--- Day changed Thu Jul 23 2009
01:34 < ybit> fenn, kanzure: in your copy of pythonOCC, in line 6 of pythonOCC/src/gp.py ..do you have "import _gp"?
01:35 < ybit> because when trying to run pymates.py: ImportError: No module named _gp
01:35 < ybit> a  File "/home/heath/builds/pythonOCC-wo0.2/src/OCC/gp.py", line 6, in <module>
01:35 < ybit>     import _gp
01:35 < ybit> ImportError: No module named _gp
02:52 < kanzure> Tim Schmidt might be worth stalking. He used to be on the multimachine mailing list, and is now on the reprap-michigan mailing list, and is also some sort of nix hacker who complained about fstab-rsync.
02:52 < kanzure> timschmidt@gmail.com
02:53 < kanzure> ybit: I had to comment out that line.
02:53 < kanzure> ybit: actually, I didn't comment out pythonOCC/src/gp.py because I had gp.py somewhere else (it was installed to my pythonpath)
02:55 < kanzure> ybit: one disparity between what you've done and how fenn and I have done it is that fenn and I both used scons to get things working
02:56 < kanzure> ybit: and at the end of our installations, we did some links like "sudo ln -s /usr/share/pyshared/OCC /usr/lib/python2.6/dist-packages/OCC" - check http://adl.serveftp.org/dokuwiki/pythonocc
02:58 -!- Irssi: #hplusroadmap: Total of 21 nicks [0 ops, 0 halfops, 0 voices, 21 normal]
03:16 < genehacker2> I think he's part of the core reprap group kanzure
04:34 < maradydd> do gitwiki/djangit support templates the way mediawiki does?
05:28 -!- marainein [n=marainei@220.253.25.222] has joined #hplusroadmap
07:06 < kanzure> maradydd: I don't know. django has its own custom templating language.
07:09 < kanzure> wtf http://lesswrong.com/lw/4g/eliezer_yudkowsky_facts/
07:10 -!- SmoKeyCastle [n=ingo@122.111.57.103] has joined #hplusroadmap
07:13 < kanzure> hello SmoKeyCastle
07:13 < SmoKeyCastle> hey
07:14 < kanzure> what's up?
07:14 < SmoKeyCastle> not much.
07:15 < kanzure> what brings you around these parts? :)
07:16 < SmoKeyCastle> Not much. stalking mainly.
07:16 < kanzure> yeah we do that frequently in here
07:16 < kanzure> anyone in particular I can help you stalk?
07:16 < SmoKeyCastle> haha
07:17 < SmoKeyCastle> I have been having a look at your website
07:17 < SmoKeyCastle> pretty interesting
07:17 < kanzure> it will consume your soul
07:17 < kanzure> unless you're careful.
07:18 < SmoKeyCastle> rTMS how much of an effect does that have?
07:18 < kanzure> it depends on the coil geometry and the 'strength' of the magnetic field
07:18 < kanzure> there was an article in the new york times once about it. the journalist totally lost his ability to draw kittens, for instance.
07:19 < kanzure> (for the duration of the stimulation)
07:20 < SmoKeyCastle> so the theory goes that there is a certain frequency that our brains run at peak performance on and this magnetic coil cap is able to enforce that? sounds like it can have some side effects though.
07:20 < kanzure> no, that's not the theory at all
07:20 < SmoKeyCastle> oh ok I must have misread
07:20 < kanzure> think of it as if it's MRI except it's punching your neurons
07:26 -!- marainein [n=marainei@220.253.25.222] has quit ["Ex-Chat"]
07:28 < kanzure> discworld?
07:34 < kanzure> the usual channels: ##electronics, #math, #not-math, #physics, #not-physics, #biology, ##chemistry, ##neuroscience, #space, #robotics
07:35 < kanzure> other channels: #diybio, #emc, #emc-devel, #cam, ##SL4, #mireprap, anything else?
07:35 < kanzure> oh I guess there's #engineering but they don't talk about engineering in there
07:35 < SmoKeyCastle> makes sense
07:35 < kanzure> of course I'm not mentioning all of the software channels since that's obvious
07:36 < kanzure> there's also #madscientists but nobody goes there anymore
07:51 < CIA-44> skdb: kanzure * r3ced699e2801 /pymates/tests.py: using a cone instead of a cylindrical peg in the unit test
07:53 < fenn> terry pratchett rubs me the wrong way
07:54 < fenn> all his stuff is supposed to be funny but it just ends up as annoying
07:54 < kanzure> someone in #yaml claimed to have worked on the discworld mud
07:54 < kanzure> MUD
07:55 < fenn> i guess that explains the RPG-ish tutorial
07:55 < kanzure> nah all (good) programmers are obsessed with RPGs
07:57 < fenn> "Eliezer Yudkowsky has no need for induction or deduction. He has perfected the undiluted master art of duction."
07:57 < fenn> i KNEW it!!
07:57 < kanzure> what more evidence of a cult do you need
07:57 < fenn> quack quack quack
07:58 < kanzure> s/duction/ducktion/
08:02 < fenn> "Eliezer Yudkowsky thought he was wrong one time, but he was mistaken."
08:02 < fenn> "An AI programmed to maximize utility will tile the Universe with tiny copies of Eliezer Yudkowksy."
08:02 < SmoKeyCastle> lol
08:02 < fenn> that one hurts
08:05 < fenn> "Eliezer Yudkowsky's map is more accurate than the territory."
08:37 < drazak> ergh
08:37 < drazak> wtf is with this cell line
08:38 < drazak> they've died on us twice
08:53 -!- drazak is now known as draz|lab
08:56 -!- SmoKeyCastle [n=ingo@122.111.57.103] has left #hplusroadmap []
09:38 < CIA-44> skdb:  * rce6891e8f431 /codingstyle.txt: more
09:40 < CIA-44> skdb: kanzure * re58899bf4df2 /pymates/tests.py: added rotation to the unit test. bus commit.
