--- Day changed Mon Aug 03 2009 | ||
fenn | yes that sounds useful | 00:00 |
---|---|---|
kanzure | and then write a method that makes yaml be ok with certain tags not corresponding to classes it knows about | 00:00 |
fenn | or better yet ignore them | 00:00 |
kanzure | maybe exception handling? | 00:00 |
fenn | if that works | 00:00 |
kanzure | doesn't yaml throw an exception upon coming across a foreign tag? | 00:00 |
fenn | i dont think you can get that much info about an exception at runtime | 00:00 |
fenn | yes but how do you know it's in your list of expected missing tags | 00:01 |
kanzure | hm there's probably no way to force yaml to continue | 00:01 |
kanzure | ConstructorError | 00:01 |
fenn | yeah better to just make a dummy object | 00:02 |
fenn | metadata lists the classes used for this package; maybe it could list tags too | 00:03 |
fenn | that's sort of what i meant anyway | 00:04 |
fenn | with the classes field | 00:04 |
kanzure | yes but I was originally thinking about moving template.yaml into metadata.yaml | 00:04 |
fenn | 'these are the classes you need in order to parse data.yaml' | 00:04 |
kanzure | moving template.yaml into metadata.yaml will not work (unless I can do some fancy yaml exception handling) | 00:04 |
fenn | oh well | 00:05 |
fenn | skdb.load(foo) should just make dummy objects with the corresponding tags then | 00:06 |
kanzure | how would it do that if yaml crashes before you can figure out what the tags are | 00:06 |
fenn | if you want to look at template.yaml without loading the python classes | 00:06 |
fenn | template.yaml still a separate file | 00:06 |
fenn | so you do that between files | 00:06 |
fenn | maybe the document separator '---' could be used, but i still don't really know how that works | 00:07 |
kanzure | I get it, but what do you mean by "between files" | 00:07 |
fenn | load metadata.yaml; do stuff(); load template.yaml | 00:07 |
kanzure | okay | 00:08 |
kanzure | you have any idea where ConstructorError is defined? | 00:08 |
kanzure | there's yaml.YAMLError but nothing else really | 00:08 |
fenn | you shouldnt get any error | 00:08 |
kanzure | aha, yaml.constructor.ConstructorError | 00:08 |
kanzure | when I do a "try, except:" how do I get the "this" or "self" in the exception code block? | 00:10 |
kanzure | aha | 00:10 |
kanzure | except SomeError, myvar: | 00:10 |
fenn | hah http://lesswrong.com/lw/xr/in_praise_of_boredom/ came up when searching for scifi stories about aliens without a sense of boredom | 00:11 |
kanzure | it's almost a holiday tradition for singularitarians to discuss overlord ai that attempt to please all of our senses | 00:11 |
kanzure | and boredom into infinity-future-doubleplus-bad scenarios | 00:12 |
kanzure | "expected paperclip maximizer" sigh | 00:12 |
QuantumG | you don't need aliens.. you can talk to people without a sense of boredom.. go ask your grandparents if they were ever bored growing up. | 00:13 |
fenn | they were too busy getting chased by wooly mammoths to get bored | 00:13 |
kanzure | try: yaml.load("hah: !blah") | 00:13 |
fenn | or was it the other way around | 00:13 |
kanzure | except yaml.constructor.ConstructorError, vroom: | 00:13 |
kanzure | print vroom.problem_mark | 00:13 |
kanzure | or print vroom.problem | 00:13 |
fenn | um | 00:14 |
kanzure | from now on I vow to name all of my variables after official NASCAR sounds | 00:14 |
kanzure | including the classic 'kaboom' | 00:15 |
fenn | where did you get 'problem' from? | 00:15 |
kanzure | found it through the all knowing power of bpython | 00:16 |
fenn | weird that it shows up in dir() but not help() | 00:16 |
kanzure | someone needs to beat the bpython developers with a baseball bat over the head (hell, through the head) | 00:17 |
kanzure | scrolling should be implemented by now | 00:17 |
kanzure | neat. so, to figure out if the problem is what you think it is, | 00:18 |
kanzure | vroom.problem_mark.buffer[(vroom.problem_mark.index)] | 00:18 |
kanzure | which should return "!" if it's a tag error | 00:19 |
kanzure | and prepend : to the ] if you want to get the tag name | 00:19 |
fenn | tag could have data after it | 00:19 |
fenn | foo: !tag bar | 00:19 |
kanzure | are spaces and newlines and EOF the only way to terminate a tag name? | 00:20 |
fenn | nothing a regex couldnt fix | 00:20 |
fenn | i think so | 00:20 |
