--- Day changed Sun Sep 06 2009
00:06 < CIA-32> skdb: kanzure * r cef5c3d /unittests/test_geom.py: fuse some more boxes and test their volumes
00:07 < katsmeow> ybit, you need crt electron guns?
00:14 < CIA-32> skdb: kanzure * r e6ca3cf / (30 files in 2 dirs): move unit tests around
00:19 -!- mason-l [n=x@202-89-188-136.static.dsl.amnet.net.au] has quit ["Ex-Chat"]
00:20 < katsmeow> ybit, speak up before he busts the crts up
00:23 < CIA-32> skdb: kanzure * r abc7ef9 / (doc/proposals/graphtheory.py graphtheory.py): move graph theory code out of the main directory and into the abyss
00:27 < kanzure> todo: write unit tests for my assembly part overlap detection methods.
00:27  * kanzure sleeps
00:33 < fenn> kanzure: torque is N*m/rev, not N*m as is often said
00:33 < fenn> dumbasses who think angles don't exist
00:49 -!- mason-l [n=x@202-89-188-136.static.dsl.amnet.net.au] has joined #hplusroadmap
00:50 < genehacker> kanzure does sciencemadness work for you?
00:55 < genehacker> oh wait nvm
01:04 -!- xp_prg [n=xp_prg3@208.54.15.112] has quit ["This computer has gone to sleep"]
01:07 < fenn> kanzure: BRepAlgoAPI_Fuse(brick1.shapes[0], brick2.shapes[0]) takes just as long as a boolean common operation (10 seconds)
01:46 -!- xp_prg [n=xp_prg3@67.188.6.132] has joined #hplusroadmap
01:57 -!- genehacker [n=noko@pool-173-57-48-104.dllstx.fios.verizon.net] has quit [Read error: 113 (No route to host)]
02:49 -!- katsmeow is now known as katsmeow-afk
03:28 -!- kso512 [n=kso@cpe-66-68-153-104.austin.res.rr.com] has quit [Read error: 110 (Connection timed out)]
03:45 -!- bct [n=bct@S0106001c1041555b.ed.shawcable.net] has joined #hplusroadmap
04:21 -!- xp_prg [n=xp_prg3@67.188.6.132] has quit ["This computer has gone to sleep"]
08:01 < ybit2> katsmeow-afk: yes i do
08:02 < ybit2> as the fenn mentioned once before though, the choke coil may be wrong for sem
08:02 < ybit2> don't know just yet
08:13 < kanzure> hello bct 
08:30 < kanzure> fenn: should I bother writing the unit tests then?
08:30 < kanzure> also it doesn't take that long on my laptop to do Fuse
08:30 < kanzure> 2.6sec to run through test_geom.py
08:31 < kanzure> actually the unittest module claims 0.746sec
08:34 < kanzure> and BRepAlgoAPI_Common takes less than 1sec to run on two boxes
08:36 -!- |kardan| [n=kardan@p54BE369B.dip.t-dialin.net] has joined #hplusroadmap
08:40 < kanzure> fenn: what's wrong here?
08:40 < kanzure> opt = brick1.options(brick2)[0]
08:40 < kanzure> opt.interface1.part is None
08:54 -!- kardan| [n=kardan@p54BE454F.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)]
09:10 < kanzure> load_CAD() makes that not be None? wtfc
09:35 < kanzure> grr the bullshit from STEPImporter isn't using stdout or stderr
09:36 < kanzure> but maybe it's because somebody forgot to turn off the debug flags on the configure script when compiling opencascade for the dot debs?
09:49 -!- wrldpc2 [n=benny@ool-ad03fe34.dyn.optonline.net] has joined #hplusroadmap
09:54 -!- goonie_ is now known as goonie
10:05 < kanzure> pythonOCC-0.3/src/samples/Level2/DataExchange$ python import_step_single.py &>/dev/null
10:05 < kanzure> well that's one way to get rid of the annoying output
10:05 < kanzure> although that gets rid of absolutely everything
10:08 < kanzure> heh things work for once: brick1.interfaces.index(opt.interface2)
10:09 < fenn> i don't see how it could be neither stdout or stderr and still show up
10:09 < kanzure> it's stdout but in the library
10:09 < kanzure> python only controls python stdout apparently
10:09 < kanzure> unless there's a way to "lock" the "global" stdout or something?
10:10 < fenn> i dunno
10:10 < kanzure> *but in the OCC library
10:51 < Utopiah> not sure it was shared here earlier http://www.wikid.eu/ "WikID, the Industrial Design Engineering Wiki"
10:55 < kanzure> doesn't seem to have much content
10:59 < fenn> what content there is doesn't seem to be useful
11:00 < fenn> negatively useful in fact
11:00 < fenn> "a coffee mill changes beans to coffee, a chair supports (prevents one from becoming tired), and a poster provides information (decreases uncertainty)"
11:01 < fenn> is it just me or are all those sort of non-sequitur
11:01 < kanzure> fenn: http://adl.serveftp.org/yaml-test.py
11:01 < fenn> or plain wrong even
11:04 < fenn> why is it a 'str object'?
11:07 < kanzure> because it's a string in the yaml snippit
11:07 < kanzure> hmm
11:07 < fenn> why is it a string?
11:08 < fenn> it should be 'Blah' or 'instance'
11:08 < kanzure> no
11:08 < kanzure> look at the yaml string
11:09 < kanzure> yaml will read "interfaces:" and see that it has two values in the list
11:09 < kanzure> in the sequence
11:09 < fenn> oh
11:09 < kanzure> so it's correct
11:09 < fenn> well, bad kanzure
11:09 < kanzure> no this is fine
11:09 < kanzure> this is supposed to show that post_init_hook is not being called
11:09 < kanzure> and now I can play around with FennObject.from_yaml quickly
11:09 < fenn> my_blah = yaml.load("!blah\ninterfaces:\n- !interface hello\n- !interface hello2
11:10 < kanzure> why should it be an interface?