09:40 < CIA-44> skdb: kanzure * r0786d0ee583c /pymates/tests.py: comments
09:40 < CIA-44> skdb: kanzure * r121f9ce9e779 /pymates/pymates.py: comments, threw in show_interface_points (where did it go?)
09:40 < CIA-44> skdb: kanzure * r5b9ee64fb396 /pymates/pymates.py: fixed- show_interface_points never really did vanish
09:40 < CIA-44> skdb: kanzure * r6702b3f2c621 /codingstyle.txt: Merge branch 'master' of ssh://bryan@adl.serveftp.org/var/www/skdb
09:53 -!- timschmidt_ [n=chatzill@68.40.5.224] has joined #hplusroadmap
09:54 < timschmidt_> hey.  internet went wonky for a minute
09:54 < kanzure> hey timschmidt_ 
09:55 < kanzure> so, you mentioned an interest in automated manufacturing and bootstrapping
09:55 < kanzure> are you familiar with gingery?
09:55 < timschmidt_> yes
09:55 < kanzure> are you familiar with apt-get?
09:55 < timschmidt_> of course
09:55 < kanzure> well then
09:55 < kanzure> you're cought up
09:55 < kanzure> the git repo is in the /topic
09:55 < kanzure> have fun
09:56 < fenn> yarr
09:56 < timschmidt_> So I ask for a widget and said software tells me what to make, in what order, so that by the end I have what I need to make a widget?
09:56 < fenn> didnt we have some sort of 'overview of skdb' page (before you typed all over it at least)
09:56 < kanzure> timschmidt_: yes
09:56 < timschmidt_> excellent
09:57 < timschmidt_> I've wished for said software in the past.  Thank you for working on it.
09:57 < fenn> timschmidt_: it's amazing how hard it is to explain that concept to some people
09:58 < timschmidt_> I tried to get Zach to build something similar into thingiverse...  I think it looked too hard for him to be interested though
09:58 < kanzure> timschmidt_: thingiverse sucks. sorry.
09:58 < kanzure> these guys need to learn about packaging standards
09:58 < kanzure> and what a dependency is, for that matter
09:59 < kanzure> one subproject is "djangit", which might be a suitable frontend
09:59 < timschmidt_> meh.  It's one spot on a continuum.  Anything that makes all of this easier.  Agreed about the dependencies though.
10:00 < kanzure> you should check out the repository asap :)
10:00 < fenn> i ought to caution that the software doesn't actually do anything yet
10:01 < timschmidt_> heh
10:01 < timschmidt_> ok
10:03 < timschmidt_> One of the projects I've been working on is a wrench-buildable repstrap...  you guys may be interested.  Allow you to skip a few dependencies.
10:04 < timschmidt_> In other words, you go to the hardware store, buy some common materials, assemble them at home with a wrench (and perhaps some cutting with a hacksaw), and you've got a CNC machine.  No complex tools necessary.
10:07 < kanzure> yes, skipping dependencies is always nice
10:07 < kanzure> I think that's called a cutset in graph theory
10:10 < fenn> timschmidt_: what about cast concrete designs?
10:10 < kanzure> ok fenn and I were wondering about how to do the packages
10:10 < kanzure> each package could potentially contain either a CAD file or a python script to generate the parametric CAD file
10:11 < kanzure> but also metadata describing what the package contains (for an equivalent to "apt-cache search" and "apt-cache show" I guess)
10:11 < kanzure> (also describing dependencies and related information)
10:11 < timschmidt_> fenn: one of the other projects I'm involved with (the multimachine) has done some research into cast concrete designs...  but AFAIK, none of our members has built one.
10:18 < timschmidt_> about packaging: CAD files are only part of the problem.  The multimachine, for instance, uses already-cast parts (engine blocks) for most of it's construction, and requires few tools, but a _very_ skilled machinist to assemble.  CAD files are important, but instructions for humans 'in the loop' should also be top-level objects.
10:20 < kanzure> automatically generated instructions are planned
10:20 < fenn> one of the goals is to define an abstract language which could be translated to either english human-readable instructions or robot control code (gcode for example)
10:20 < kanzure> gcode->english is nearly impossible, but some-other-format->gcode might be
10:21 < timschmidt_> right
10:25 < Smari> gcode is messed up.
10:26 < kanzure> fenn: re: pcr-example, I guess the question to ask (so that we can figure out the answer to your packaging question) is where exactly in the toolchain different sequences are generated by skdb.
10:27 < kanzure> does anyone have ideas for how to do a unit test that must test for correct rotation? what should I check?
10:41 < CIA-44> skdb: kanzure * ra7ecf6ffabc6 /pcr-example: added to pcr-example
10:46 < fenn> add some points to the object which you can get coordinates for and check their coordinates
10:46 < kanzure> how do you add points to an object itself?
10:47 < kanzure> you mean to a TopoDS_Shape?
10:50 -!- mage2 [n=mage@66.179.208.36] has quit [Remote closed the connection]
10:52 < kanzure> import OCC.SGEOM #not found. How do you get pythonOCC/src/SWIG_sr..darwin/SGEOM.i to compile?
10:53 < fenn> isnt there some occ thing that combines multiple TopoDS_Shapes?
10:53 < fenn> SGEOM is some other thing
10:53 < fenn> it's actually called GEOM
10:53 < fenn> not included with OCC
10:54 < kanzure> the parametric tests seem to require it
10:54 < fenn> "the salome GEOM module"
10:54 < fenn> is that really necessary just for parametric shapes?
10:55 < fenn> http://sourceforge.net/projects/salomegeometry/
10:57 < fenn> looks exactly like OCC to me http://www.salome-platform.org/home/presentation/geom/
10:57 < kanzure> well at least OCC.GEOMImpl is required
10:58 < kanzure> sgeom/src/GEOMImpl/ hm.
10:59 < fenn> are you trying to get PAF working?
11:00 < kanzure> yes
11:01 < fenn> from OCC.SGEOM import GEOM_Solver,GEOM_Parameter
11:01 < kanzure> what about it?