kanzure | what about special characters? | 00:20 |
fenn | comment | 00:20 |
kanzure | yaml should ignore those anyway | 00:20 |
kanzure | not sure if that's in the error or not | 00:20 |
kanzure | ok it's in the error | 00:21 |
kanzure | I wonder if this is the only type of ConstructorError | 00:23 |
kanzure | if so, you could just regex vroom.problem (not vroom.problem_context) | 00:23 |
kanzure | which looks like this: "could not determine a constructor for the tag '!bar'" | 00:24 |
fenn | there's another tricky yaml thing i've been too scared to use yet, called 'implicit path resolver' | 00:24 |
kanzure | what does it do? | 00:25 |
fenn | basically if the mapping value (foo in the above example) has a certain string like 'foo' then parse it as this sort of object | 00:25 |
kanzure | I thought you did that with the dice example? | 00:25 |
kanzure | 26d4 | 00:25 |
fenn | but it's not even in the pyyaml docs and may not exist elsewhere besides pyyaml | 00:25 |
fenn | that's just a regular old implicit resolver | 00:25 |
kanzure | so you do typing based off of key name? | 00:26 |
kanzure | how is that different from __setvalue__()? | 00:26 |
fenn | help(yaml.add_path_resolver) | 00:27 |
fenn | you mean __setstate__ ? | 00:27 |
kanzure | yes | 00:27 |
fenn | setstate lets you do anything, pretty much | 00:28 |
fenn | oh wait | 00:28 |
fenn | setstate is a method of the class | 00:28 |
fenn | that doesnt work so well if you dont know what type to parse the data as | 00:28 |
kanzure | right | 00:29 |
kanzure | well that was a pointless exercise (since document separators are probably a better idea) | 00:31 |
kanzure | heh from SA today, some goon doctor decided to cut open a poison ivy rash and pour rubbing alochol into it | 00:38 |
kanzure | and now can't stop puking, vomiting, can't hold down any water, is dehydrated and apparently decided to go outside in a blacksuit for a few hours (he's a funeral minister or something) | 00:38 |
fenn | yay open sores medicine | 00:40 |
fenn | i hereby claim trademark on the phrase 'open sores medicine' - infringement punishable by death | 00:41 |
kanzure | quick, register opensores.org, biosores.org, bioopensores.org, sorebioopen.org, .. | 00:42 |
kanzure | what else are you going to do with millions and millions of dollars in funding? | 00:42 |
kanzure | heh "square peg in a round hole" http://www.inventgeek.com/Projects/BreadBox/Overview.aspx | 00:47 |
fenn | wow the internet really has thought of it all before | 00:49 |
fenn | wait a minute, that was jon cline i was talking to about this | 00:49 |
-!- genehacker [n=noko@pool-173-57-41-223.dllstx.fios.verizon.net] has quit [Read error: 110 (Connection timed out)] | 00:50 | |
kanzure | who is the author of this page? | 00:50 |
kanzure | "jared buock" | 00:50 |
fenn | kanzure: how did you come across this link? | 00:51 |
kanzure | fenn: jonathan cline | 00:52 |
kanzure | email to diybio | 00:52 |
fenn | right now i see it | 00:53 |
fenn | jared bouck sure gets up to a lot of good | 00:57 |
fenn | paintball turret with wireless heads up display | 00:57 |
kanzure | i want that mounted on the apartment door by tomorrow morning | 00:58 |
fenn | i think everything on that site is the same author | 00:58 |
fenn | hehe 'stealth server' http://www.inventgeek.com/Projects/ProjectSilver/Overview.aspx | 01:05 |
CIA-43 | skdb: kanzure * r0a9556696a9a / (core/skdb.py core/threads.py doc/todo/pymates-todo): worked on the Package class- not quite done with basic package loader. | 01:12 |
-!- wrldpc2 [n=benny@pool-173-48-253-182.bstnma.fios.verizon.net] has joined #hplusroadmap | 03:40 | |
-!- wrldpc2 [n=benny@pool-173-48-253-182.bstnma.fios.verizon.net] has quit [] | 04:06 | |
-!- wrldpc2 [n=benny@pool-173-48-253-182.bstnma.fios.verizon.net] has joined #hplusroadmap | 04:11 | |
-!- nsh [i=sbp@wikipedia/nsh] has quit ["Leaving."] | 05:24 | |
-!- nsh [i=sbp@59.176.232.72.static.reverse.ltdomains.com] has joined #hplusroadmap | 05:25 | |
-!- Netsplit holmes.freenode.net <-> irc.freenode.net quits: CIA-43 | 05:30 | |
-!- Netsplit over, joins: CIA-43 | 05:30 | |
-!- drazak [n=drazak@drazak.net] has quit [Read error: 104 (Connection reset by peer)] | 05:40 | |
-!- drazak_ [n=drazak@drazak.net] has joined #hplusroadmap | 05:40 | |
-!- Netsplit holmes.freenode.net <-> irc.freenode.net quits: drazak_, CIA-43 | 05:53 | |