11:10 < fenn> otherwise you get an error because it's a string
11:10 < kanzure> the point is that *any* list doesn't load
11:10 < kanzure> no you get an error because it doesn't have an attribute
11:10 < kanzure> that you supposedly added in post_init_hook
11:10 < kanzure> see?
11:10 < fenn> please make modification and test before continuing bitching
11:10 < kanzure> the string error is *correct*
11:11 < kanzure> it's true that str has no attribute "part"
11:11 < kanzure> Blah.post_init_hook sets str.part
11:12 < fenn> 'foo'.part = 'bad kanzure'
11:13 < fenn> AttributeError: 'str' object has no attribute 'part'
11:13 < kanzure> hey wtf is "yield" doing in yaml/constructor.py
11:13 < fenn> it's a generator..
11:13 < kanzure> why
11:13 < kanzure> why does it have to be a generator
11:13 < fenn> because they work better on huge files
11:13 < kanzure> blah
11:14 < fenn> yaml is designed to work with streams i think
11:14 < fenn> make up your mind
11:14 < kanzure> ?
11:14 < kanzure> my mind about what?
11:14 < fenn> the other day you were bitching about how people didn't use generators
11:15 < fenn> and how it broke your 15k bookmarks
11:15 < kanzure> but now it's breaking something else because they didn't bother to implement post_init_hook for us
11:15 < kanzure> this is really something that they should have done
11:15 < fenn> are you sure?
11:15 < kanzure> yes if it's a generator
11:15 < fenn> i'm not totally sold on the idea of post_init_hook
11:15 < kanzure> what would be the other way to do it?
11:15 < fenn> in __init__ somehow
11:16 < kanzure> same problem
11:16 < kanzure> in __init__ it's only "partial" apparently
11:16 < kanzure> and then later yaml throws the lists into the dictionary or something
11:16 < kanzure> see, it's because of the generator
11:16 < fenn> you mean it fills out the lists later, but you have a pointer to an empty list in the attribute
11:16 < fenn> instead*
11:16 < kanzure> yep
11:22 < kanzure> construct_yaml_seq in yaml/constructor.py is basically: data = []; yield data; data.extend(self.construct_sequence(node))
11:23 < fenn> seems out of order doesn't it
11:24 < fenn> should be: data.extend(self.construct_sequence(node)); yield data
11:24  * kanzure tries
11:24 < fenn> maybe they did it that way for some reason
11:24 < kanzure> like loading big files?
11:24 < fenn> right
11:26 < kanzure> doesn't seem to fix it
11:26 < kanzure> although I'm doing FennObject.yaml_loader.construct_yaml_seq = my redefined method
11:29 < kanzure> oh it's not even called
11:30 < kanzure> FennObject.yaml_loader.__bases__[4]
11:30 < kanzure> >>> FennObject.yaml_loader.__bases__
11:30 < kanzure> (<class 'yaml.reader.Reader'>, <class 'yaml.scanner.Scanner'>, <class 'yaml.parser.Parser'>, <class 'yaml.composer.Composer'>, <class 'yaml.constructor.Const
11:30 < kanzure> ructor'>, <class 'yaml.resolver.Resolver'>)
11:31 < kanzure>  22 FennObject.yaml_loader.__bases__[4].construct_yaml_seq = _construct_yaml_seq
11:31 < kanzure> didn't work. hrm.
11:31 < fenn> good.
11:31 < kanzure> somehow I can't seem to overwrite construct_yaml_seq
11:32 < kanzure> yaml.constructor.Constructor.construct_yaml_seq = _construct_yaml_seq
11:32 < kanzure> also doesn't work
11:33 < kanzure> running out of ideas
11:34 < kanzure> http://adl.serveftp.org/yaml-test.py
11:37 < kanzure> AttributeError: 'FakeInterface' object has no attribute 'extra_attr'
11:40 < CIA-32> skdb: fenn * r a98b061 /packages/threads/threads.py: minor changes
11:40 < CIA-32> skdb: fenn * r ca37fa5 /core/yamlcrap.py: appears to fix the null sequence upon loading problem
11:40 < fenn> yay sleepycode
11:42 < kanzure> ok
11:42 < fenn> you will have to ask my hypothalamus wtf i just did
11:43 < kanzure> not your hippocampus?
11:43 < kanzure> fenn fails brains
11:43 < fenn> a hippocampus says what?
11:43 < kanzure> you set deep=True on construct_mapping in from_yaml
11:43 < fenn> yes but why?
11:44 < fenn> does that construct the node tree entirely first?
11:44 < kanzure> yes, it's what happens when you specify your own __setstate__ too
11:45 < kanzure> oopsies
11:46 < fenn> if self.deep_construct: for dummy in generator: pass
11:46 < kanzure> from copy import copy, deepcopy; from skdb import *; lego_pack = load_package("lego"); lego_pack.load_data(); brick1 = deepcopy(lego_pack.parts[0]); brick2 = deepcopy(lego_pack.parts[0]); brick1.load_CAD()
11:46 < kanzure> AttributeError: 'Lego' object has no attribute 'load_CAD'
11:46 < kanzure> oh I didn't import geom
11:46 < kanzure> catastrophe inverted
11:47 < fenn> brick1.interfaces[0] == brick1 ?
11:47 < kanzure> >>> brick1.interfaces[0].part
11:47 < kanzure> Lego(name=2x2 round brick, num_pegs=4, num_holes=4)
11:47 < kanzure> >>> brick1.interfaces[0].part == brick1
11:47 < kanzure> True
11:47 < kanzure> yay
11:49 < fenn> do i get a raise?
