2009-12-02.log

--- Day changed Wed Dec 02 2009
fenncan i embed videos in openoffice powerpoint?00:03
fennor animated gifs?00:04
ybiti'm going to guess yess00:15
ybit#openoffice ?00:15
ybitgots to love how the e key is stuck the first night, grr00:16
kanzurefenn: videos, yes. gifs i wouldn't expect to work00:18
-!- jasonwohlfahrt [n=chatzill@cpe-72-179-52-26.austin.res.rr.com] has joined #hplusroadmap00:29
fennanyone on windows?00:34
fenncould please verify if this works in $DEFAULT_WINDOWS_MOVIE_PLAYER? http://fennetic.net/irc/squirrel_tracker.flv00:35
QuantumGya need third party software to play flvs on windoze.00:37
genehacker2no00:37
fennpoop00:37
genehacker2what is it?00:37
fennwell what codec does windows like?00:37
QuantumGplays for me in third party software00:37
fenni dont trust anything to be installed00:38
QuantumGright, so make a wmv00:38
fenni mean i dont expect it to be there when i need it00:38
QuantumGor an avi00:38
fennso its just the format not the codec?00:38
QuantumGor an mpg00:38
genehacker2http://xkcd.com/00:39
QuantumGwell the codec you used for flv is "flash video 1" .. so clearly that's not gunna be around :)    00:39
QuantumGan mpg will play just about anywhere.00:39
QuantumGbut if you really want peons to look at it, put it on youtube ;)00:40
fenni want it in the powerpoint file00:40
fennnoboy said this was optimal00:40
QuantumGdon't care about size?00:41
genehacker2I think there's away to put .flv's in powerpoint00:42
genehacker2no there isn't00:43
genehacker2convert it to .avi00:43
fennok try again? http://fennetic.net/irc/squirrel_tracker.avi00:43
* fenn notes that the conversion was done at 00:38 and the rest of the time was waiting for the upload to finish00:44
genehacker2no00:45
genehacker2but supposedly you can put avi into powerpoint00:45
fenn"no" what?00:45
genehacker2it doesn't work00:46
fennrawr00:46
genehacker2there's away to put youtube videos into powerpoint00:48
genehacker2without any add ins00:48
genehacker2http://jennylu.wordpress.com/2009/01/13/youtube-into-powerpoint-wikihow-has-advice/00:49
fennif only ssh were correct: 52% 2208KB   1.2MB/00:52
genehacker2you might be able to put a webpage on your usb drive that emulates youtube00:52
fennum no thanks00:52
genehacker2and have powerpoint open that00:52
genehacker2just convert it to .swf00:53
genehacker2or figure out a format that will play in excel00:55
genehacker2*powerpoint00:55
fennegad.. ronja has done a 170km free space optical link00:58
fennall these ham call signs sound/look the same00:59
fennwhat is the fucking point01:00
fennronja is showing movies of raytraced parts spinning http://ronja.twibright.com/3d/01:12
fenni love how they automatically include the calculated mass of the brick chimney01:15
fenngrr so jealous http://ronja.twibright.com/3d/hardware_1.png01:17
fennunits 'exp(1)'01:31
fennDefinition: 2.718281801:31
-!- jasonwohlfahrt [n=chatzill@cpe-72-179-52-26.austin.res.rr.com] has quit [Read error: 60 (Operation timed out)]01:32
-!- jasonwohlfahrt [n=chatzill@cpe-72-179-52-26.austin.res.rr.com] has joined #hplusroadmap01:46
kanzurehey jason01:46
kanzurestill on for thursday?01:47
-!- xp_prg [n=xp_prg3@c-98-234-218-161.hsd1.ca.comcast.net] has quit ["This computer has gone to sleep"]01:52
-!- katsmeow is now known as katsmeow-afk01:54
-!- genehacker2 [i=genehack@wireless-128-62-46-67.public.utexas.edu] has quit [Read error: 110 (Connection timed out)]04:28
-!- jasonwohlfahrt [n=chatzill@cpe-72-179-52-26.austin.res.rr.com] has quit [Read error: 110 (Connection timed out)]04:52
-!- wrldpc2 [n=benny@ool-ad03fe34.dyn.optonline.net] has quit []05:17
-!- futuresoon [n=futureso@cpe-68-175-67-66.nyc.res.rr.com] has joined #hplusroadmap07:59
-!- genehacker [i=genehack@wireless-128-62-46-67.public.utexas.edu] has joined #hplusroadmap08:09
-!- genehacker2 [i=genehack@wireless-128-62-141-52.public.utexas.edu] has joined #hplusroadmap09:12
-!- genehacker [i=genehack@wireless-128-62-46-67.public.utexas.edu] has quit [Read error: 110 (Connection timed out)]09:13
-!- jasonwohlfahrt [n=chatzill@cpe-72-179-52-26.austin.res.rr.com] has joined #hplusroadmap09:36
-!- jasonwohlfahrt [n=chatzill@cpe-72-179-52-26.austin.res.rr.com] has quit [Read error: 60 (Operation timed out)]09:51
-!- genehacker2 [i=genehack@wireless-128-62-141-52.public.utexas.edu] has quit [Read error: 60 (Operation timed out)]09:56
-!- genehacker [i=genehack@wireless-128-62-141-52.public.utexas.edu] has joined #hplusroadmap10:06
-!- Xeones [n=Xeones@cpe-68-175-67-66.nyc.res.rr.com] has joined #hplusroadmap10:16
-!- Xeones [n=Xeones@cpe-68-175-67-66.nyc.res.rr.com] has left #hplusroadmap []10:16
-!- wrldpc2 [n=benny@ool-ad03fe34.dyn.optonline.net] has joined #hplusroadmap10:20
-!- genehacker [i=genehack@wireless-128-62-141-52.public.utexas.edu] has quit [Read error: 110 (Connection timed out)]10:43
kanzurehttp://designfiles.org/10:50
kanzurehm it's snowing in dallas11:10
-!- jasonwohlfahrt [n=chatzill@cpe-72-179-52-26.austin.res.rr.com] has joined #hplusroadmap12:18
bkeroSNowing here soon12:30
-!- jasonwohlfahrt [n=chatzill@cpe-72-179-52-26.austin.res.rr.com] has quit [Read error: 110 (Connection timed out)]12:45
-!- genehacker [i=genehack@wireless-128-62-103-189.public.utexas.edu] has joined #hplusroadmap13:07
CIA-33skdb: kanzure * r a59ae6c /doc/presentations/hplus-summit-2009/outline: an old outline for the h+ summit 2009 talk on how to make a civilization seed13:08
kanzurehm i keep forgetting how to set stickybit13:14
CIA-33skdb: kanzure * r d140a12 /thirdparty/graphsynth.py: update graphsynth to work with gxml files without canvas objects13:23
CIA-33skdb: kanzure * r d8e940d /thirdparty/graphsynth.py: fix graphsynth output for arcs13:39
-!- genehacker [i=genehack@wireless-128-62-103-189.public.utexas.edu] has quit [Read error: 60 (Operation timed out)]14:07
-!- genehacker [i=genehack@wireless-128-62-103-189.public.utexas.edu] has joined #hplusroadmap14:19
fenndo i know Chris Pearson? http://gingery.wordpress.com/14:29
fennhm or is it jessica ginger/charlesworth14:35
CIA-33skdb: kanzure * r 83b1e90 /geom/geom.py: update geom.py to not complain when you load !part14:36
fennhmm. "Just realised that being 30 is about just saying "yup I'm going back to California! no matter what!l How should/can I make it happen?"14:45
fennmaybe i have some sort of old-person virus14:46
fennthat makes me want to go to california14:46
bkeroI would not like to go back to california. :(14:49
fennyou aren't 30 yet14:50
fennkevin kelly is into lifelogging? http://www.meetup.com/quantifiedself/14:52
kanzuresteve keeps yapping about the "quantified self" people in california14:54
fenn"the Intellectual Implosion of Kevin Kelly"14:55
kanzurekk.org is why i know you14:56
-!- jasonwohlfahrt [n=chatzill@216-54-230-114.static.twtelecom.net] has joined #hplusroadmap14:57
kanzurehey jason14:57
kanzurehuh ddclient works for zoneedit15:02
fennwhat's that some windows band-aid?15:05
fennoh it's dyndns.. what was the firewall thing? zonealarm?15:06
fennrofl http://kk.org/ct2/Socialven.jpg15:15
kanzurehi futuresoon 15:38
futuresoonhey kanzure15:38
futuresooni'm a drupal developer. that's what i do15:38
kanzureright, so, replab web stuff15:38
futuresoonright15:38
kanzurei'm not sure drupal is a good idea considering the code base we already have15:39
futuresoonit may be a stretch but, what about rdf?15:39
kanzurewhy not use what we have15:39
futuresoonwell i still need to look at what you have15:39
kanzure:)15:39
futuresoonand i do agree it shouldn't be another instructables15:39
futuresoonit should actually progress the automation concept15:40
kanzuredo you know about apt-get?15:40
futuresooni use ubuntu15:40
kanzureskdb is apt-get 'cept for hardware15:40
futuresoonyeah i love the concept. i've been reading15:40
kanzureso, the server is a front-end to skdb15:40