11:01 < fenn> well, that's what it needs from salome
11:02 < kanzure> what is "it"?
11:02 < fenn> PAF
11:02 < kanzure> no, OCC.PAF.Context requires some other things as well
11:02 < kanzure> GEOMImpl, for instance
11:02 < fenn> anyway those functions seems to pull in all the rest of the geometry module
11:05 < kanzure> for now I'm still working on figuring out how to test rotation
11:05 < kanzure> I think if I translate and rotate a point, the point's location would be updated (?)
11:06 < fenn> rotating a point doesnt do much
11:06 < kanzure> if you translate it first, it does do something
11:06 < kanzure> just like if you translate and rotate an object
11:06 < fenn> ok but what does that prove
11:06 < kanzure> that rotation works?
11:07 < fenn> but doesnt your data format specify rotation about the origin (i.e. no translation first)
11:08 < kanzure> are we talking about different things?
11:08 < kanzure> I just want a way to check if a point on the object is properly rotated
11:09 < fenn> carry on
11:09 < kanzure> so I need to figure out how to access a point on the object
11:09 < kanzure> preferably not the interface-point
11:10 < fenn> right
11:10 < kanzure> point.Transform(OCC.gp.gp_Trsf()) seems to do the trick. but it just applies a transform. maybe I'll just set it to a certain point on the object.
11:11 < fenn> what is your "object"'s type?
11:11 < kanzure> in the unit test, one of the objects is a cone
11:11 < fenn> TopoDS_Shape?
11:11 < kanzure> yes
11:11 < fenn> how do you attach points to that?
11:14 < fenn> i dont see any OCC point object in the unit test
11:14 < fenn> oops nm, old version
11:15 < fenn> still there's no points attached to the cone
11:15 < kanzure> right
11:15 < kanzure> I'm looking up how to do this
11:16 < kanzure> oh you have to use TopExp_Explorer
11:16 < fenn> how do i say 'the default to_yaml'?
11:16 < kanzure> does the class inherit from something that has a to_yaml() ?
11:16 < fenn> i want to delete the to_yaml if there's no yaml_repr
11:17 < fenn> but i cant really do that inside the to_yaml function
11:17 < kanzure> what is the context
11:17 < fenn> so instead i was thinking something like def to_yaml(): if not yaml_repr: return default_to_yaml
11:18 < kanzure> does hasattr() return True for methods?
11:18 < fenn> yes
11:18 < kanzure> does that solve your problem?
11:18 < fenn> no
11:18 < fenn> look at to_yaml in FennObject
11:19 < fenn> right now it returns __repr__ if there's no yaml_repr, but that's wrong
11:19 < kanzure> how about yaml.YAMLObject.to_yaml()
11:21 < kanzure> you know what, it probably does some recursion bullshit
11:21 < kanzure> because yaml.YAMLObject.to_yaml() supposedly tests for the existence of to_yaml() already in your class right?
11:25 < fenn> no actually i'm getting a "tag not specified" error
11:29 < kanzure> oh god OCC for java http://jcae.sourceforge.net/occjava-doc/org/jcae/opencascade/jni/TopExp_Explorer.html
11:30 < fenn> it's only to be expected
11:41 -!- mage2 [n=mage@66.179.208.36] has joined #hplusroadmap
--- Log opened Thu Jul 23 11:42:02 2009
11:42 -!- kanzure [i=bryan@dhcp-84-36.me.utexas.edu] has joined #hplusroadmap
11:42 -!- Irssi: #hplusroadmap: Total of 22 nicks [0 ops, 0 halfops, 0 voices, 22 normal]
11:42 -!- Irssi: Join to #hplusroadmap was synced in 1 secs
11:46 < kanzure> OCC.Sweep
11:48 < fenn> i cant figure out how to call BRepPrimAPI_MakeSweep
11:48 < fenn> but BRepPrimAPI_MakePrism works fine
11:49 < kanzure> hahah
11:49 < fenn> well, except for that missing bit in the center since it's not actually a solid
11:49 < kanzure> import OCC.Units
11:49 < fenn> but i think i can fix that with a union
11:52 < kanzure> see opencascade/OpenCASCADE6.3.0/ros/src/Units/
11:52 < kanzure> see opencascade/OpenCASCADE6.3.0/ros/src/UnitsAPI/Units.data
11:52 < ybit> http://tech.groups.yahoo.com/group/diy_3d_printing_and_fabrication/
11:53 < CIA-44> skdb:  * rb5b3312f8632 / (processes.yaml skdb.py): more data represenation crap
11:53 < ybit> 10:00 < timschmidt_> One of the projects I've been working on is a  wrench-buildable repstrap...
11:53 < ybit> link?
11:55 < kanzure>  ROTATION STIFFNESS: kg*mm²/(s²*rad)
11:55 < ybit> 10:17 < fenn> one of the goals is to define an abstract language which could be  translated to either english human-readable instructions or robot  control code (gcode for example)
11:56 < ybit> are you talking about lojban?
11:56 < kanzure> no
11:56 < fenn> ybit: http://builders.reprap.org/2009/02/another-off-shelf-linear-bearing.html
11:56 < ybit> there was a guy working on a #lojban compiler last time i checked
11:56 < fenn> not what i mean by 'language'
11:56 < fenn> more like 'formalism'
11:56 < ybit> ..last i checked was almost a year ago
11:57 < fenn> probably some mix of yaml and python to continue the theme
11:58 < ybit> i knew you were thinking along those lines but it would still be cool to speak lojban :P
11:58 -!- maradydd [n=mlp@88.147.30.78] has quit [Read error: 110 (Connection timed out)]
11:58 < ybit> thanks for the link
12:04 < ybit> 02:53 < kanzure> ybit: one disparity between what you've done and how fenn and  I have done it is that fenn and I both used scons to get  things working
12:04 -!- xp_prg [n=xp_prg3@99.2.31.217] has joined #hplusroadmap
12:04 < ybit> maybe later i will install debian somewhere on another machine and dedicate it to skdb work, that might help the workflow as well..