11:49 < kanzure> you wrote nine characters
11:49 < kanzure> no
11:50 < kanzure> you must write 100klines/week to get a raise
11:50 < fenn> that should mean i get a raise and a cookie too
11:50 < fenn> for not writing piles and piles of crap
11:50 < kanzure> mr. anderson, a company is like a centipede
11:50 < kanzure> if any one of its employees has a problem, the company has a problem
11:50 < kanzure> do I make myself clear?
11:50 < fenn> all legs and no brain?
11:50 < fenn> a poisonous monster?
11:50 < kanzure> well not if you're in the myzeopta period
11:50 < fenn> where's my rabbit
11:51 < kanzure> silurian period
11:51 < fenn> who comes up with these names?
11:52 < kanzure> old men who still play with dinosaurs
11:52 < fenn> "Murchison named the sequences for a Celtic tribe of Wales, the Silures, following the convention his friend Adam Sedgwick had established for the Cambrian."
11:52 < fenn> obviously i don't know my welsh tribes
11:52 < kanzure> too bad he didn't name them after whales
11:52 < fenn> i thought it sounded rather tolkein
11:52 < kanzure> Shamoo, Shamploo, 
11:53 < fenn> tolkien*
11:53 < kanzure> "in the shamoo period, giant man-eating centipedes roamed the cerniferous forests"
11:54 < fenn> right, so now that the problem is solved, you have to continue to stampede in some direction
11:54  * fenn suggests the todo list
11:54 < kanzure> well I was actually working on something else originally
11:54 < fenn> oh
11:54 < kanzure> so I'll be getting back to that
11:55 < kanzure> unless you have some particular suggestions ..
11:56 < fenn> no, i'm going to continue making fun of the product design wiki
11:57 < kanzure> maybe you should be more constructive
11:57 < kanzure> and add a graph with boxes and arrows
11:57 < kanzure> I'm sure they'll love it
11:59 < kanzure> sounds like they already have at least one of those
12:03 < fenn> and no manufacturing step anywhere on it
12:04 < fenn> is it bedtime yet?
12:04 < kanzure> didn't you just wake up?
12:05 < fenn> i woke up at 11
12:05 < fenn> pm
12:05 < kanzure> oh it was me that just woke up
12:06 < fenn> why did you wake up in the middle of the night?
12:06 < kanzure> I didn't. went to sleep at 12 and woke up at 8.
12:10 < fenn> um. what's up with the latest email to OM?
12:12 < kanzure> J. Andrew Rogers?
12:13 < kanzure> btw JAR follows Eugen Leitl around on the net a bit
12:13 < kanzure> he is an annoying crypto guy
12:13 < fenn> no, you forwarding something you just wrote via eugen leitl?
12:14 < kanzure> you wrote it
12:14 < kanzure> he picked it up
12:14 < kanzure> thought I'd forward it since that text hasn't been on om yet
12:19 < fenn> JAR sure likes to make sweeping generalizations
12:19 < fenn> i just want one specific example of WTF he's talking about
12:26 < fenn> time python -c 'import skdb'
12:51 < kanzure> "overhumanism" ? http://www.antipope.org/charlie/blog-static/2009/09/chromed_jackboots.html#comments
13:19 < kanzure> fenn: AttributeError: 'module' object has no attribute 'Lego'
13:20 < kanzure> this is in import_package_classes
13:20 < kanzure> (Package.import_package_classes)
13:20 < kanzure> the line is: cls = getattr(module, class_name)
13:20 < kanzure> where module = packages/lego/lego.py
13:20 < fenn> i don't have Package.import_pakage_classes
13:21 < kanzure> it's the same thing as import_package_classes though
13:22 < kanzure> Package("lego")
13:22 < kanzure> load_package("lego")
13:24 < kanzure> dummy = Package(package)
13:24 < fenn> right, so i can find the packages directory
13:25 < fenn> which could be located anywhere
13:25 < kanzure> yes I know
13:25  * kanzure fixes
13:25 < fenn> i guess that should be using settings instead
13:25 < kanzure> yes
13:27 < kanzure> yay it works now
13:27 < kanzure> should I delete all this load_package, open_package crap?
13:28 < kanzure> technically it still works, but is less awesome
13:39 -!- strages [n=strages@c-68-62-216-5.hsd1.al.comcast.net] has quit [Read error: 54 (Connection reset by peer)]
13:49 -!- mason-l [n=x@202-89-188-136.static.dsl.amnet.net.au] has quit [Read error: 110 (Connection timed out)]
13:50 -!- mason-l [n=x@202-89-188-136.static.dsl.amnet.net.au] has joined #hplusroadmap
13:51 < kanzure>   File "/home/kanzure/code/skdb/packages/screw/screw.py", line 68, in __init__
13:51 < kanzure>     res = unit()
13:51 < kanzure>   File "/home/kanzure/code/skdb/packages/threads/threads.py", line 43, in tensile_area
13:51 < kanzure>     assert Unit(self.pitch).compatible('mm/rev')
13:51 < kanzure> how did it make that particular leap?
13:51 < kanzure> that's from test_package.py
13:52 < kanzure> oh.
13:52 < kanzure> unit = getattr(self.thread,k)
13:57 -!- genehacker [n=noko@pool-173-57-48-104.dllstx.fios.verizon.net] has joined #hplusroadmap
14:15 -!- xp_prg [n=xp_prg3@c-67-188-6-132.hsd1.ca.comcast.net] has joined #hplusroadmap
14:44 -!- dira [n=chatzill@86.99.40.183] has joined #hplusroadmap
14:45 < kanzure> hello dira
14:45 < dira> hey
14:45 < dira> what's up?