kanzure(it's all crammed into web.py)15:40
kanzureit literally loads up opencascade and skdb in the background15:40
kanzureas python modules15:40
kanzurei stayed away from mod_php and mod_python for obvious reasons i hope15:41
kanzurereloading opencascade on each page request.. bad voodoo15:41
futuresooni guess that rules out django?15:41
kanzurenope, django was a candidate IMHO15:41
kanzurebut i tried it15:41
kanzurei was writing "djangit", a git wiki for django15:41
kanzurehttp://github.com/kanzure/djangit15:42
kanzurebut i started to realize i hated django15:42
kanzureso now web.py is the main candidate i think15:42
kanzureer15:42
kanzurenot web.py the library, but the code in skdb.git15:42
kanzureit runs off of cherrypy. basically you just drop in the skdb classes and it makes a web server out of them15:42
kanzurehttp://adl.serveftp.org/skdb/web/web.py15:42
kanzurehttp://cherrypy.org/15:42
futuresooni'm still learning python. drupal is my main jam. not sure how much help i'll be able to be15:43
kanzureit's pretty far along15:43
kanzurethere's no paypal integration yet15:43
futuresoonyeah that's the type of thing drupal makes simple15:43
kanzureyes i'm sure it's simple15:43
futuresooni see that the longer term roadmap for these types of things is in rdf?15:43
kanzurebut i'm not so sure about using multiple platforms15:44
kanzurewell not quite rdf. there's something we've been using called "yaml"15:44
futuresooni think i read that at the oskamak wiki?15:44
kanzurehttp://adl.serveftp.org/skdb-packages/screw/metadata.yaml15:44
kanzureoscomak is paul fernhout's attempt at convincing me to use rdf and his triples language :)15:44
kanzurebut it's yet-another-data-storage-format15:44
kanzureyaml is for serializing objects into and out of memory, in a human-readable format15:45
futuresooni see15:45
kanzurethis is why i'm cautious about multiple platforms15:45
kanzurewe should all speak the same languages15:45
kanzuredrupal is great, and i was originally a php programmer guy15:45
kanzure(i was really turned off by phpnuke though. i guess i'm old)15:46
kanzurefenn: are you around?15:46
kanzurei need to hop a bus, and fenn is able to answer various dev questions in the mean time15:47
fennya15:47
futuresooncool15:47
fenni think drupal django and friends are too complicated and heavy15:47
futuresoonfenn: what about if drupal or whatever were just outputting yaml style metadata15:48
fennthat doesn't make any sense15:48
fenndrupal is a web framework15:48
futuresoonyou're familiar with rdfa?15:48
fenni know what rdf is, and considered using it (maybe i will)15:49
fenni15:49
fenni'm no rdf wizard though15:49
futuresoonwell right there in the html markup of a drupal page could be rdf that a user generated by filling out forms15:49
futuresoondrupal 7 will output rdf automatically and drupal core will be better suited when drupal 7 comes out in general, but i've made queryable rdf triple stores in drupal 615:50
fennwhat is the "a" in rdfa?15:50
futuresooni forget15:50
fennfilling out forms is not practical15:51
futuresoon"in attributes"15:51
fennthe data is way too complex and non-standard to be worth making forms for people to fill out15:51
fennthis is why i didnt go with a relational database15:51
fennthey are focused on tables and data that is essentially similar15:51
futuresoonwell i think the thinking was that it's easy to create and map forms to rdf when your'e doing it in drupal15:52
fennpeople keep telling me to make a wizard that dummies can follow to create anything, but i just dont see it happening15:52
futuresoonso arbitrary data structures are less of a problem15:52
futuresoonthat wizard is cck http://drupal.org/project/cck15:52
futuresoonor http://drupal.org/project/rdfcck15:53
-!- jasonwohlfahrt [n=chatzill@216-54-230-114.static.twtelecom.net] has quit [Read error: 110 (Connection timed out)]15:53
-!- Netsplit zelazny.freenode.net <-> irc.freenode.net quits: nsh, bct15:54
-!- Netsplit over, joins: bct15:54
-!- nsh [i=sbp@59.176.232.72.static.reverse.ltdomains.com] has joined #hplusroadmap15:54
fennugh why can't they just link to the youtube url for these tutorial videos15:55
futuresooni built this a couple days ago before i knew there were people on replab.org working already http://replab.sigmanode.com15:56
futuresoonit uses cck15:56
futuresoonto make autocomplete forms15:56
fennthat is what we pejoratively refer to as 'wiki bullshit' :) no offense15:57
futuresoonwell if you map it to RDF is what i'm saying15:57
futuresoonanyway some of it depends the timeframe. a lot of this is going to come online better than it is now in a couple months whether anyone here works on it or not15:57
futuresoonafk16:05
fennwell i dont get it16:06
-!- drazak_ [n=drazak@drazak.net] has joined #hplusroadmap16:08
Utopiahfuturesoon: are you at http://openfarmtech.org/ ?16:08
-!- drazak [n=drazak@drazak.net] has quit [Read error: 104 (Connection reset by peer)]16:09
fenni think the point everyone's been  missing all along is that there's plenty of content out there, but it's all in terrible formats like plain text and images buried in some cms database16:11
fennweb forms like this just encourage lazy shoddy contributions i think16:12
fenn"look it displays without errors, i must have done it correctly"16:13
futuresoonfenn: i don't think everyone should be able to haphazardly enter it (unless the community of people who can do it grows fast enough that that's worthwhile)16:15
futuresoonbut it does seem to make sense that anybody should be able to use it, whether they're just reading instructions visually or not. more eyes makes for shallower bugs, that type of thing16:15
futuresoonUtopiah: i've been lurking/reading their weblog for a while, haven't posted16:16
Utopiahfuturesoon: ok, but I mean you are not on-site16:16
futuresoonno16:16
fennfuturesoon: what's wrong with just letting users edit the rdf directly? (besides that its a visual clusterfuck)16:17
futuresoonthe major problem with rdf unlike in yaml is that presumably a lot of the benefit in rdf eventually comes from using specific classes that not just anybody is going to know to input from examples. i mean i don't know how interoperable various yaml datasets ever get16:18
fennhow do you discover what the correct rdf class to use is?16:18
futuresoonthis might be the type of thing that i should try and figure out what you guys have already done and whether there's even something i can envision in drupal/rdf that has any benefits---these are just general benefits i'm talking about in a general way16:19
futuresoonfenn: when someone implements an rdf ontology that you like, you just map all your existing data to it, or other people can take your data and do that16:19
fennkanzure always talks about not "putting a gun to everyone's head" to make them use your ontology.. 16:20
futuresooni think that happened when google released their own ontology recently for people, products, etc.---the rdf community was like "that already exists" and just made their own mapping16:20
futuresoonfenn: absolutely16:20
futuresoonrdf is a multiplicity of worldviews on what's the best ontology16:20
futuresoonanyway, i wouldn't try to do a drupal rdf implementation other than as proof of concept before drupal 7 comes out in a few months anyway16:21
futuresooni originally came to kanzure like "i'm a drupal dev. what can i do"16:21
futuresoonnow i'm seeing rdf (from something i read on oscomak) looks promising here16:21
fennheh16:21
futuresoonbut the real first thing for me is to start seeing what is up in this project that you already have going16:21
fenni keep telling paul to just use rdf, but he's stubborn because he invented it first, or something16:21
futuresoonheh16:21
futuresoonis paul a yaml guy?16:21
fennno16:22
futuresoonpaul is kanzure?16:22
QuantumGfenn: got a link to that google ontology?16:22
