialup.adsl.anteldata.net.uy] has quit [Client Quit] | 18:44 | |
-!- eudoxia [~eudoxia@r190-135-73-58.dialup.adsl.anteldata.net.uy] has quit [Ping timeout: 260 seconds] | 18:46 | |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has quit [Ping timeout: 276 seconds] | 18:48 | |
-!- delinquentme [~delinquen@c-76-125-242-200.hsd1.pa.comcast.net] has quit [Quit: ChatZilla 0.9.87 [Firefox 6.0.2/20110902133214]] | 19:32 | |
-!- strages [~makers@c-71-207-222-170.hsd1.al.comcast.net] has quit [Ping timeout: 260 seconds] | 19:52 | |
-!- strages [~makers@c-71-207-222-170.hsd1.al.comcast.net] has joined ##hplusroadmap | 19:53 | |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has joined ##hplusroadmap | 21:12 | |
kanzure | maze generator in html5/js http://mazery.sjackson.net/ | 21:30 |
kanzure | huh my fan is running high after that | 21:31 |
-!- augur [~augur@129.2.129.32] has quit [Remote host closed the connection] | 22:01 | |
-!- diane [~diane@cpe-69-205-70-55.rochester.res.rr.com] has quit [Quit: Lost terminal] | 22:11 | |
-!- augur [~augur@208.58.5.87] has joined ##hplusroadmap | 22:28 | |
-!- mayko [~mayko@71-22-217-151.gar.clearwire-wmx.net] has quit [Quit: Leaving] | 22:29 | |
-!- foucist [~foucist@ps14150.dreamhost.com] has joined ##hplusroadmap | 22:33 | |
-!- Technicus [~nomadicus@66-190-14-100.dhcp.stpt.wi.charter.com] has quit [Ping timeout: 260 seconds] | 22:43 | |
CryptoQuick | kanzure: still here? :) | 23:14 |
kanzure | yes | 23:15 |
-!- foucist [~foucist@ps14150.dreamhost.com] has quit [Quit: leaving] | 23:17 | |
CryptoQuick | shit, forgot to put notifications back on :P | 23:19 |
CryptoQuick | so, bud, I did a bunch more research after we talked | 23:19 |
CryptoQuick | http://cq.1k.ru/cad_research/ | 23:19 |
CryptoQuick | some of that is yours :) | 23:19 |
CryptoQuick | also, I've done a bit of coding on my renderer | 23:20 |
CryptoQuick | but yeah, I need to know where to really start on representing a parametric solid | 23:20 |
CryptoQuick | I think it'd probably be best to port what we need from ESOLID, see how they tackled some issues, you know... but bring it over in coherent pieces, yeah? | 23:21 |
kanzure | by representing do you mean representing or rendering | 23:21 |
kanzure | funkhouser.pdf i wish all my pdfs were named funkhouser | 23:21 |
CryptoQuick | right! : | 23:22 |
CryptoQuick | :D | 23:22 |
CryptoQuick | rendering on hold for now | 23:22 |
CryptoQuick | I mean, like, having a mathematical description of our parametric solid INSIDE the computer (the FILES are INSIDE the COMPUTER!?!?!?) | 23:22 |
kanzure | okie dokie | 23:22 |
kanzure | the esolid representation is a little weird in a few places | 23:23 |
CryptoQuick | no kidding!? :P | 23:23 |
kanzure | after writing all that code for esolid i was reading that thesis the other day and the section on the representation of trimming curves kinda blew my mind | 23:23 |
kanzure | it's a little weird in a few places because it's C++ and lots of things are passed by pointers or stored as a pointer to an index and then given a fucking awful variable name | 23:23 |
CryptoQuick | haha, I saw that! | 23:24 |
kanzure | btw, in lolcad there's a STEP generator (in lolcad.py) that will be replaced by https://github.com/mpictor/StepClassLibrary | 23:24 |
kanzure | i think tpaviot made one out of SCL that generates python classes so i'm willing to give that a shot.. | 23:24 |
kanzure | but my generator was hard-coded and is broken | 23:25 |
CryptoQuick | mmh | 23:25 |
kanzure | the typical way to write a STEP generator is to write an EXPRESS schema, run it through your EXPRESS compiler, and get back a bunch of STEP classes and read/write tools that you then integrate into your codebase | 23:25 |
kanzure | (which is what SCL is about) | 23:25 |