14:45 < dira> how's the collision thing going ?
14:46 < kanzure> opencascade's BRepAlgoAPI_Common and BRepAlgoAPI_Fuse methods apparently take too long when a brep model is anything more interesting than a giant box.
14:50 < dira> is it sorta mesh model Collision detection?
14:53 < kanzure> no
14:54 < kanzure> the method using common and fuse boolean operations is just a way of combining brep models, which are not meshes
14:54 < kanzure> and then I take the volume of the resulting meshes
14:54 < kanzure> ack
14:54 < kanzure> resulting breps :)
15:14 < kanzure> fenn: have you ever noticed that the option.connect() output is in the wrong order?
15:14 < kanzure> >>> option.connect()
15:14 < kanzure> connecting 2x2 round brick's stud  to 2x2 round brick's anti stud brick1[6] to brick2[2])
15:14 < kanzure> but if you investigate, brick2.interfaces[6].connected is what is set (not 2)
15:16 < kanzure> er nevermind
15:16 < kanzure> silly me
15:18  * dira watches kanzure talking to himself
15:31 -!- ve [n=a@94-193-95-252.zone7.bethere.co.uk] has quit [Read error: 110 (Connection timed out)]
15:35 -!- xp_prg [n=xp_prg3@c-67-188-6-132.hsd1.ca.comcast.net] has quit ["This computer has gone to sleep"]
15:49 < kanzure> what is a DSFiller?
15:49 < kanzure> http://adl.serveftp.org/lab/opencascade/doc/ReferenceDocumentation/ModelingAlgorithms/html/classBOPTools__DSFiller.html
15:55 < kanzure> hm I bet the skdb import time is spent making that units file every single time
15:57 -!- strages [n=strages@c-68-62-216-5.hsd1.al.comcast.net] has joined #hplusroadmap
16:02 < kanzure> http://exotk.org/ wrapper for opencascade from 2002
16:04 -!- genehacker [n=noko@pool-173-57-48-104.dllstx.fios.verizon.net] has quit [Read error: 148 (No route to host)]
16:13 < kanzure> http://opencascade.blogspot.com/2008/12/why-boolean-operations-are-so-sloooooow.html
16:16 < kanzure> http://opencascade.blogspot.com/2008/12/why-are-boolean-operations-so-sloooooow.html
16:16 < kanzure> er
16:16 < kanzure> http://opencascade.blogspot.com/2008/12/why-are-boolean-operations-so-sloooooow.html
16:17 < kanzure> well I fail at linking
16:17 < kanzure> there's a part 2 which is more revealing
16:19 -!- ve [n=a@94-193-95-252.zone7.bethere.co.uk] has joined #hplusroadmap
16:36 -!- ve [n=a@94-193-95-252.zone7.bethere.co.uk] has quit [Connection timed out]
17:03 -!- wrldpc2 [n=benny@ool-ad03fe34.dyn.optonline.net] has quit []
17:04 -!- wrldpc2 [n=benny@ool-ad03fe34.dyn.optonline.net] has joined #hplusroadmap
17:25 -!- nchaimov [n=cowtown@c-24-21-45-17.hsd1.wa.comcast.net] has joined #hplusroadmap
17:33 < kanzure> assembly_volume_estimate takes less than half a second for Package("lego").parts[0]
17:41 < CIA-32> skdb: kanzure * r 524be53 / (7 files in 4 dirs): fixed Package, wrote more geom tests, wrote more package tests
17:51 -!- genehacker [n=noko@pool-173-57-48-104.dllstx.fios.verizon.net] has joined #hplusroadmap
18:32 -!- wrldpc2_ [n=benny@ool-ad03fe34.dyn.optonline.net] has joined #hplusroadmap
18:32 -!- wrldpc2 [n=benny@ool-ad03fe34.dyn.optonline.net] has quit [Read error: 104 (Connection reset by peer)]
18:32 -!- wrldpc2_ is now known as wrldpc2
18:45 -!- genehacker2 [n=noko@pool-173-57-48-104.dllstx.fios.verizon.net] has joined #hplusroadmap
18:51 -!- wrldpc2 [n=benny@ool-ad03fe34.dyn.optonline.net] has quit [Read error: 110 (Connection timed out)]
18:52 -!- QuantumG [n=qg@206.248.137.22] has joined #hplusroadmap
18:52 < QuantumG> did ya miss me?
18:54 < kanzure> not much
18:54 < kanzure> what's up?
18:54 < QuantumG> was looking at single board computers last night
18:54 -!- wrldpc2 [n=benny@ool-ad03fe34.dyn.optonline.net] has joined #hplusroadmap
18:55 < QuantumG> and basically came to the conclusion that an eeepc did everything I wanted for cheaper.
18:55 < QuantumG> 'cept that it weighs 900g with battery
18:55 -!- genehacker [n=noko@pool-173-57-48-104.dllstx.fios.verizon.net] has quit [Read error: 113 (No route to host)]
18:57  * kanzure considers setting up a source code documentation militial front
18:57 < QuantumG> a low power sbc with wifi, usb and programmable IO.. that doesn't cost more than a netbook... 
18:57 < kanzure> anyone who writes undocumented code gets shot
18:57 < QuantumG> heh
18:58 < kanzure> what the fuck is this: http://adl.serveftp.org/lab/opencascade/BOPTools_PaveFiller.cxx
19:00 < QuantumG> cxx
19:00 < QuantumG> hmm
19:00 < kanzure> notice how all the comments are: // purpose:
19:00 < QuantumG> I've heard some people use that extension to me generated cpp
19:00 < kanzure> and hten it's totally blank
19:00 < QuantumG> so not finding comments in it is not that surprising
19:00 < kanzure> what's the point of a comment saying "purpose" and then not filling it in?