fennhe came up with something called 'pointrel' in the 1980's16:22
fennpaul fernhout is a guy who writes long winded emails on the openmanufacturing list16:22
fennhe wrote oscomak16:22
futuresoonahh i see16:23
fennor talked about writing it at least16:23
futuresooni came to all this stuff first through arduino and then openfarmtech. i still don't have a lot of manufacturing experience apart from soldering open source chargers for ipods16:23
fennbut he got distracted with pointrel16:23
fennyeah i had a revelation when i built ladyada's usbtinyisp16:24
fenn"you mean i can download hardware?"16:24
futuresoonapt-get :-)16:24
fennapt-get was later, when i was thinking about openfarmtech16:25
fennit seemed marcin was making something like GNU16:25
fennbut there could be many different "technology distributions"16:25
fennand of course the thing that made linux not impossible to use was a package manager16:26
fennotherwise you were constantly fiddling around with little details16:26
fennkits should be pre-vetted to make sure they "just work"16:27
fennso i came up with a way to describe what works with what16:27
futuresoondo you think you could describe what a mock yaml entry for some random thing might be?16:28
fennit's not as simple as with software though, because you can't just instantly instantiate any package at zero cost16:28
futuresoonnaturally16:28
fennsure16:28
futuresooni like marcin's phrase of "scrap metal plus labor"16:28
fennmeh16:28
futuresoonmarketing BS? maybe :-P16:28
fenneverything is just matter energy and information, but saying that doesn't tell you how to do it16:28
futuresoonsure16:28
fennit is something you dont really understand until you make stuff from scratch though16:29
futuresoonbill of materials sounds like one thing where you'd want standard data16:29
futuresoonis there a concept of "bill of materials" in this?16:30
fennthat's why it's so easy for consumers to get attached to a brand or style.. it's something magical, not just a series of movements someone made at some point in time16:30
fennyes, sort of16:30
fennbill of materials as traditionally stated is more like a shopping list16:30
fennwe want to reference the actual packages required to include16:31
futuresoondependency management, sure16:31
QuantumGso does skdb include a list of required skills and skill levels?16:31
fennlike an electrolytic capacitor; i dont want to just say 'buy this part number' i want to describe what it really is enough that you could make it from scratch if necessary16:31
futuresoonfenn: okay but if it's an integrated circuit, you just include a part number?16:32
fennQuantumG: not yet; never came to any good conclusions on how to do that unambiguously16:32
fennfuturesoon: yes but remember not all IC's are black boxes16:32
fenndo you know about opencores?16:32
fenni can instantiate an 8051 processor on any fpga chip16:32
QuantumGya.. it's an interesting problem.. but certainly being able to say "intermediate welding skills required" would be handy.16:32
futuresoonfenn: that's interesting and i'd like to hear more about that at some point, but it sounds like a judgment call whether to buy it or make it16:33
futuresooni mean what's faster? what's reasonable16:33
fennQuantumG: there was talk with the fab academy about providing skill training certificates for people who completed their course16:33
fennfuturesoon: depends... :)16:33
futuresoonfenn: especially if you could include a link to an open source manufacturer, you could support that enterprise, create advertising for peoples' open source business models, save people time16:34
fennthe idea is to define a preferences vector to weight different costs16:34
fennyep16:34
futuresooni believe in multi-dimensional analysis :-)16:34
fennthe point is there's no one true answer16:35
fennit depends on a person's goals/values and circumstances16:35
fennshipping is really expensive in sweden for example16:35
futuresoonso you want to advertise a critical path not even for most people, but just the plurality of people16:35
futuresoonor put your energy there etc16:35
QuantumGand yeah, you should be able to say:  skill(stick welding, intermediate) || skill(mig welding, intermediate) || skill(tig welding, beginner)   for something that's easier to do with tig than with stick or mig.16:36
fenner. it would calculate the critical path on an individual basis, depending on your preferences and circumstances16:36
fennhopefully we can fudge some positive network effects into the calculation16:36
futuresoonQuantumG: one potential with RDF is that someone could make a new instance of the web stack and map their own answers to say just that16:36
fennconversely i'd like to propose inferring skill sets from projects one has previously completed16:38
fennit's not a one to one mapping though16:38
fenni mean you could have tig welded it but that doesn't mean you're an intermediate stick welder16:38
QuantumGyeah, for example you might have taken on the project but gotten someone else to do that skill16:39
fennso, how do i make pretty pictures of these ideas? this is what i'm wrestling with today16:39
fennabout the closest thing i have is stuff like http://designfiles.org/lab/techtrees/Terran_Tech_Tree.jpg16:39
QuantumGjust having the required skills in your BOM would be an improvement over most kits that use prose to describe the required skills.16:40
fennhttp://designfiles.org/lab/fenn/pngs/zerg_tree.png16:41
fenni'd love to create those on the fly with graphviz and real technology16:42
fennhave to build the db first though, which apparently is harder than it looks16:43
fennchicken and egg problem16:43
futuresoondesign choices that you won't be able to undo?16:43
fenngetting anyone to contribute without already having impressive graphics16:44
fenngetting motivated about a project that you've been talking about for years and has gone nowhere16:44
futuresoonso what stops this from being something that can be part of a major drupal install?16:44
futuresoonthe data structures are already arbitrary YAML type stuff right?16:45
fenni dont want to have to deal with heavyweight software, that's all16:45
fennopencascade is bad enough, i'm actually thinking about dropping it despite the months invested in figuring it out16:45
futuresoonsecond time you've mentioned that. i'll look up what that is16:45
fennhttp://adl.serveftp.org/dokuwiki/occ16:46
fennlarge part of the problem is there is no open source cad format or program16:46
fennopenscad is interesting16:47
fennin the end we will probably use a big mishmash of stuff16:47
fennalibre (not Free), brl-cad, blender, openscad, AoI, heekscad16:48
fenni dont really get why google hasn't open sourced sketchup16:48
fennnow i am rambling16:48
futuresoon:-)16:49
futuresoonyeah i just need to see what's already in skdb. for bills of materials, and dependency management between projects, rdf seems to make sense. obviously the CAD stuff if that's your data it doesn't benefit from having people fill out forms16:50
futuresoonwhat i had envisioned was instructables with dependency management and links (which could be monitored as to whether they were dead) to places to buy the bills of materials if you couldn't make them16:51
futuresoonyou could state what materials you already have and filter everything by prerequisites you meet16:52
fennremember i said it wasn't like software - you dont want to require a specific version of a thing (a certain model of milling machine for example)16:53
futuresoonright, well each node could have revision tracking16:53
fenndefining non-specific requirements turns out to be difficult16:53
fennno you miss the point16:53
fennit could be a router or a VMC or a bridgeport; they all perform adequately16:53
futuresoonright so you're saying if there's 5 versions of something, how do you know of the various combinations and permutations which one's good anymore16:53