12:10 < kanzure> ybit: it shouldn't matter that you're using gentoo
12:11 < ybit> it would improve workflow, instead of translating debian commands to gentoo 
12:11 < ybit> echo "deb http://www.opennovation.org/ubuntu jaunty main contrib non-free" | sudo tee -a /etc/apt/sources.list
12:11 < ybit> sudo apt-get install libopencascade-dev scons python-psyco swig checkinstall
12:11 < wrldpc2> "you must be a current cryonics suspension member"  @http://www.universalimmortalism.org/  uhh what does this mean exactly?
12:15 < kanzure> you must have alcor and your insurance company talking with each other 
12:16 -!- splicer [n=patrik@h189n3c1o261.bredband.skanova.com] has joined #hplusroadmap
12:26 < fenn> organic packaged cookie dough: rational?
12:34 < fenn> "Cyclone Power Technologies attempts to quash internet rumours that his company is developing a robot that runs its biomass-powered engine by digesting fallen soldiers on the battlefield"
12:36 < kanzure> so are those rumors being squashed by the biomass-powered engine that "is most definitely not a robot"?
12:36 < fenn> Los Angeles is trying to "enforce regulations that require condom use in porn films"
12:36 < fenn> that seems about as misguided as gun control
12:36 < kanzure> hm maybe I should think twice before moving there
12:40 < kanzure> http://adl.serveftp.org/lab/screw.skdb.tar
12:53 < ybit> kanzure: you were talking about finding people which might be able to help, something like this might be of use: http://www.jibble.org/piespy/
13:00 -!- wrldpc2 [n=benny@pool-173-48-253-182.bstnma.fios.verizon.net] has quit []
13:07  * fenn grumbles about kwalify
13:07 < fenn> why isnt there a basic "url" type?
13:08 < fenn> surely url is as rigorously defined as "int" or "string"
13:12 < fenn> "Diego Gambetta and Steffen Hertog suggest that in the psyche of engineers lurks something that makes them more predisposed to terrorism..."
13:21 < CIA-44> skdb: kanzure * r99aac6c039d8 /pymates/ (pymates.py tests.py): made some changes to the tests
13:21 < CIA-44> skdb: kanzure * rb0e0f93e44a9 / (processes.yaml skdb.py): Merge branch 'master' of ssh://bryan@adl.serveftp.org/var/www/skdb
13:21 < kanzure> http://github.com/kanzure/shelltrance/blob/0abbce2af8c699781fef383aa374458b55840e74/trance-skdb
13:21 < kanzure> don't kill me
13:24 < ybit> argh, why spiders jump on my computer screen? they are so interesting, but major distractions
13:24 < ybit> i can't kill it either because it's such a cool spider
13:25 < ybit> i'm jealous he can jump about 20x his body size and climb on wallls
13:26 < ybit> not so jealous that he's bound to be smashed
13:28 < fenn> ybit dip it in liquid nitrogen
13:29 < Smari> kanzure, that's just silly. :)
13:29 < ybit> if only some were available now
13:29 < kanzure> Smari: what's silly?
13:29 < fenn> kanzure keeps wanting me to write this crap
13:29 < kanzure> fenn: because you're clueless
13:36 < Smari> kanzure, the stuff you pasted... it's very cool, but still a bit crazy.
13:37 < kanzure> Smari: fenn and I are trying to come up with the packaging format example
13:39 < Smari> mhm
13:39 -!- wrldpc2 [n=benny@pool-173-48-253-182.bstnma.fios.verizon.net] has joined #hplusroadmap
13:40 < kanzure> http://github.com/kanzure/shelltrance/tree/master
13:40 < Smari> Simple: .tar.gz containin a manifest file in a standardized format which refers to the rest of the files in the package. Files in the package then also come in some standardized formats.
13:41 < kanzure> yes, but then what about screw.yaml ?
13:41 < fenn> Smari: the problem we're running into is there's no clear separation between metadata and data
13:41 < fenn> in order to thoroughly test compatibility you need all of the data about the object
13:42 < fenn> so the idea is to do a first order "do the units of the attributes match up" test
13:42 < kanzure> I was hoping that there would be some way to do a "search" over the packages so that you can maybe see which one you want to download
13:42 < kanzure> supposedly that would be searching over metadata, right?
13:42 < fenn> and then test all of those wrt domain-dependent tags
13:43 < fenn> github is really slow today
13:44 < fenn> shelltrance is still wrong
13:44 < kanzure> do you refer to the git repo or the shelltrance.txt file
13:45 < fenn> trance-skdb
13:45 < fenn> it shouldnt be #! anything
13:45 < fenn> if it's a transcript of an interactive session
13:45 < fenn> i'll fix it
13:45 < fenn> fsov "fix"
13:46 < kanzure> http://github.com/kanzure/shelltrance/blob/1e1c5a88776c128587a193c5a4954acf7c149edb/trance-skdb
13:46 < kanzure> there?
13:53 < fenn> blah
13:53 < fenn> http://adl.serveftp.org/git/gitweb.cgi?p=shelltrance.git;a=blob_plain;f=trance-skdb;hb=79bfe7a68e1c50b7d82ac9adc004a570b741ac5c
13:53 < fenn> that took way too long
13:54 < kanzure> hm. skdb/process.yaml shows up when searching google for filetype:yaml
13:55 < kanzure> #15~
13:56 < kanzure> cool, queries like this also lead to it: "arbor milling" "abrasive jet" "internal broaching"
13:58 < kanzure> hm.. http://mirror.sit.wisc.edu/pub/CPAN/authors/RECENT-1W.yaml
13:59 < kanzure> Smari: how would you make it?
13:59 < kanzure> Smari: I mean, when should it download python versus just data etc. ?
14:05 < Smari> kanzure, I'd say keeping it non-executable is a good idea for security reasons.