19:00 < QuantumG> 6.FF
19:01 < QuantumG> oh, I love people like that
19:01 < kanzure> no you don't
19:01 < QuantumG> "I read this paper about how to implement an algorithm and it had psuedocode in it, so I'm going to put that in my comments and then implement it"
19:02 < kanzure> IMHO giving a reference to a paper would be better than this shit
19:02 < QuantumG> "if you want to know what my code does, go read the paper.. and no, it's not publically available."
19:02 < kanzure> ah you mean they don't understand it
19:02 < QuantumG> variable names like "VF" 
19:02 < QuantumG> and functions like "PerformVF"
19:02 < QuantumG> what a dipshit
19:02 < kanzure> I should show you my professor's code
19:03 < kanzure> http://adl.serveftp.org/equer.lisp
19:03 < QuantumG> speaking of retarded programmers.. I checked out the opencog code the other day to see if there had been any change.. they seem to have added a few more kitchen sinks to the codebase but *still* no-one has joined any of it together into something that *does something*
19:04 < kanzure> imho I've never seen any indication that goertzel can code
19:04 < kanzure> so I've mostly ignored opencog
19:04 < QuantumG> he's an academic
19:04 < QuantumG> no-one else there can code either though
19:04 < kanzure> nah he's just timothy leary reincarnated .. as an academic.
19:04 < kanzure> did you see equer.lisp?
19:04 < QuantumG> and people who can code are scared off by them
19:05 < QuantumG> yes, but its lisp, so I kinda expected it to be horrid 
19:06 < kanzure> while I'm complaining,
19:06 < kanzure> someone wrote this:
19:06 < kanzure> compatibility_equations.cs
19:06 < kanzure> er
19:06 < kanzure> http://adl.serveftp.org/compatibility_equations.cs
19:06 < kanzure> in the lab that I work in.
19:07 < kanzure> she subsequently got a master's thesis for this "innovative" work
19:07 < kanzure> but if you look at it, it's just six conditionals
19:07 < QuantumG> to be honest, it's a master's thesis, I don't actually expect much more
19:08 < kanzure> if that's true then I should be able to write a few lines of bullshit and get a degree too
19:08 < QuantumG> master's thesis = baby's first unpublishable paper.
19:09 < kanzure> maybe you just have low standards
19:09 < QuantumG> I'm just cynical because I've hung out with post-grads :)
19:10 < QuantumG> I dunno how profs do it :)
19:10 < kanzure> wait till you meet the postdocs
19:11 < QuantumG> them too
19:12 < QuantumG> anyways, I'm thinking of building a blimp
19:12 < kanzure> out of what?
19:12 < QuantumG> where blimp is a generous description
19:12 < QuantumG> computer controlled floating garbage bag would be more accurate
19:13 < kanzure> you mean American Airlines?
19:13 < QuantumG> :)
19:14 < QuantumG> 1. buy some clothes line, garbage bags and rent a helium canister.. fly some tethered experiments, collect some data, etc.
19:14 < kanzure> what data?
19:15 < QuantumG> how long they stay up.. how accurate first principles are at predicting lifting weight, etc
19:15 < QuantumG> observe how bad wind affects it
19:17 < QuantumG> 2. attach a biprop radio controlled plane fuselage to it.. get a feel for how much power is needed to stabilize it.
19:17 < QuantumG> get some idea of the battery requirements.
19:18 < QuantumG> 3. work out a mass budget and redo the experiments with that amount of ballast.
19:19 < QuantumG> and that should give me enough information to work out the kinda of sensors, actuators and computational requirements to automate a stable platform in medium winds.
19:21 < kanzure> heh I'd be tempted to run a little regression analysis kernel hooked up to some altimeter or gps system
19:21 < kanzure> so that it would try to figure out a governing equation for what the hell to do to maximize height or to stabilize height at least (or whatever goal you want)
19:21 < QuantumG> yeah.. is gps sufficient?  
19:21 < kanzure> probably not
19:21 < kanzure> don't altimeters exist?
19:21 < kanzure> altitude meters?
19:21 < QuantumG> well, ultimately I want a rock steady platform
19:21 < QuantumG> for camera mounting
19:22 < QuantumG> cause that's fun :)
19:22 < kanzure> how do you expect to be rock study with unpredictable winds?
19:22 < kanzure> NWS doesn't get down to sub-meter resolution for wind contours :p
19:22 < QuantumG> well, as best as is acheivable
19:31 < genehacker2> a solar balloon is what you get when you use garbage bags
19:32 < genehacker2> have to deal with the pesky sun after all
19:33 < genehacker2> I don't have the exact numbers, but helium expands a lot when you heat it.(compared to say nitrogen)
19:34 < genehacker2> you might use some sort of mechanical active stabilization system
19:34 < genehacker2> or some quad thruster mechanism to stabilize your camera
19:35 < genehacker2> quite interested in this project
19:36 < QuantumG> I was going to use white garbage bags :)
19:37 < kanzure> why not black?
19:37 < QuantumG> and yeah, I have no idea how to control buoyancy
19:37 < QuantumG> no reason
19:37 < kanzure> black will cause more heat
19:37 < genehacker2> garbage bags suck
19:38 < genehacker2> they lose helium faster than say mylar or anything metal coated
19:38 < QuantumG> they're also cheap and available
19:38 < genehacker2> got a walmart nearby?
19:38 < QuantumG> no, but go on
19:39 < genehacker2> buy space blankets
19:39 < genehacker2> about a buck for ~2 meter squared piece of metal coated mylar
19:39 < genehacker2> anyway how big is your blimp going to be?