fenni dont want to have to buy/build a router if i already have a VMC to do the job16:53
QuantumGso will your db include the kinds of tests a part needs to pass before it is integrated?16:54
fennso this is why all the process/technique functionality stuff16:54
fennQuantumG: more like guides, like the debian lint or whatever it's called16:55
fennlintian16:55
fennfuturesoon: there's not a lot of data in the system yet, so the format is still flexible.. this is why i find it hard to tell people what to do16:56
futuresoonyeah16:56
futuresoonfenn: okay so i'm having an RDF thought16:56
fennbut then they send me stuff and it's all wrong :(16:56
futuresoonwhat if you had a data structure that says "isDependentOn" and also "replacesDependencyOn"16:57
fennso i end up rewriting all the data and spending about as much time as doing it myself from scratch16:57
futuresoonanyway i could see how this stuff gets complicated16:57
-!- jasonwohlfahrt [n=chatzill@cpe-72-179-52-26.austin.res.rr.com] has joined #hplusroadmap16:58
futuresoonfenn: are you familiar with SPARQL?16:59
futuresoonyou said you know about RDF a little?16:59
futuresooncause all these contingencies sound like the real apt-get is the SPARQL query and you're just modifying web forms so that the SPARQL query you get will handle all these eventualities17:00
futuresoonis that something that's possible?17:02
fennSPARQL is just SQL for RDF right?17:04
futuresoonbasically17:05
fennthen no17:05
futuresoonso if a VNC claimed to replace a dependency of a router in a module, it could say so17:05
fennyes17:05
futuresoonand presumably you'd just include anything out of the tools you already have that replaces any of the module's dependencies17:05
futuresoonwhere a module would be the SPARQL query without replacing17:06
fennright17:06
futuresooniow, the actual module is to look at what you have and replace its dependencies with things those things where it can17:06
fennbut, not all the data is explicit, there will have to be some intermediate calculations to figure out what is possible, which is where things get "interesting"17:06
futuresoonany examples of that?17:06
fennas engineers we love to make everything modular and black boxy17:06
futuresooni mean obviously that makes sense but something concrete17:06
kanzureback. reading.17:07
futuresoonhey kanzure17:07
kanzurethanks fenn for picking up where i slacked off17:07
futuresoonkanzure: the short version is that i'm trying to envision whether RDF isn't the perfect thing for this17:07
futuresoonand what your thoughts might be on that17:07
fennrdf has the nice quality of being infinitely extended17:07
fennunfotunately it's a pain in the ass to edit17:07
kanzurehttp://replab.sigmanode.com makes no sense to me17:08
futuresoonkanzure: i wasn't on the google group yet17:08
futuresoonso i was just "starting anywhere" as they say17:08
futuresooni had no idea real package handling was already underway17:08
fennthe google group hasn't come to any conclusions yet17:08
fennthe replab group i mean17:09
kanzure16:21 < futuresoon> i originally came to kanzure like "i'm a drupal dev. what can i do" <-- oh. i read it more like "i'm a drupal dev. everyone convert to drupal now! because i know how to use it"17:09
futuresooni can see that too, but what i'm saying is, i was just trying to get a replab.org community begun so they could start selling their CEB presses and getting some instructables "wiki bullshiTt" so to speak17:09
kanzurefuturesoon: i am not paul. paul fernhout is http://pauldfernhout.net/ and i'm http://heybryan.org/17:10
futuresooni much prefer real package management if that's technologically feasible17:10
futuresoonkanzure: i see that now17:10
fennwe have packages in git repositories that you can automatically sync/download17:10
fennkanzure: i havent actually tried skdb-get.py yet17:10
fennbut i suppose it works17:10
kanzurefuturesoon: mock yaml entries are available, like http://adl.serveftp.org/skdb-packages/screw/metadata.yaml for instance17:11
fennbtw why is VOICED in there?17:11
kanzurefuturesoon: it would calculate the critical path dynamically. that's what all this dependency management stuff is about.17:12
kanzureQuantumG: yeah there's a dependency resolution example floating around in the unit tests folder (stupid, i know) based off of booleano's syntax17:12
futuresoonkanzure: so "name" "functionality" and "parts" might be some basics?17:14
futuresoonparts being the dependencies which might be other yaml files?17:16
kanzurefuturesoon: part of the problem with instructables is that the instructions suck. so we were working on a way to fix that; think rendering instructions to 3D animations, text, graphics, etc.17:16
fennparts is the stuff that is in the package17:16
kanzurefuturesoon: you can check out the metadata.yaml file to see dependencies 17:16
fenna materials package might have no parts at all17:16
futuresoonoh, yup. i see the word dependencies now17:17
kanzurefuturesoon: oh you were replied to my messages. heh. i'm still not caught up yet17:17
futuresoonkanzure: and rdf was not selected because it's easier to manually edit yaml than rdf?17:17
futuresoonkanzure: heh no worries17:17
* fenn pouts17:18
kanzurefenn: what's wrong?17:18
fennhe's asking you why you decided all the things i decided17:18
fennmaybe i am just overly sensitive to this17:18
futuresoonfenn: well it was your answer. it's a good answer17:18
futuresoonhey, there's no doubt that rdf is harder to manipulate than yaml17:19
futuresooni'm just thinking interoperability in an alleged future of rdf that can't be proven will ever happen17:19
fenni was originally going to do alternative xml syntax (SLiP)17:19
-!- genehacker [i=genehack@wireless-128-62-103-189.public.utexas.edu] has quit [Read error: 110 (Connection timed out)]17:19
futuresoonok let me look that up17:19
kanzurestory time17:19
kanzurethe other week i was involved in a group called 3 day startup17:20
kanzureso i spent the weekend not sleeping and doing lots of coding17:20
fennhttp://slip.sourceforge.net/17:20
kanzureand i was on a team with some other programmers17:20
futuresoonfenn: yup i see it17:20
kanzureand after talking to see what our shared skillset was, and showing them some code i had prototyped quickly,17:20
kanzurethey all started to ask me why we shouldn't just use a mysql database17:20
kanzure(ok, not all at once, but just randomly over the weekend)17:20
kanzurei don't know how to explain this17:21
kanzurebut people seem to fixate on particular tools?17:21
fennit's called mind share17:21
kanzurewhen i was doing php work in 2003, i would write the same lines of php over and over again to query a mysql db17:21
kanzurebut i was young and stupid, so it's forgivable17:21
kanzureand everyone who was a "web designer" just stole templates from other websites17:21
futuresooni never write mysql so i'm not as good at it as a web developer should be17:22
fennonce you start using a tool a lot, you start thinking in terms of that tool17:22
kanzurebecause that's what they knew how to do..17:22
futuresoonsounds like what i'm doing17:22
kanzurei don't think fenn or i have anything specifically against xml, rdf, slip, json, or anything else you throw at us17:22
kanzurebut at some point we needed to find some common ground17:22
kanzureor sparql17:22
kanzuregah brainlapse17:23
fennhey if you want to make a "more websitey" template for this, that would be great: 17:23
fennum. heybryan.org:8081 is down?17:23
kanzureer17:23
kanzureone sec17:23
kanzuretry now?17:23
fennfuturesoon: for this http://heybryan.org:8081/package/17:23
fennbecause apparently it's too plain-looking for kanzure17:23
kanzureyes17:23
fenneverything is gray on gray for me though, regardless what website it is, and that's how i like it17:24
fennfuturesoon: really unlikely that this already exists, but it would be cool if drupal had a git backend17:26