14:06 < kanzure> Smari: er, what I am asking about is architecture-related
14:06 < kanzure> if you had a "super" fablab, how should this work?
14:06 < Smari> Of course we can't do much to prevent second-order execution (i.e., devices being described which contain payloads)
14:06 < kanzure> how do you search for which packages you want to download?
14:08 < kanzure> er, or how do you do it now (without skdb or anything like that)
14:09 < kanzure> from what I can tell most people don't know how to search for whether or not a machine or part exists that does what they are thinking of
14:09 < kanzure> fenn: maybe part compatibility shouldn't be related to search..
14:09 < kanzure> so the metadata will just then be a list of packages, their dependencies, names, descriptions. :/
14:11 < kanzure> xpkg: it's what xkcd would be called if it was a package manager instead of a webcomic
14:11 < kanzure> http://peter.michaux.ca/articles/xpkg-generic-package-managment-software
14:13 < kanzure> doesn't seem to do much
14:15 < fenn> http://adl.serveftp.org/lab/fenn/screw.skdb.tgz
14:15 < timschmidt_> ybit: http://builders.reprap.org/2009/04/first-example-of-wrench-build-machine.html
14:18 < kanzure> also, btw, smart apparently has a plugin-abble "policies" that rank solutions to dependency problems: http://labix.org/smart
14:20 < fenn> Smari: packages should be signed; can you explain why allowing execution is bad?
14:23 < kanzure> fenn: did you fix metadata.yaml?
14:23 < fenn> no
14:23 < timschmidt_> nearly all Linux packages contain scripts which are executed upon installation / removal.
14:23 < timschmidt_> So long as the packages come from a trusted source, there's no problem with that
14:24 < timschmidt_> (they are, after all, installing things which you will execute on your own at a later date)
14:24 < kanzure> xpkg would be neat if somebody was actively maintaining it
14:25 < ybit> timschmidt_: nice :)
14:27 < fenn> hm. i've been told to ask smari what is "group theory" and what it has to do with package management
14:28 < ybit> by whom?
14:28 < fenn> a little bird
14:28 < kanzure> tweet tweet tweet
14:28 < ybit> :)
14:30 < timschmidt_> ybit: I'm proud of the pillow block bearings.  They're stupid simple (which is a feature!) - just a 608 skate bearing, JB Weld, and a 3/4" pipe hanger.  It seems obvious, but replaces a part that costs $25+ each with something that costs ~$0.50.
14:30 < fenn> what's the JB weld for?
14:31 < kanzure> fenn: be sure to look at http://labix.org/smart eventually
14:31  * fenn looks
14:32 -!- timschmidt_ [n=chatzill@68.40.5.224] has quit [Read error: 104 (Connection reset by peer)]
14:33 -!- timschmidt [n=chatzill@68.40.5.224] has joined #hplusroadmap
14:33 < timschmidt> http://builders.reprap.org/2009/03/1-no-tools-pillow-block-bearing.html
14:33 < timschmidt> I used two bearings and a clamp there.  After testing, I found that only one was necessary, and a flat surface works as well as a clamp.
14:36  * ybit is off to work
14:38 < fenn> i'm doing basically the same thing on my gingery lathe motor pulley, but i dont use any jb-weld
14:39 < fenn> it's also just a strap of metal with holes drilled in it, not a 3/4" pipe hanger
14:39 < kanzure> http://labix.org/python-constraint hehe
14:39 < fenn> i was thinking about putting a strip of rubber between the bearing and hanger
14:45 < fenn> dpkg -S /usr/bin/rename ought to find _something_
14:48 < fenn> think it's from util-linux, but still seems odd that it doesn't show up
14:49 < fenn> $ dpkg -S /bin/ls
14:49 < fenn> coreutils: /bin/ls
14:50 < fenn> $ dpkg -S /usr/bin/ddate
14:50 < fenn> util-linux: /usr/bin/ddate
14:51 < kanzure> so rename comes with ddate :)
14:51 < fenn> no it doesnt
14:52 < fenn> i'm just demonstrating that dpkg -S works for the package util-linux
14:52 < fenn> still dont know where it comes from
14:55 < kanzure> bkchem - python-based chemical structures editor
14:58 < kanzure> pkpgcounter - computes number of pages or quantity of ink needed to print documents
15:00 < kanzure> python-apt - Python interface to libapt-pkg
15:00 < kanzure> heh python-beagle
15:00 < timschmidt> someone asked about concrete as a building material earlier...  I remembered that the OpenLathe people have been working with it for some time.
15:00 < timschmidt> http://openlathe.wikidot.com/
15:00 < fenn> never heard of that; how long has it been around?
15:01 < timschmidt> unsure
15:02 < fenn> compare and contrast: http://fennetic.net/machines/bootstrap http://openlathe.wikidot.com/lathe-bed
15:03 < fenn> looks like nobody's touched it since jun 2008
15:04 < kanzure> python-gastables - compressible flow gas table modules for Python
15:05 < kanzure> heh graphical interface? blah
15:05 < timschmidt> I think most activity happens in their Yahoo group...
15:06 < kanzure> the fact that they are using a yahoo group tells you something about them
15:06 < kanzure> I joined the list a while back
15:06 < timschmidt> certainly.
15:06 < timschmidt> Most machinists aren't super-computer savy though.
15:06 < kanzure> or I tried to. this might have been one of the groups that doesn't allow new members.
15:06 < kanzure> timschmidt: you're hanging out with the wrong machinists
15:06 < timschmidt> lol
15:08 < timschmidt> I think it's mostly a symptom of the times.  There were a LOT more machining jobs in this country 50 years ago.
15:08 < genehacker2> but then all the equipment went to china
15:08 < timschmidt> well, mostly the Chinese built their own equipment, but yes.
15:09 < genehacker2> some of it did
15:09 < kanzure> yee-haw, let's ship a 5 ton CNC center over the ocean
15:09 < timschmidt> right.  Which is why they built their own.