19:40 < QuantumG> aiming to lift about 1kg of payload
19:41 < genehacker2> ok so let's see here
19:42 < QuantumG> would love to do less.. for example, my radio controlled biprop is only 100g .. but sbcs are annoyingly expensive
19:42 < genehacker2> SBCs?
19:42 < QuantumG> single board computers
19:42 < QuantumG> using a netbook is massively more affordable
19:42 < genehacker2> how many channels is your biprop?
19:42 < genehacker2> a NETBOOK?
19:43 < genehacker2> in the air?
19:43 < QuantumG> tethered :)
19:43 < genehacker2> so the size of your balloon is big
19:43 < genehacker2> then you got to make it really aerodynamic
19:44 < genehacker2> you might want to make a zeppelin
19:44 < genehacker2> otherwise you're going to need to pressurize it
19:44 < QuantumG> you're quickly exceeding my capabilities
19:44 < genehacker2> wind is going to be a problem
19:44 < genehacker2> unless you live in a windless area
19:46 < genehacker2> how do you intend to turn garbage bags into a blimp?
19:46 < genehacker2> welding, taping, or gluing are your options
19:46 < QuantumG> dude, I intend to fill the bags with helium and tie them off
19:47 < QuantumG> this is why I said blimp was terribly generous :)
19:47 < genehacker2> oh
19:47 < genehacker2> a small party helium tank usually only fills about 1 garbage bag
19:48 < genehacker2> so a tethered balloon?
19:48 < genehacker2> why put a biprop on it?
19:48 < QuantumG> a CL class tank will fill 50 party balloons, or 7 of the garbage bags I'm looking at.. 
19:48 < genehacker2> hmmm...
19:49 < QuantumG> because that's the (broken) radio controlled plane I have :)
19:50 < genehacker2> and what's the purpose of putting a plane on there?
19:50 < genehacker2> if you want to maneuver it around...
19:50 < QuantumG> to see if it can drag around the balloons
19:52 < genehacker2> I'm betting not
19:52 < genehacker2> if there is even a slight breeze
19:52 < genehacker2> it's heading straight to the ground
19:52 < QuantumG> that's ok, failing is learning something to.. but not trying worse
19:52 < QuantumG> s/to/too/
19:53 < genehacker2> I'd also advise rigging up a self destruct cord or tying them nice and tight so they don't end up getting away and shooting into the upper atmosphere
19:54 < genehacker2> like one of solar balloon's did
19:54 < QuantumG> yeah, I'll be tying the plane to the ground a bit tighter than I tie the balloons to the plane.. but I'm not sure I really care if I lose either :)
19:55 < genehacker2> heh
19:55 < genehacker2> might want to put a parachute on that netbook
19:55 < QuantumG> yeah.. if I ever get up to that stage.
19:56 < genehacker2> it's not that hard really
19:56 < genehacker2> http://waterocket.explorer.free.fr/parachute.htm
19:57 < genehacker2> what's this for anyway?
19:57 < QuantumG> amusement
19:57 < genehacker2> photography?
19:57 < QuantumG> yeah, it'd be nice to be able to film something
19:58 < genehacker2> so there are formula's out there for figuring out the atmospheric drag on stuff based on the cross sectional are
19:58 < genehacker2> a
19:58 < QuantumG> yeah, I've done it before
19:58 < QuantumG> the equations that is
19:58 < QuantumG> for rocket stuff
19:59 < genehacker2> you can also calculate the thrust your biprop produces by using formulas (propellor efficiency doesn't exceed certain bounds) and the rating on your motors
19:59 < genehacker2> and actual test data if you want
20:00 -!- xp_prg [n=xp_prg3@c-67-188-6-132.hsd1.ca.comcast.net] has joined #hplusroadmap
20:02 < genehacker2> also making gores to make a blimp shaped balloon isn't that hard
20:03 < QuantumG> what do you know about buoyancy control?
20:03 < genehacker2> I built my own system
20:03 < genehacker2> for my solar powered balloon
20:04 < genehacker2> so you can drop weights and let helium out
20:04 < QuantumG> as in solar panels or as in solar balloon?
20:04 < genehacker2> solar thermal balloon
20:04 < QuantumG> ahh cool
20:05 < genehacker2> you could use ballonets
20:05 < genehacker2> if your balloon can hold pressure that is
20:05 < genehacker2> you can heat and cool the air in your balloon
20:06 < genehacker2> of course when you want to change bouyancy this involves change the volume of your balloon
20:06 < QuantumG> I never understood ballonets.. do ballonets need to be inside the balloon?
20:07 < genehacker2> yes
20:07 < QuantumG> right, that makes more sense
20:07 < genehacker2> ballonets decrease the amount of air it displaces
20:07 < QuantumG> the changing volume bit
20:07 < genehacker2> is quite pesky
20:08 < QuantumG> pumping air is, umm, currently beyond me :)
20:08 < genehacker2> because you can't make your whole blimp just a helium balloon
20:08 < CIA-32> skdb: kanzure * r 8e61ed1 /core/units.py: make skdb load faster by not excessively generating combined.dat in core/units.py
20:08 < genehacker2> then use nichrome heaters and water
20:09 < genehacker2> nichrome burns a hole in a balloon letting the helium out
20:09 < genehacker2> a water pump pumps water out when you want to go up
20:12 < CIA-32> skdb: kanzure * r 0890feb /readme: updated the readme
20:13 < QuantumG> so yeah, just coming up with some computer controlled buoyancy control would be an interesting problem.
20:15 < genehacker2> http://www.cockeyed.com/inside/acre/acre.html
20:15 < genehacker2> http://home.earthlink.net/~archvillain/images/lj/skycombat-article3small.jpg
20:16 < kanzure> hmm.. http://github.com/kanzure/skdb/graphs/punch_card
20:18 < QuantumG> heh
20:22 < genehacker2> analog punch card?