fenni am sick of data jails17:26
kanzurei was pretty upset that i had to write my own module for django to work with git17:26
kanzureand even then i couldn't stop it from wanting to use a database17:27
futuresoonhttp://drupal.org/project/versioncontrol_git <--- not sure if this is anything like what you want?17:27
kanzureand it doesn't have a flatfile database backend (django i mean)17:27
kanzure"At the moment, only a back end for CVS has been written (see references)."17:28
futuresoonkanzure: flatfile would include .rdf dbases right?17:28
kanzureah nevermind17:28
fennthat was the project proposal17:28
kanzurefenn: sorry, you're right17:28
kanzurefuturesoon: flatfile could mean anything :)17:28
fennalmost anything17:29
kanzureheh true that. sqlite doesn't count in my book17:29
fenni dont think a bare git repo would count17:29
kanzurebut it's closer than mysql and postgresql etc17:29
fennflatfile means "show me the code!"17:29
kanzurefenn: no i'm talking about a plugin to make django use ..17:29
kanzureer.17:29
kanzurei have to remember that you don't actually know django17:29
kanzurethe way django works is that you instantiate the database and load it up with the models written in model python files 17:29
fennuh, nevermind i got confused17:30
kanzuredjango has some management code that does this for you, wrapping around the database api17:30
kanzureand then the database takes away your information and puts it somewhere, probably in /var/mysql/ or something17:30
kanzurefuturesoon: please feel free to test/break the http://heybryan.org:8081/package/ demo. you can "edit" files17:31
kanzureit should be one or two lines of python to add in a preview CAD image generated based off of the yaml that a user edits17:31
kanzure(i already wrote a CAD preview generator dealy, but haven't tied it in)17:31
futuresoonkanzure: i should wget the lego files for instance and then do something?17:32
futuresoonedit the yaml?17:32
kanzurewait wait, what?17:32
kanzureclick the "edit" link17:32
fennyou could set up your own copy of the website but that's not what he means17:32
kanzurehttp://heybryan.org:8081/package/lego:master/data.yaml/:edit17:33
futuresoonctrl-f for 'edit' doesn't give me anything on the lego page17:33
kanzureyou have to click a file first17:33
fennactually i dont see any edit link17:33
kanzurelike data.yaml17:33
fenn:x17:33
kanzuresorry17:33
kanzurei suck17:33
kanzurei hear you laughing, fenn :(17:33
futuresoonno prob, okay i changed some of the CNC paths17:33
kanzureyou did? there were paths?17:34
futuresooni made a -1 a -217:34
kanzurewhat?17:34
futuresoonthe first y-vector17:34
kanzurecan you link me?17:34
futuresoonthat i saw17:34
kanzureoh17:34
futuresoonhttp://heybryan.org:8081/package/lego:master/data.yaml/:edit17:34
futuresoonsame link i guess17:34
kanzureis this it?17:34
kanzurehttp://heybryan.org:8081/package/lego:anonymous/data.yaml/5ffe4c3060a4196bc21932d306190327c64db23217:34
kanzureah the first y_vector17:34
kanzureyep i see the change17:34
futuresoonthat includes my edit17:34
kanzureneat17:34
futuresoonjust have to work on that edit button now :-P17:35
futuresoonjust playing17:35
* fenn mumbles something about line by line diffs17:35
futuresoonyeah a diff sure17:35
kanzureblah i don't have that implemented yet17:35
kanzuresome help would be awesome..17:35
fenni cant help but feel all this git filesystem crap is redundant development effort17:35
kanzurewhat?17:35
fenndoesn't this exist somewhere already?17:35
futuresoonhttp://drupal.org/project/versioncontrol_git <--- not sure if this is anything like what you want?17:36
kanzurethere are a lot of git/ruby libraries17:36
futuresoondid that turn out to be relevant or not?17:36
fenni mean with cherrypy17:36
fennmaybe that's asking too much17:36
kanzurefuturesoon: i know how drupal plugins work- in general- but how do drupal modules work, like versioncontrol_git?17:36
futuresooni know it may sound like the drupal "hard sell" but there's a lot of stuff you get out of the box with drupal17:36
futuresoondrupal modules provide hooks to other modules so they can become depenencies of those modules17:37
futuresoonthat's about it17:37
futuresoonthe primary usefulness of drupal (apart from the large development community in general) are cck and views17:37
futuresoonin particular i think rdfcck would be of interest17:37
futuresoonbecause then you can have the arbitrary data structures that rdf provides and a nice web interface for entering their data and versioning them17:38
kanzurelike this? http://drupal.org/project/rdfcck17:38
futuresoonyup17:38
futuresooni referenced that earlier17:38
kanzuresorry17:38
futuresoonno prob17:38
kanzurei'll try harder17:38
futuresooni've played with that to some extent17:38
futuresoonhah17:38
futuresooni've played with most of the sparql/rdf things that drupal has17:38
* fenn looks at http://www.stackfoundry.com/wedge/ (re git for cherrypy)17:39
futuresoonmy primary interest in drupal is that it's going leaps and bounds to trivialize some of the difficulties with rdf at present17:39
kanzurefenn: anyvcs isn't quite ready/functional yet17:39
kanzure:(17:39
fenncome on future, hurry up and get here17:39
futuresoonheh17:40
kanzurei think the main problems still aren't solved17:40
kanzurescrew the data format problem17:40
futuresoonkanzure: ok17:40
futuresoonso what are the main issues17:41
kanzureinstruction representation17:41
kanzurein terms of classes and code, not in terms of the data format17:41
kanzureone sec17:41
kanzurehttp://adl.serveftp.org/skdb/doc/proposals/action.py17:41
fenninstructions are a separate issue from dependency tracking17:41
kanzurei think we have dependencies down, even if we don't have our resolution engine yet17:41
fennthe philosophy maybe17:42
kanzurefuturesoon: the main reason for providing forms on the website for people to edit stuff is so that people who aren't computer wizards can get involved17:43
kanzurebut17:43
kanzureif they have to be computational linguist experts in order to help out at all, what's the point?17:43
futuresooni don't think drupal can replace the need for something in a good language like python to actually provide instructions to CNC XYZ machines and what have you17:43
fennthere is a lot of stuff like 'find me the relevant data on this ANSI standard' but people dont seem to want to do that kind of legwork17:43
kanzurefenn: because nobody has access to ANSI standards :(17:44
fennyou can find it in machinery's handbook, which you can download from me for the asking17:44
futuresoonbut if you had the data structures, even if it took rdf wizards to create the forms, it would allow people to make simple edits in the event they can't build/make what they apt-get17:44
kanzurefuturesoon: one idea i was thinking of was letting people type mostly plaintext instructions, but in a templating syntax with parametric variables or something17:44
kanzureand then package maintainers would come by and integrate the information more thoroughly17:44
fenni think that is dangerous17:45
kanzurefuturesoon: is it important to generate these forms? i mean, i can think of a way to easily do that in python, but it doesn't seem important to me.17:45
fennthe plaintext crap will outpace the rate of assimilation and it will turn into thingiverse/instructibles overnight17:45
kanzuretemplates aren't plaintext crap necessarily17:46
fennyeah they're crap in sheep's clothing17:46
kanzurelet's say you pass a few variables to the Screw __init__'s17:46
kanzureer17:46
kanzureScrew's __init__17:46
fennsuperficially looks like good data... until you try to do anything with it17:46
kanzurethen could you remind me why i made this wiki/git thing17:46
fenni don't know :(17:47
futuresooni can make a case for webforms, but if the CAD files or something have metadata associated with them, webforms for those may not be important17:47