15:09 < fenn> i saw a show on discovery channel where they disassembled an entire factory in germany, labeled everything, shipped to china and reassembled, all in like 3 months
15:09 < kanzure> there's a company that does this
15:09 < kanzure> Matrix Services
15:11 < genehacker2> the only thing I've heard of going overseas was a gear cutter
15:13 < fenn> this was a steel mill
15:13 < fenn> not small
15:13 < genehacker2> wow
15:14 < kanzure> http://neuroimaging.scipy.org/site/index.html
15:15 < CIA-44> skdb:  * r6c5370e42cf9 /occ_shell.py: thought i committed usage info earlier
15:19 < katsmeow-afk> the steel mill the Terminater movie was shot in was packed for shipping to China the day after shooting ended
15:20 < katsmeow-afk> for a couple reshoots, like the liquid terminator changing shapes in the "molten steel", they had to do mockups, the factory didn;t exist by then
15:21 < katsmeow-afk> for things the Chinese cannot ship, they buy entirely: for oil sands in Canada, they also bought the pipeline to the coast and the right-of-way it sits on
15:22 < timschmidt> For a sense of humor, they bought The Onion.
15:22 < katsmeow-afk> so in some places in BC Ca, you haveto go thru China to get to the rest of BC
15:22 < CIA-44> skdb:  * r9bd74bbef3a9 / (screw.py skdb.py): split screw into its own file; to be included in a package
15:23 < timschmidt> http://www.theonion.com/content/columnists/well_ive_sold_the_paper_to
15:23 < genehacker2> really?
15:25 < genehacker2> heh speaking of liquid metal terminators and modular robots
15:27 < CIA-44> skdb:  * r37ff5d05de51 / (screw.py tests.py): fix for namespace changes
15:28 < genehacker2> The same sensors could have detected the weightlessness of falling, and the pre-programmed panic sequence could then have ordered the T-1000 cells to form into a parachute. The hot air rising from the molten steel would have lifted it up out of harm's way.
15:28 < genehacker2> http://www.islandone.org/MMSG/9609lego.htm
15:29 < fenn> genehacker2: it also could have just killed sarah when it had the chance
15:29 < fenn> not very terminator-y
15:29 < genehacker2> bah
15:30 < katsmeow-afk> yeas, that whole scene wasn't believeable,, except they establkished before that when multiple damage is done, the terminator can't respond
15:30 < genehacker2> its just something to keep in mind when building our robot armies...
15:31 < katsmeow-afk> yeas, that defect was fixed in the tv series of liquid metals
15:31 < genehacker2> meh it's movie physics
15:32 < katsmeow-afk> the woman liquid terminator seemingly wasn't damageable, and remianed sentient/sapient during attacks, and could reach 20ft in either direction, and carry electricity, unlike the quasi-liquid terminator in T3
15:36 < kanzure> today I seem to be able to do anything except work on pymates
15:39 < CIA-44> skdb:  * r8cdfa3a7e731 /screw.py: compatibility assertion
15:55 < CIA-44> skdb:  * r9f36149a52a4 / (skdb.py units.py yamlcrap.py): split yaml and units into their own files
15:55 < fenn> might want to test that
15:57 < CIA-44> skdb:  * rc85a402ecf7e /skdb.py: gah. forgot to save after deleting. units and yaml in their own files for real now
16:00 < draz|lab> welp, the 293FT's are transfected
16:06 < CIA-44> skdb:  * ra8c6aaf5751d / (skdb.py thread.py): split Thread into its own file - still slightly broken (cant find !thread for some reason)
16:38 < wrldpc2> My family runs a machine shop for the record.  Just putting that out there.  They're in Lewiston, ME -- not the most current machine tech but they have a full compliment of lathes, a forge, etc.
16:57 < QuantumG> bastard
17:06 < kanzure> wrldpc2: Are they generally agreeable people?
17:06 < wrldpc2> absolutely
17:07 < wrldpc2> I very much would like to help in any way that I can.  Just saying.
17:07 < kanzure> there are many ideas that come to mind
17:07 < wrldpc2> They can do custom stuff.
17:09 < kanzure> if I sent you schematics, CAD files, etc., would they be able to tell me if they would want some money to make it or not?
17:09 < fenn> i think they'd certainly be able to answer the question
17:13 < wrldpc2> Let me get you an email ..
17:13 < wrldpc2> sales@down-east.com
17:13 < wrldpc2> That's my uncle Steve
17:14 < wrldpc2> tell him Ben Peterson (Bob's son) his brother referred you.
17:14 < wrldpc2> I have NO idea what their CAD capabilities are ..  heh
17:14 < wrldpc2> the shop is equipped with 1950s tech latest
17:14 < wrldpc2> there is a computer there
17:14 < wrldpc2> lol
17:14 < wrldpc2> schema sure
17:15 < wrldpc2> I'm just an idiot.
17:16 < kanzure> 1950s tech is good enough
17:16 < fenn> not for making use of g-code
17:16 < kanzure> probably better..
17:16 < wrldpc2> I gotta jet.  They will DEFINITELY help you.  If I knew what I was talking about I might be able to get some of the work done gratis so keep that in mind.
17:16 < wrldpc2> go xhumans go!
17:17 -!- wrldpc2 [n=benny@pool-173-48-253-182.bstnma.fios.verizon.net] has quit []
17:19 < genehacker2> ok wrldpc2
17:19 < genehacker2> compile this:
17:19 < genehacker2> http://www.thingiverse.com/thing:816
17:20 < kanzure> "The spectra of photon emission detected from the palm skin span from 500 to 700 nm, with primary and secondary emission peaks at 630–670 nm and 520–580 nm, respectively"
17:22 < genehacker2> yup
17:22 < genehacker2> we glow red
17:22 < genehacker2> and green
17:28 < genehacker2> http://en.wikipedia.org/wiki/Immersion_lithography
17:30 < fenn> genehacker2: the rolamite bearing is pretty cool
17:31 < fenn> genehacker2: i think you'd have to make all the pieces tapered to take up the slop but still be printable
17:31 < genehacker2> tapered?