20:23 < kanzure> showing when we've written code for skdb (well, committed it)
20:31 < kanzure> assembly_volume_estimate([lego1, lego2]) works, but assembly_volume_actual gives RuntimeError 'StdFail_NotDone' due to BRepBuilderAPI not liking the BRepAlgoAPI_Fuse operation or something
20:31 < kanzure> I guess I can do something other than fuse to figure out the volume, like taking the sum of the volume (assembly_volume_estimate) and subtracting the volume of the boolean common
20:41 < genehacker2> can you calculate area too?
20:42 < genehacker2> there's a nice litle formular that gives you mold solidification time when you plug in area and other factors
20:42 < kanzure> yes
21:12 -!- wrldpc2 [n=benny@ool-ad03fe34.dyn.optonline.net] has quit [Read error: 104 (Connection reset by peer)]
21:12 -!- wrldpc2 [n=benny@ool-ad03fe34.dyn.optonline.net] has joined #hplusroadmap
21:40 -!- dira [n=chatzill@86.99.40.183] has quit ["ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]"]
21:53 < QuantumG> rant:
21:53 < QuantumG> People who think human level AI is only 10 years away have actually never tried writing a human level AI :)
21:53 < QuantumG> If we knew exactly how to do it now, we'd be working on the engineering problems for 10 years.
21:53 < QuantumG> However, I will say that in 10 years time it will be easier to start a project like human level AI. You'll still fail, but it'll be cheaper to fail. So maybe in 10 years we'll start learning how to do it.
21:59 < kanzure> what the fuck is ai?
22:00 < QuantumG> don't get into a semantic argument :)
22:00 < kanzure> sorry, I just think you're bullshit
22:01 < kanzure> however I think this of everyone who talks about ai
22:01 < kanzure> because they all have the same answer
22:02 < QuantumG> well, when I say a human level AI I have a very practical answer... a software program that is capable of doing a wide range of tasks given to it in natural language to an acceptable degree.  And the reason why I say "human level" is cause currently these tasks can only be done by humans.  
22:03 < QuantumG> I'm not interested in Turing tests
22:03 < QuantumG> or any of the other nonsense
22:04 < kanzure> meh
22:05 < QuantumG> I have a practical expectation of what a human (say, employee) should give me in response to my instructions, and that's what I want from human level ai.  If the human can't deliver, I fire em.. if the program can't deliver, I ask for a refund :)
22:05 < kanzure> sorry I must have lost the point
22:05 < kanzure> are you trying to build a brain or not?
22:06 < katsmeow-afk> are these expectations including physical acts, or merely understanding the directions?
22:06 < QuantumG> kanzure: not right now
22:07 < QuantumG> katsmeow-afk: well, my employees are software developers :)
22:07 < katsmeow-afk> ah
22:07 < QuantumG> I rarely tell them to do physical things..
22:07 < kanzure> you have employees?
22:07 < QuantumG> ya
22:07 < kanzure> hi, my name is kanzure
22:07 < kanzure> who are you?
22:08 < QuantumG> your name is brian :)
22:08 < kanzure> bryan
22:08 < kanzure> but really, who are you?
22:08 < kanzure> and why do you have employees?
22:08 < QuantumG> I'm Trent
22:08 < kanzure> yes, mr. waddington :p
22:08 < kanzure> but that doesn't answer my question
22:09 < QuantumG> well, I'm not sure why the two are related.. I don't really define myself by my profession.. but to answer your question, I own a part of a software company.
22:09 < kanzure> just have never heard of it
22:09 < kanzure> have known you for a few years by now
22:09 < QuantumG> yeah
22:09 < QuantumG> I don't really talk about who I work for/with .. after a few bad experiences
22:10 < QuantumG> until I'm no longer working for/with them.
22:10 < kanzure> sounds shady
22:10 < QuantumG> that's the internet for you
22:11 < QuantumG> "you insulted me on Slashdot, I'm going to hack your company's website."
22:11 < kanzure> maybe ten years ago
22:11 < kanzure> now it's "I'm going to cloud your ajax xss attack vectors, n00b"
22:11 < QuantumG> hehe
22:11 < QuantumG> I actually had someone call one of my employers
22:12 < QuantumG> it was embarrassing.
22:12 < kanzure> for the thousands of emails that I send out on the internet with my phone number in them,
22:12 < kanzure> nobody calls me
22:12 < kanzure> this is probably a good thing
22:12 < kanzure> but it's kind of surprising
22:12 < kanzure> people seem to be afraid of the telephone
22:13 < QuantumG> oh, and even going to the trouble of not talking about my employment I still get people posting my address, telephone numbers, pictures of my house from streetview, etc..  fucktards.