kanzureit was my understanding that we wanted to load up the yaml in a sandbox17:47
fennpart of me wants to believe that 99% of humans aren't total morons17:47
futuresooni think where the webforms come in is if you want to represent data to people on the web where you'd benefit from network effects of people checking that data17:47
kanzurebut what's wrong with a giant <textarea> like there is currently?17:48
futuresoonasking me?17:48
* kanzure nods17:48
fenncurrently there's no syntax or semantic checking17:48
futuresoonwell if i have a special microsyntax that allows me to say what's a dependency in plaintext areas17:48
kanzurefenn: yeah that can be fixed though17:48
fennhow hard woul it be to get that operational?17:49
futuresoonthen that doesn't allow a web application where i can allow people to filter packages by prerequisites they already meet17:49
kanzurefenn: not very.. i think it would be a matter of yaml.load(shit_the_user_posted_to_the_server) and then route out errors and complain17:49
fennyou just have to try skdb.load(whatever) and print any errors17:49
kanzurestop reading my brain waves damn it17:49
kanzurefuturesoon: yes that's right17:49
fennno you!17:49
futuresoonthe most basic but useful thing i can think of where you'd want webforms is if you have users who would start building stuff but they don't even see projects that they can do because they don't know what prerequisites they meet17:49
kanzurefuturesoon: but i was talking more low-level actually. er, let me ask my question a different way17:49
futuresoonok17:49
kanzurefuturesoon: do you know about the HTML "input" and "textarea" elements?17:50
futuresoonsure17:50
futuresoonin that sense all the fields i'm thinking of are textareas, checkboxes, various things17:50
kanzureis there any particular reason to give users "input" elements instead of the giant "textarea" element that you saw on heybryan.org:8081 ?17:50
futuresoonwell textareas cover 95% of it17:50
fennthey do?17:50
kanzurethe yaml can be broken up into multiple small "input" elements, but that limits what a user can type in17:51
futuresooni would certainly think so17:51
fenni thought the point of cck was it puts the form values into predefined rdf17:51
fenntextarea is just unformatted crap as far as the computer is concerned17:51
kanzurefenn: cherrypy passes all form variables as a dictionary to your python function17:51
futuresoonrdfcck (but not cck more generally) supports turning cck data into rdfa (or straight rdf "flat file" databases)17:51
fennbut only on a per-form basis17:51
fennyou can't do anything with a huge block of text17:52
futuresoonbut people only input it as textareas, after that it can be represented many ways17:52
futuresooncck and views support all sorts of things----if it were a google map, the "input" coudl be that a user clicks on the map17:52
futuresoonand how that's represnted could also be a map17:52
kanzurei feel like we're talking past each other17:52
fennbtw why do people refer to it as 'flat file database'? is there some preferred default form for storing rdf?17:52
futuresoonkanzure: right, i dont' get the thing about textareas other than that i agree, textarea inputs make sense most of the time17:53
kanzureyes i know data can be converted to various forms on the back end. i was asking about some ui preference thingy17:53
kanzureblah17:53
kanzureforget it :)17:53
kanzureeverything's fine17:53
fennsingle big textarea = wiki bullshit = no semantic tagging17:53
futuresoonfenn: my thinking too17:53
kanzureif you open up a text editor to edit some yaml file,17:53
kanzureyou can add in all sorts of attributes and values where-ever you want17:53
kanzurebut if you have a pre-generated form, you're limited to editing the values of the pre-existing attributes17:54
kanzureyou can have some dynamic javascript stuff to add more form elements17:54
kanzurebut to be honest, what does that really buy you? :(17:54
futuresoonlet me show you something17:54
fennuser guidance17:54
kanzureis that important?17:54
fennpeople don't read remember17:54
futuresoonhttp://www.youtube.com/watch?v=L6NUFxWVEOw17:54
futuresoonthis is for making arbitrary ontologies17:54
futuresoonmapping these to cck fields is also arbitrary17:54
kanzuresigh17:54
kanzureyes you've gone over this17:55
kanzurethe mapping isn't the issue17:55
-!- genehacker [i=genehack@wireless-128-62-38-104.public.utexas.edu] has joined #hplusroadmap17:55
genehackerI need to ask those replab people how far they want to replicate17:55
kanzurewhy17:55
genehackerI'm just wondering some things about what they actually want to accomplish17:56
fenn"to an economically significant degree"17:56
kanzureask us17:56
kanzurewe're pretty much the most of them..17:56
kanzureexcept marcin, because he's too cool for irc or something :)17:56
fennreplab is more about access for non institutional groups than about replication per-se17:56
genehackerare they just going to make structural components for things like the laser cutters, 3d printers, etc17:56
kanzurewhat is a "structural component"?17:57
kanzurethat sounds very vague17:57
fennelectronics, free designs, active components like tool bits: all fair game17:57
fennnobody would agree that making your own laser diode is economically feasible at this point17:57
futuresoonkanzure: i think i have an answer to what it "buys" you to have these javascripty web widgety doodads17:58
genehackerare they going to go as far as making their own bearings, linear slides, extruded aluminium hex bar, screws, nuts and bolts?17:58
fennyou damn kids and your newfangled doodads17:58
genehackeror will they buy those?17:58
futuresoonkanzure: i think the answer is that if it's something you'd like to represent, sort, filter, in a web app, then it makes sense to have a web interface for a community to add value to that data17:58
fennwhy can't you be happy with vi like me and my grandpa17:58
futuresoonheh, i love vim17:58
kanzurefuturesoon: but it can already do that the way it is17:59
genehackerpah everyone knows notepad is the one true text editor17:59
genehackerit came first17:59
futuresoonkanzure: but as soon as it's something that you don't care how it's represented or manipulated or linked up over the web, it matters less. does that make sense?17:59
fenngenehacker: yes i think so17:59
kanzurenot really, no.17:59
genehackersounds ambitious17:59
futuresoonkanzure: it would be very hard for me to believe that without cck and views you can build a community around the data. some of that data maybe doesn't profit from having community around it as others though18:00
fennnot when you start shopping for extruded aluminum hex bar and find out you can't afford it18:00
genehackeralso from what ingots? bar stock?18:00
fennwtf you dont make bar stock from bar stock; that's idiotic18:00
futuresoonif you need an engineer to make a CAD file and that CAD file has metadata, it may not matter if the webforms are made a little bit easier for people to collaboratively edit18:00
kanzure"i love vim. vreally, i vo! now destroy the laser cannon before captain nerbov gets here!" <- sorry, when you said "i love vim" i started thinking of you as a russian scientist from the 50s18:01
genehackerI'd be more worried about making the spider dies then18:01
futuresoonheh18:01
futuresoonnuclear wessels18:01
kanzureright18:01
genehackerso ingots or bar stock?18:01
genehackeror mine the raw materials18:01
fenngenehacker: have you heard about makerbeam?18:01
genehackeryes18:01
fennthe guys behind that are really into replab18:02
kanzurein fact they created the replab mailing list18:02
futuresoonyeah some of those guys told me i should talk to kanzure18:02
fennyeah :(18:02
fennso here we are18:02
kanzurefuturesoon: oh, i thought genehacker said that18:02