CryptoQuick | well, I think at this point I would like to learn about how this is really done, then I'd like to write a javascript version of a simple CSG and/or boundary modeler | 23:26 |
CryptoQuick | "simple" | 23:26 |
kanzure | okay | 23:26 |
CryptoQuick | also, kanzure, what do you think of this? | 23:32 |
CryptoQuick | http://pyjs.org/ | 23:32 |
CryptoQuick | I wouldn't mind using that, really, as it meets the condition that this runs clientside. | 23:32 |
CryptoQuick | it would also mean we could use proper classes~ | 23:32 |
kanzure | uh python can also run client side | 23:33 |
CryptoQuick | since when | 23:33 |
CryptoQuick | I mean in the browser, you dork :P | 23:33 |
kanzure | i see | 23:34 |
CryptoQuick | important! | 23:34 |
CryptoQuick | but yes, actually, this could meet a couple ends, I suppose | 23:34 |
kanzure | i think i used pyjs once | 23:34 |
CryptoQuick | I mean, if we create a nice tidy working CAD kernel, in python, it could be used with PyJS... | 23:34 |
CryptoQuick | and also be used for your needs whatever they might be on the desktop | 23:35 |
kanzure | well, my idea for lolcad was to have a client that was just an interpreter | 23:37 |
kanzure | and then a window would show the changes in real time | 23:37 |
CryptoQuick | a python interpreter? | 23:38 |
kanzure | yeah | 23:38 |
CryptoQuick | so, same workflow as OpenSCAD | 23:38 |
kanzure | except without the inherent UI coupling | 23:39 |
kanzure | autocad has autolisp.. so why not a modern scripting language | 23:39 |
kanzure | heekscad has heekspython actually | 23:39 |
CryptoQuick | and then I could do as others have done with OpenSCAD, tacking on a UI | 23:39 |
kanzure | yeah | 23:41 |
CryptoQuick | well, why hasn't this nice simple tidy cad kernel been made yet, kanz? :P | 23:42 |
kanzure | because i'm busy as all hell and nobody's been buggng me | 23:42 |
kanzure | bugging | 23:43 |
CryptoQuick | well shit dude | 23:43 |
CryptoQuick | >:T | 23:43 |
CryptoQuick | kanzure: we need to set up some administration infrastructure first, if you don't mind. | 23:47 |
CryptoQuick | where are we putting our code, where are we keeping documentation, that sort of thing. | 23:47 |
CryptoQuick | I know you have answers to this already, but Google is being less than helpful. | 23:47 |
kanzure | how about lolcad.git | 23:47 |
kanzure | what username would you like | 23:47 |
* CryptoQuick <-- | 23:48 | |
kanzure | ok one sec | 23:48 |
kanzure | username? | 23:48 |
kanzure | oops | 23:48 |
kanzure | email? | 23:48 |
* CryptoQuick <-- @gmail.com | 23:48 | |
kanzure | ok your username is crpytoquick@diyhpl.us | 23:51 |
kanzure | see pm for password | 23:51 |
CryptoQuick | did you really spell it like that? | 23:51 |
kanzure | fuck | 23:51 |
kanzure | um.. | 23:51 |
kanzure | no | 23:51 |
kanzure | i spelled it right | 23:51 |
CryptoQuick | cool :) | 23:51 |
CryptoQuick | so, uhm | 23:54 |
CryptoQuick | I ssh into this or no? | 23:54 |
kanzure | yep | 23:55 |
kanzure | ssh cryptoquick@diyhpl.us | 23:55 |
CryptoQuick | yeah | 23:55 |
CryptoQuick | I figured it out now | 23:55 |
CryptoQuick | :P | 23:55 |
kanzure | don't tell me you don't.. know how to use ssh | 23:55 |
CryptoQuick | haha, I was having a derp moment, shut up | 23:56 |
CryptoQuick | passwd changed~ | 23:57 |
kanzure | oh shit.. i screwed up | 23:57 |
kanzure | i meant to type newuser, did adduser instead | 23:57 |
kanzure | oh well. you can have a full account.. | 23:57 |
kanzure | newuser sets you up on pinyshell, a hosting service for community git | 23:58 |
CryptoQuick | buh | 23:58 |
CryptoQuick | do what you must to fix it :P | 23:58 |
kanzure | nah go ahead | 23:58 |
CryptoQuick | ? | 23:58 |
kanzure | just don't abuse the box i'm not made of political power | 23:58 |
CryptoQuick | i'll try not to put cp on it | 23:59 |
CryptoQuick | but no guarantees | 23:59 |
--- Log closed Mon Sep 12 00:00:01 2011 |
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!