22:14 < QuantumG> I've had people harass my mate in Canada who provides the account I'm ircing from
22:14 < QuantumG> and used to provide my website before I moved it to dedicated hosting
22:14 < QuantumG> maybe this all says more about me than them
22:14 < kanzure> you might just be a collosal jerk :p
22:15 < QuantumG> no doubt
22:22 < ybit2> http://www.electraflyer.com/electraflyerc.php *drool*
22:22 < ybit2> http://www.youtube.com/watch?v=fWoLsJz8J5U 
22:23 < ybit2> i wonder what type of glider the electraflyer-x is built from
22:23 < ybit2> http://www.electraflyer.com/
22:23 < katsmeow-afk> ybit, speak up before he busts the crts up
22:24 < QuantumG>  	Empty weight w/battery packs
22:24 < QuantumG> Maximum Take-off Weight (MTOW)
22:24 < QuantumG> 	380 lbs (172 kg)
22:24 < QuantumG> 625 lbs (283 kg)
22:24 < QuantumG> wow
22:24 < kanzure> wish I could remember what I was doing
22:25 < ybit2> :)
22:25 < QuantumG> before or after ranting with me about ai :)
22:25 < kanzure> during
22:25  * ybit2 will take either: http://en.wikipedia.org/wiki/Schempp-Hirth_Duo_Discus or http://en.wikipedia.org/wiki/DG_Flugzeugbau_DG-1000
22:25  * katsmeow-afk tells him that ybit doesn't want the electron guns
22:26 < ybit2> awww
22:27 < ybit2> i'll take them, if anything, they make for fun sculptures 
22:28 < QuantumG> http://www.reuters.com/article/entertainmentNews/idUSTRE5850F320090906
22:28 < QuantumG> sure hope its better than Slacker Uprising
22:31 < katsmeow-afk> a glider makes fun scuptures?
22:32 < ybit2> an electron gun
22:33 < katsmeow-afk> oh, i don't think he wold save the crt necks after busting them, tomake sculptures
22:33 < ybit2> i made mine look like a robot and it was a great crt monitor decoration until i decided to look inside all of its components
22:34 < ybit2> anyone have glider knowledge?
22:34  * ybit2 searches for a flying enthusiasts channel
22:35 < ybit2> [03:34] [473] #aviation You need to be invited to that channel
22:35 < ybit2> odd
22:35 < kanzure> there's a uav channel somewhere
22:35 < katsmeow-afk> FCC is rather strict
22:35 < katsmeow-afk> err FAA
22:35 < kanzure> there's an open source uav project channel on freenode
22:35 < kanzure> there's also #highaltitude or something
22:35 < ybit2> i just want to know where the propeller is, or is it tied to an airplane..
22:36 < ybit2> the two gliders that were linked
22:36 < katsmeow-afk> ??
22:36 < ybit2> what are you ?? about? 
22:36 < katsmeow-afk> linked gliders?
22:37 < ybit2> 22:25  * ybit2 will take either: http://en.wikipedia.org/wiki/Schempp-Hirth_Duo_Discus or http://en.wikipedia.org/wiki/DG_Flugzeugbau_DG-1000
22:37 < katsmeow-afk> those are gliders, they have no props
22:37 < ybit2> figured i might as well get a two seater when i get into this business
22:37 < ybit2> http://en.wikipedia.org/wiki/File:DG_Flugzeugbau_DG-808B.jpg
22:38 < ybit2> motor gliders like these are what i'm looking for
22:38 < ybit2> http://en.wikipedia.org/wiki/Motor_glider
22:38 < QuantumG> isn't a motor glider a plane? :)
22:39 < katsmeow-afk> oh, that folds back into the fueslage
22:39 < QuantumG> ahh, I see
22:39 < QuantumG> A motor glider is a fixed-wing aircraft that can be flown with or without engine power. 
22:39 < katsmeow-afk> you can see the hatch open behind the wings, on top
22:41 < ybit2> something like the HK36 Super Dimona is probably what i'm looking for
22:41 < ybit2> i think i linked to a youtube vid of an electric glider this past week or so
22:42 < ybit2> http://www.youtube.com/watch?v=WcWSI03NKo0
22:42 < ybit2> that's it
22:46 < ybit2> problem with this is that you have to pay fees to the airport for the aircraft to park there
22:58 < katsmeow-afk> foldup
22:58 < katsmeow-afk> or, pull the wings off shorter spars
22:58 < QuantumG> a plane you can put on the back of your car, like a boat, is a *long* time coming.
22:59 < QuantumG> quite possibly that drivable plane concept is going to be available first.
22:59 < katsmeow-afk> of course, but there have beena few that trailered
22:59 < ybit2> for only $150k
22:59 < ybit2> or more, just guessing
22:59 < QuantumG> people pay that for a boat.
23:00 < katsmeow-afk> pull off the rear draggy wheel, insert trailer hitch, fold or detach wings, drive off
23:00 < QuantumG> I guess the difference is that you can't go get drunk on your plane.
23:00 < ybit2> katsmeow-afk: that's an idea
23:00 < QuantumG> catch birds.
23:00 < ybit2> http://www.solar-flight.com/projects.html is nice too btw
23:11 < ybit2> http://pastebin.com/m289978ed 
23:14  * kanzure compiles sgeom
23:15 < ybit2> http://groups.google.com/group/rec.aviation.homebuilt/topics?pli=1
23:18 < ybit2> interesting fact: Wind machines were used in Persia as early as 200 B.C.[2]
23:19 < katsmeow-afk> unless you cold a sailboat as a wind macine
23:19 < katsmeow-afk> count
23:20 < katsmeow-afk> i have this spot on my index phalange where i grabbed some red hot steel with a ole in the glove
23:20 < katsmeow-afk> so excuse my typoing
23:20 < ybit2> it was a windmill
23:20 < ybit2> The first windmills were developed to automate the tasks of grain-grinding and water-pumping and the earliest-known design is the vertical axis system developed in Persia about 500-900 A.D. 
23:21 < ybit2> where did this 200b.c. come from..
23:28 < QuantumG> some other wind machine perhaps?
23:50 < kanzure> is "watching code compile" a legitimate hobby?
23:51 < QuantumG> I dunno about hobby, but its certainly a living
23:51 < kanzure> you mean to say that people should be paying me for this?
23:52 < QuantumG> oh, and [insert dis about Gentoo here]
23:52 < kanzure> it's not gentoo
23:53 < genehacker2> yes
23:53 < genehacker2> watch that code compile!
23:56 < kanzure> oh no I missed a parse error on line 38124