kanzurewhich is funny because i've known genehacker for over a year now18:02
futuresoonsam something18:02
kanzureputman18:02
futuresoonyou and sam were debating on the mailing list i saw18:02
* kanzure shakes his head18:02
futuresoonright18:02
fennkanzure: on irc, nobody can hear you laugh/scream18:02
genehackerso to make makerbeam we need spider dies18:02
futuresoonkanzure: i think your'e BOTH right :-)18:03
genehackerfor an extruder18:03
kanzurefuturesoon: about what?18:03
futuresooni think sam putnam is emphasizing the web app community aspect and you're saying that unless it's a package that actually works to make machine readable stuff, it's just another instructables18:03
kanzureyay someone listens18:04
kanzurefuturesoon: hey, you distracted me >:( 18:04
fennfuturesoon: so are you the guy who sam was talking about building a funding community website thingy like oshwbank.org ?18:04
kanzurefuturesoon: i was gonna give you a potential, small easy task 18:04
genehackeralso why do they want to only punch out circular forms18:04
futuresoonno, i sent marcin an email a few days ago, my first contact. sam got back to me and saw the site stub i made18:04
futuresoonand now i'm here18:04
kanzurefuturesoon: the git diff view on web.py that fenn was asking for ;-) it can be made by adding a function to the Package class in web.py18:04
genehackerwith a sufficient press it'd be possible to make so much more18:04
kanzurefuturesoon: how did you meet marcin?18:04
futuresoonkanzure: just sent him an email a few days ago when i saw replab.org needed a web developer18:05
kanzureer but how did.. uh18:05
kanzurehey do you do templates? :)18:05
futuresoonhah18:05
futuresooni need to learn me some python18:05
kanzureand by templates i don't mean stealing them18:05
fenngenehacker: i think the ironworker thing is pointless.. he should just use the plasma cutter he already has18:05
futuresoonnot sure what i'm being asked anymore? :-)18:05
kanzurefuturesoon: the easiest way is to set up an interpreter. go give http://bpython-interpreter.org/ a try18:05
genehackeriron worker thing?18:05
kanzurefuturesoon: i was wondering if you know CSS/xhtml/how to make stuff not look like http://heybryan.org/18:06
fenngenehacker: i thought that's what you were talking about "only punch out circular forms"18:06
futuresoonkanzure: i didn't steal that replab.org template18:06
futuresooni made it from gimp in 2 days18:06
kanzurefuturesoon: sorry i wasn't saying you did18:06
futuresoonfrom 960.gs18:06
futuresoonno no, i don't take offense18:07
kanzurei didn't know you made it18:07
futuresooni have ALSO stolen templates :-)18:07
futuresoonand have no problem doing so18:07
kanzurei've just been meeting people who claim nto be web designers, but they just steal templates18:07
futuresooni'm just expanding my horizons18:07
fenni dont see what's wrong with stealing templates18:07
futuresoonright, nothing wrong with it18:07
futuresoonthis time i happened not to18:07
kanzurepeople will say "hey that looks like that other site!"18:07
fennboo hoo18:07
kanzurethat's not bad?18:07
fennjust dont steal it from your direct competitor18:07
kanzurewell18:07
kanzureshould've told me that earlier18:07
kanzurei did that once..18:08
fennidiot18:08
kanzurei was 11 :(18:08
futuresoonthe replab.sigmanode.com site has 4 layouts depending on whether there's left, right, both or no sidebars18:08
futuresoonobviously like most templates18:08
genehackerplasma cutters need tungsten don't they?18:08
futuresoonexcept i made this one18:08
fennbtw can we take down that fritzing picture18:08
kanzurefritzing?18:08
kanzurefenn: i'll probably be leaving to go to a meeting later this evening, soon18:08
genehackermade what?18:09
fennthe breadboard bubble flyin submarine phonograph thing18:09
genehackera plasma cutter head?18:09
kanzurefenn: i was wondering if you took a look at doc/presentations/hplus-summit-2009/outline18:09
fennyes very good18:09
kanzurehttp://adl.serveftp.org/skdb/doc/presentations/hplus-summit-2009/outline18:09
kanzureok18:09
fenndo you update that regularly/automatically?18:09
fennadl/skdb18:09
kanzurei do it regularly but not automatically18:09
genehackerhmm... where is that submarine picture thing18:10
genehackerI need to save it18:10
genehackerand make it in real life18:10
fennhttp://fritzing.org/18:10
kanzureoright18:10
kanzurehaha, yes we probably should18:10
fennit is a cool picture18:11
genehackerneeds more fins18:11
genehackerneeds to have the ability to swarm18:12
futuresooni love the fritzing concept18:15
genehackerthat periscope needs to be shortened18:15
-!- futuresoon [n=futureso@cpe-68-175-67-66.nyc.res.rr.com] has quit [Remote closed the connection]18:24
genehackeralso does anyone know of any good resources on NC18:24
genehackercontrol18:24
genehackerreally old really simplified NC machines18:24
genehackerand how they worked18:25
-!- futuresoon [n=futureso@cpe-68-175-67-66.nyc.res.rr.com] has joined #hplusroadmap18:26
-!- jasonwohlfahrt [n=chatzill@cpe-72-179-52-26.austin.res.rr.com] has quit [Read error: 110 (Connection timed out)]18:26
genehackerI guess that submarine could use fluidic control to be really really cheap to make18:27
genehackerwouldn't have much capability18:27
genehackerbut it'd be cheap18:27
-!- El_Matarife [n=El_Matar@ppp-69-148-69-156.dsl.rcsntx.swbell.net] has joined #hplusroadmap18:27
futuresoonis there a skdb.py?18:35
futuresooni'm a python noob no doubt, but if it says import skdb does that mean there's a skdb.py?18:35
futuresoonwell we'll see how the install script goes18:46
futuresooni need to start getting some resources to build some cnc xyz machines18:46
futuresoonthat's basically what all this is intended to run, right? things like in the replab roadmap/18:47
-!- genehacker [i=genehack@wireless-128-62-38-104.public.utexas.edu] has quit [Read error: 110 (Connection timed out)]18:53
kanzurefuturesoon: yeah there's an skdb.py19:32
kanzurebut there's also a module19:32
kanzurefuturesoon: yes19:32
kanzurefuturesoon: see skdb/core/skdb.py19:32
kanzurebut skdb/core/__init__.py is also somewhat informative19:32
futuresoonokay19:34
futuresooni might have a pythonpath problem19:35
futuresoonskdb/core/__init__.py is just from skdb import *19:35
futuresoonmaybe you mean settings.py?19:36
futuresoonok i think i'm close now19:37
futuresoonnow i get AssertionError: /usr/local/share/skdb/threads: skdb package not found or unreadable19:37
futuresoonfrom packages/screw/screw.py19:38
futuresoonokay now it runs it just doesn't output anything19:39
futuresoonshould i be using it in interactive mode?19:39
futuresooni should have been running the lego demo anyway, that's what i wanted19:41
futuresoonskdb.geom not loaded, load_CAD not available19:41
futuresoonbut it runs now19:41
futuresoonmore so19:41
futuresooni get some compatibility assessments about anti stud and stud cup19:42
futuresooni assume that means it's working although paths.py still has some OCC.gp stuff i have to work out19:42
kanzurewhat are you running19:48
kanzure"no output" when you run what?19:48
futuresooni ran the lego demo and it basically worked although it threw some exceptions19:48
kanzureyou should try running skdb/paths.py - that's an interactive 3D visualizer19:48
futuresoonthat time it was the screw.py19:48
kanzuredid you see various legos in 3D?19:48
futuresoonpaths.py relies on OCC.gp which isn't working now19:48
futuresoonthought the installer mentioned OCC but i guess i don't have it19:48
kanzureyeah you need that to do anything interesting with skdb19:49
futuresoonthe install script you have up there19:49
kanzurekind of a flaw :(19:49
futuresooncan i apt-get it? i don't know what a gp extension is19:49
kanzurehopefully the installation notes are clear enough19:49
kanzureyes you can apt-get it per the notes19:49
kanzuresee skdb/doc/installation19:49
futuresoonk19:50