17:32 < fenn> hm. actually i think it won't work because the silicone will stretch, and rolamite band has to be stiff
17:32 < fenn> otherwise the band will scuff against the outer race
17:32 < genehacker2> I designed the band so that it's in tension
17:32 < fenn> i know
17:33 < fenn> what keeps the inner rollers from moving outward?
17:33 < genehacker2> tension
17:33 < fenn> the band will stretch
17:33 < fenn> you'd have to add nylon thread to the band or something
17:33 < genehacker2> there is also ~1 mm of clearance between the band and the bearing race
17:34 < fenn> you see what i'm saying?
17:34 < genehacker2> not really
17:34 < genehacker2> so put nylon thread into it to decrease the modulus?
17:35 < genehacker2> oh shoot that might be a problem
17:36 < genehacker2> yup looks like I need to increase the modulus
17:36 < genehacker2> http://www.youtube.com/watch?v=y-BK_iEJ7Go does this avoid that problem?
17:37 < fenn> genehacker2: http://imagebin.org/56933
17:38 < genehacker2> yeah I see what you mean
17:38 < genehacker2> now how can I compensate for that?
17:38 < fenn> increase the modulus
17:38 < genehacker2> any other way?
17:39 < fenn> i'm sort of bullshitting.. i mean add a fiber that's flexible but stiff in tension
17:39 < genehacker2> that's what increasing the modulus means
17:39 < fenn> i guess
17:40 < genehacker2> perhaps increase the clearance?
17:41 < genehacker2> now then how to get the nylon in there
17:42 < fenn> also i'm not sure how it's better than just a roller bearing
17:42 < genehacker2> it can be replicated
17:42 < genehacker2> doesn't require high tolerances
17:43 < fenn> why can't you just print a tapered roller bearing?
17:43 < genehacker2> CAD it up
17:43 < fenn> i dont have any cad
17:44 < fenn> well, i'm lazy, so there
17:44 < genehacker2> why tapered?
17:44 < fenn> to take out the slop
17:44 < genehacker2> ????
17:44 < fenn> so you dont need "high" tolerances
17:44 < genehacker2> in rolamites, the roller need not be perfectly round
17:44 < kanzure> eh? http://www.minitage.org/
17:46 < genehacker2> hmmm...
17:46 < genehacker2> kanzure's in the ADL
17:47 < genehacker2> kanzure can you access the stratasys?
17:49 < fenn> genehacker2: couldnt you do some sort of flexible chain made from plastic? then you could use the stratasys
17:49 < fenn> oh wait that would defeat the purpose of the rolamite
17:50 < fenn> um. how about just a thin band of abs
17:50 < fenn> polypropylene would be best i think
17:50 < fenn> abs is too brittle nevermind
17:52 < genehacker2> it might not be
17:53 < genehacker2> just need to figure out how much strain it can take
17:53 < genehacker2> meh
17:53 < genehacker2> probably wouldn't last too long
17:53 < fenn> abs has crap fatigue strength too
17:53 < genehacker2> damn
17:53 < genehacker2> definately won't work
17:55 < genehacker2> now then how to encapsulate the nylon
17:55 < genehacker2> perhaps some sort of 2 step mold process?
17:56 < fenn> when i was thinking about how to make my own timing belts i came up with this procedure:
17:57 < fenn> make an internal mold (arbor) and smear a thin layer of silicone over it
17:57 < fenn> then wrap it evenly with nylon, and apply more silicone
17:57 < fenn> repeat for desired thickness/strength
17:57 < genehacker2> sorta what I mean by 2 step molding
17:57 < fenn> then slide the tube off the arbor by your desired belt thickness, and slice with a razor blade
17:58 < fenn> the demolding might need to be more complicated since the belt will be under tension due to shrinkage and being wrapped with a certain amount of tension
17:59 < fenn> that might not matter so much for a smooth belt though
17:59 < genehacker2> shrinkage from molding
17:59 < fenn> shrinkage from the silicone curing
17:59 < genehacker2> silicone doesn't shrink that much
17:59 < genehacker2> 0.0001 something
17:59 < fenn> not much but it doesnt have to be much
18:00 < genehacker2> if you mean shrinkage due to stretching that's easy
18:00 < genehacker2> I already take that into acount
18:06 < genehacker2> interesting
18:06 < genehacker2> we might be able to make dielectric elastomer actuators
18:07 < genehacker2> looks like the suck though
18:15 < genehacker2> hmm...
18:15 < genehacker2> Nylon fiber has a young's modulus of 3000 MPa
18:23 < genehacker2> http://users.tamuk.edu/kfldp00/research/Papers/dissertation.pdf
18:25 < genehacker2> this may be harder than it looks
18:33 < genehacker2> you know I'm starting to think that the best way to make this might be to use a special silicone-nylon fiber extruder
18:35 < genehacker2> heh
18:35 < genehacker2> nylon fibers in silicone happens to be an advanced composite
19:24 < kanzure> $icanhas millingmachine
19:24 < kanzure> huh "get" is an unused command name 
19:25 < genehacker2> ???
19:25 < genehacker2> 99999999999 get?
19:28 < kanzure> nget, wget, kget, debget, cget, but no "get"
19:28 < genehacker2> really?
19:28 < genehacker2> in what?
19:29 < kanzure> my system
19:29 < genehacker2> ugh
19:30 < genehacker2> headphone malfunction
19:30 < genehacker2> btw you know that acoustic control thing in diybio, do you have the paper for it?
19:30 < genehacker2> I have a use for it
19:30 < kanzure> no, but I have an older one where they did the same damn thing but without resonance
19:31 < genehacker2> I want the resonance one
19:31 < genehacker2> btw did the video of it working ever play for you?
19:32 < genehacker2> also was it with droplets or channels?
19:34 < kanzure> it was with piezos and channels but when you activated the piezos waves would form and go over a boundary, so I guess that's sort of like a drop
19:34 < genehacker2> was in a channel though?