kanzurebut you can't apt-get pythonocc, there's a few notes on how to get that working (which is what's giving you your error)19:50
kanzuresorry i'm distracted at the moment. if you yell at me loudly enough though i'll be able to answer questions :)19:50
kanzure(in a meeting)19:50
futuresoonno prob19:51
-!- genehacker [i=genehack@wireless-128-62-53-22.public.utexas.edu] has joined #hplusroadmap19:57
-!- genehacker [i=genehack@wireless-128-62-53-22.public.utexas.edu] has quit [Read error: 110 (Connection timed out)]20:26
-!- genehacker [i=genehack@wireless-128-62-38-220.public.utexas.edu] has joined #hplusroadmap20:57
-!- jasonwohlfahrt [n=chatzill@cpe-72-179-52-26.austin.res.rr.com] has joined #hplusroadmap21:27
-!- Juul_ [n=Juul@h55eb1609.dkkoost.sta.perspektivbredband.net] has joined #hplusroadmap21:39
futuresoonkanzure: you still busy?21:39
-!- Juul_ [n=Juul@h55eb1609.dkkoost.sta.perspektivbredband.net] has left #hplusroadmap ["Leaving"]21:40
kanzurefuturesoon: yeah but i'm around21:41
kanzurewhat's up?21:41
-!- futuresoon [n=futureso@cpe-68-175-67-66.nyc.res.rr.com] has quit [Remote closed the connection]21:51
-!- futuresoon [n=futureso@cpe-68-175-67-66.nyc.res.rr.com] has joined #hplusroadmap21:51
futuresooni was just looking at the pdf updates-from-austin. was going to shoot the breeze about that21:52
futuresooni can wait though21:52
fennpythonocc is a bitch22:54
fenni expect eventually it will just work all the time22:55
futuresoonfenn: would that day were today :-)23:18
futuresoonyeah i haven't gotten it to work yet23:18
-!- jasonwohlfahrt [n=chatzill@cpe-72-179-52-26.austin.res.rr.com] has quit [Read error: 60 (Operation timed out)]23:28
fennsecond time in two days i've run across the phrase "design noir"23:28
fennthird actually23:30
futuresooninteresting23:33
futuresooni doubt the nipple chair will ever go into mass production23:35
fenni still havent really figured out wtf it is about23:36
futuresoonsounds like this http://www.ugotrade.com/2009/01/28/pachube-patching-the-planet-interview-with-usman-haque/23:36
fennname sounds familiar23:37
futuresoonso replab.org23:37
fenn"operating systems for architecture" or something like that23:37
futuresoonyeah, an Urban Versioning System23:38
futuresoonhow long have you guys been working on skdb?23:38
futuresooni was looking at the updates from austin presentation23:38
fenn1.5 years since i had the idea, but only really working on it for 7 or 8 months23:38
futuresoonlooks like some of the problems of generalizing the system can be fudged for at least some of the grosser parts they're doing at openfarmtech?23:38
futuresooni see23:39
futuresoonb-rep for instance or manufacturing process specifications23:39
fennplease elaborate23:40
futuresooni know those are the awesome parts. i don't know exactly to what extent marcin has succeeded in standardizing those23:40
futuresoonhe's gotten a lot of mileage out of cutting and welding23:40
futuresoonusing bolts for gridbeam type structures23:40
fennhe hasn't done anything that requires high precision is why23:40
fenncan't make a diesel engine out of grid beam23:41
futuresoonvery true23:41
futuresoonsome of these problems like b-rep and manufacturing process specification look like they're subject to human input (you just claim that the b-rep will work out) or selecting specific ontologies and making design choices in that (having a limited number of manufacturing processes)23:43
fenn"[pachube is] A web service that enables people to tag and share real time sensor data from objects, devices and spaces around the world, facilitating interaction between remote environments, both physical and virtual."23:43
futuresoonright23:43
fennthe multiply-anticipated AR killer app23:43
fennwell, are the formats and API's open? that's all that matters in the end23:44
futuresoonformats and APIs for replab.org?23:44
fennfor pachub23:44
fennsorry i'm easily distracted23:44
futuresoonpachube intends to open source its code so anybody could run a copy23:44
futuresoongrid computing type stuff23:44
fennb-rep is a type of cad geometry, like mesh or CSG23:44
futuresoonyeah i was reading about b-rep23:45
fennit's only midnight; i shouldnt be having this much trouble understanding23:45
futuresooni guess i'm trying to figure out what the time horizon is for skdb being useful for making replab.org23:46
fennah23:47
futuresoongiven the tricky generalization problems23:47
futuresoonand whether those can be fudged temporarily23:47
fennkludges have a way of not being temporary23:47
futuresoonsee, this is why i look at rdf and think "bingo"23:47
futuresoonbecause if you had fudges in rdf, they're no longer fudges23:47
futuresoonyou just map old data structures to new ones, and those become the new standard23:48
futuresoondoes that not sound right?23:49
fenni dont think it's always so easy23:49
futuresooni don't think anybody "knows"23:50
futuresoonthere's not masses of rdf data out there yet23:50
fennlike if standard 1.0 lumps all animals together but 2.0 wants to know if they're warm or cold blooded23:50
futuresoonor at least not for things as subject to emphasis and selection as this23:50
futuresoonthat's easy in rdf23:50
fennbut the data wasn't there when it was being entered into 1.023:51
fennit's just animal 1 animal 2 animal 323:51
futuresoonbut then you just have 2.0 reference 1.023:51
futuresoonand also add new classes23:51
futuresooneach entry in 2.0 actually gives the URI for 1.0 objects23:51
futuresoonand adds info23:51
fennbut it's not a 1:1 mapping23:51
fennwhy is there no rdf data yet?23:52
futuresoonmy personal opinion is that the reason there's not lots of rdf data is because there's not content management systems that can output it23:52
fenni remember this quote from the usman haque interview23:53
fenn“…I think total openness is fatal.  It’s like lying in a blazing sun under a sky full of vultures, naked. It’s also rather rude, like babbling anything or anything that flies into your head and still expecting people to pay attention.”23:53
futuresoonyeah, i thought that quote was interesting23:53
fennbruce sterling actually23:54
futuresooni had to read that article several times to understand the implications23:54
futuresooncertainly rdf is more queryable than yaml23:54
futuresoonmore subject to manipulation23:54
fenni guess23:55
futuresooni think it's really the difficulty of even visualizing all the classes in an ontology in the first place that are one reason it's difficult to manually enter rdf23:55
fennyou know what we're doing with yaml and objects?23:55
kanzuredoes rdf have a "schema" ?23:55
fennits not just raw dictionaries; each yaml entry is parsed as a python class (or any other language)23:56
futuresoonkanzure: it has things like literals, subclasses of, etc.23:56
futuresoonfenn: things with a ! in front are python classes, right?23:56
kanzureeh23:56
kanzurethey are all python objects23:57
kanzurethe !tagtagtag stuff are particular classes though23:57
fenn!tags map to classes in whatever language you tell it to23:57
futuresooni see23:57
fenna python specific tag is something like !!python/object:blahblah23:57
kanzurein python we specify those by inheriting from yaml.YAMLObject and adding an attribute to the class "yaml_tag" which is of the same value like "!tagtagtag"23:57
kanzurefuturesoon: so, an ontology isn't necessary here23:58
fenni wouldnt say that23:58
fennthere's still an ontology, it's just more implicit in the code :(23:58
kanzurethis is all just serialization. the most important thing is code that runs and "makes sense"23:59
fenni find it hard to explicitly declare an ontology, it's like intellectual masturbation or something23:59
kanzureer, code that runs, and a way for objects to parse themselves or their own data23:59
kanzurethe code should also make sense of course ;-)23:59

Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!