19:34 < fenn> libwww-perl: /usr/bin/GET
19:36 < genehacker2> I want to repurpose it and make a acousto-fluidic FM receiver
19:36 < genehacker2> that outputs a 1 or 0 for if a certain tone is present
19:38 < kanzure> FM?
19:39 < genehacker2> frequency modulated
19:39 -!- draz|lab is now known as drazak
19:40 < genehacker2> have a bunch of resonators for each control tone and some associated fluidic circuitry
19:41 < genehacker2> it's for controlling my army of steampunk robots
19:43 < fenn> with your steam calliope of death
19:43 < genehacker2> yeah
19:44 < genehacker2> or a gramophone or something
19:44 < kanzure> http://nixos.org/nixos/screenshots/nixos-kde42-1.png what's that image in the middle?
19:45 < genehacker2> penguins designing an airplane?
19:47 < genehacker2> or that program thing
19:47 < fenn> "the penguins of madagascar"
19:48 < fenn> maybe
19:58 < kanzure> huh? the list at the bottom http://en.wikipedia.org/wiki/Architecture_description_language
20:10 < kanzure> 17:49:30 omg/jblake: "Controllers" change a system dynamically, as it is operating.                  
20:10 < kanzure> 17:49:44 omg/jblake: "Managers" plan changes to a system, before it is operated.                     
20:10 < kanzure> 17:50:35 omg/jblake: "Resources" are things that are needed for "projects", which are things that you
20:11 < kanzure> are trying to build.
21:29 < kanzure> fast & furious was over too fast and too furiously
21:31 < kanzure> apparently ERP is all done by psych majors
21:37 < genehacker2> pysch majors just do stuff at low abstraction levels
21:45 -!- genehacker2 [n=noko@cpe-67-9-157-136.austin.res.rr.com] has quit []
22:03 -!- genehacker2 [n=noko@cpe-67-9-157-136.austin.res.rr.com] has joined #hplusroadmap
22:18 < ybit> 17:57 < genehacker2> if you mean shrinkage due to stretching that's easy
22:18 < ybit> 17:57 < genehacker2> I already take that into acount
22:18 < ybit> 18:04 < genehacker2> looks like the suck though
22:18 < ybit> the hell are you guys talking about ;)
22:18 < genehacker2> a 3d printable rotary rolamite bearing I designed
22:21  * ybit was quite aware
22:21 < ybit> unless that's a new phrase for it i've yet to come across ;)
22:21 < genehacker2> oops
22:21 < genehacker2> typing erros
22:22 < genehacker2> *error
22:22 < ybit> :)
22:22 < genehacker2> that wasn't intentional
22:24 < ybit> right, thus the humor
22:25 < genehacker2> but it turned into humor
22:25 < ybit> now this i wasn't aware of, the maufacturing processes reference guide only documents ~300 processes as dictacted/suggest by the Manufacturing Consortium Industral Members who helped with the book
22:26 < ybit> fenn^
22:26  * ybit wants a list of all processes
22:26 < ybit> or at least those that weren't listed
22:57 -!- wrldpc2 [n=benny@pool-72-72-22-132.bstnma.east.verizon.net] has joined #hplusroadmap
23:03 < fenn> wah.
23:03 < fenn> "all" processes eh?
23:04 < fenn> try the vatican
23:06 < ybit> heh, i know it's a little much to ask, but dreaming is allowed in irc
23:10 < ybit> so i've been told in #defocus
23:10 -!- strages [n=strages@c-68-62-216-5.hsd1.al.comcast.net] has quit [Read error: 104 (Connection reset by peer)]
23:11 -!- strages [n=strages@c-68-62-216-5.hsd1.al.comcast.net] has joined #hplusroadmap
23:45 -!- wrldpc2 [n=benny@pool-72-72-22-132.bstnma.east.verizon.net] has quit []
23:47 < ybit> hmm
23:47 < ybit> guess i'll be needing to setup a personal server soon for pirated content
23:47 < ybit> http://filebin.ca/uwpzkb/m83-weowntheskyudachiremixremastered.mp3
23:47 < ybit> for stuff like that
23:48 < ybit> music, papers, videos
23:48 < ybit> other
23:48 < ybit> _not_ associated with http://co.de or heath.im
23:48 < QuantumG> http://www.quantumg.net/David_Kushner_-_Masters_Of_Doom.txt
23:48 < QuantumG> about the only ebook I've warez
23:49 < QuantumG> cause you just couldn't buy it anywhere at the time
23:49 < ybit> wrldpc2 has grown a beard again
23:49 < ybit> i just don't understand it, how people can do that
23:50 < QuantumG> grow a beard?
23:50 < ybit> yeah, i'm kind of jealous
23:50 < ybit> i only get peach fuzz
23:50 < QuantumG> oh.  how old are ya?
23:50 < ybit> 23
23:50 < QuantumG> I'm 32 and it'd take me about a month to grow a beard that my bearded friends would mock
23:51 < ybit> that doesn't give me hope ;)
23:51 < katsmeow-afk> not a big thing, i'm 53 and can't grow a beard either, but then i know why
23:51 < QuantumG> maybe 6 months to get something I don't know how to care for
23:51 < ybit> katsmeow-afk: :P
23:51 < ybit> katsmeow-afk: 53. wow. you have to be the coolest 53yr old i know
23:51 < ybit> well...
23:51 < ybit> nope, definitely the coolest
23:52 < katsmeow-afk> someone suggested a hypercube using the nearly 200 68331 cpu i have, but i can't figure out how to make that much bandwidth happen,,,, or why i need such a thing
23:52 < ybit> the other guy i know doesn't use irc, but he's taught me quite a bit about electronics
23:52 < QuantumG> on one of the space channels I'm on there's a 60yo guy... and another guy on there is a 12yo
23:52 < ybit> heh
23:59 < ybit> http://filebin.ca/gbbdug/DISCOVERY-SOINSANE.mp3