-!- Netsplit over, joins: drazak_, CIA-43 | 05:54 | |
CIA-43 | skdb: kanzure * rbd27d76de79b /core/skdb.py: python source file will not be immediately loaded (also, path+.py is not sufficient) | 06:44 |
-!- Netsplit holmes.freenode.net <-> irc.freenode.net quits: drazak_, CIA-43 | 06:51 | |
-!- Netsplit over, joins: drazak_, CIA-43 | 06:54 | |
-!- Netsplit holmes.freenode.net <-> irc.freenode.net quits: drazak_, CIA-43 | 06:55 | |
-!- Netsplit over, joins: drazak_, CIA-43 | 06:57 | |
kanzure | http://j.photos.cx/modeltatbridge-7d4.jpg | 07:42 |
-!- Netsplit holmes.freenode.net <-> irc.freenode.net quits: goonie, boogles, Utopiah, drazak_, katsmeow-afk, nchaimov_, splicer, strages, fenn, flazmot, (+10 more, use /NETSPLIT to show all of them) | 08:38 | |
-!- Netsplit over, joins: drazak_, wrldpc2, splicer, xp_prg, tarbo2, nsh, fenn, CIA-43, strages, ybit (+10 more) | 08:47 | |
-!- Netsplit holmes.freenode.net <-> irc.freenode.net quits: drazak_, CIA-43 | 08:54 | |
-!- Netsplit over, joins: drazak_, CIA-43 | 08:55 | |
-!- Netsplit holmes.freenode.net <-> irc.freenode.net quits: drazak_, CIA-43 | 09:15 | |
-!- Netsplit over, joins: drazak_, CIA-43 | 09:16 | |
ybit | %TERM | 09:32 |
-!- El_Matarife [n=El_Matar@adsl-68-88-200-16.dsl.rcsntx.swbell.net] has joined #hplusroadmap | 09:32 | |
ybit | echo %TERM | 09:32 |
ybit | argh | 09:32 |
CIA-43 | skdb: kanzure * r2da169d1173e /core/ (settings.py skdb.py): settings.py created, has some paths and hacks for paths for hackers | 09:35 |
CIA-43 | skdb: kanzure * r6089db52e9c4 /core/skdb.py: worked on metadata loader for packages | 09:35 |
kanzure | fenn: maybe you should put some todo notes in skdb/doc/todo/ | 09:42 |
kanzure | fenn: I'm trying to say "def __init__(stuff): self = self.yaml_loader_method_here(blah)" but that doesn't seem to work | 09:49 |
kanzure | any ideas? | 09:49 |
kanzure | nevermind, I did a workaround | 09:53 |
kanzure | was hoping that I could say lego = Package("lego") and have it load whatever "lego" package might exist, and if there's not one, then carry about its business | 09:53 |
kanzure | dummy tags | 10:27 |
kanzure | link? | 10:28 |
fenn | gah | 10:28 |
kanzure | you suck | 10:28 |
fenn | http://yaml.org/spec/current.html#id2503753 | 10:28 |
kanzure | %TAG ! tag:clarkevans.com,2002: | 10:29 |
kanzure | http://yaml.org/spec/current.html#global%20tag/ | 10:29 |
kanzure | bleh | 10:29 |
kanzure | yaml.load("%TAG ! wee\n--- !blah") | 10:32 |
kanzure | yaml.load("%TAG ! wee:\n--- !blah") | 10:32 |
kanzure | ConstructorError: could not determine a constructor for the tag 'globalblah' | 10:33 |
kanzure | for yaml.load("%TAG ! global\n--- !blah") | 10:34 |
fenn | mine says could not determine a constructor for the tag 'wee:blah' | 10:34 |
fenn | anyway you're doing it wrong | 10:34 |
kanzure | if you had a colon after wee then that's why it says wee:blah | 10:34 |
kanzure | basically it's a way of saying "load it up into this object" | 10:36 |
kanzure | so in this case we'd have a global class "wee" | 10:36 |
kanzure | and then we'd tell wee to always hasattr() for "blah" or something | 10:36 |
kanzure | for any "blah" I mean | 10:36 |
kanzure | yaml.load("%TAG !yaml! tag:\n--- !yaml!blah2") | 10:37 |
kanzure | then it says "could not determine a constructor for the tag 'tag:blah2'" | 10:37 |
kanzure | that's a really bad hack though | 10:38 |
fenn | what's a bad hack? | 10:39 |
kanzure | making a class that always says "yes I have that attribute" | 10:40 |
fenn | is that possible? | 10:40 |
kanzure | class test: def __hasattr__(self,attr): .. | 10:40 |
fenn | hasattr isnt a method | 10:40 |
kanzure | does hasattr() access __getattribute__ ? | 10:41 |
fenn | you can overload hasattr but i think that would break the rest of the program | 10:41 |
kanzure | help() says hasattr calls getattr | 10:41 |
kanzure | so I suspect __getattribute__ may be the way to go | 10:41 |
* kanzure tests | 10:41 | |
fenn | wait a minute, how does this fix the yaml tag issue? | 10:42 |
kanzure | guess it would still fudge up the file | 10:44 |
kanzure | nevermind | 10:44 |
fenn | class dummy(): pass | 10:46 |
fenn | yaml.load('foo: !!python/object:__main__.dummy \n stuff: 1234') | 10:46 |
fenn | {'foo': <__main__.dummy instance at 0x98cd04c>} | 10:47 |
fenn | when i do foo.stuff why do i get AttributeError: 'dict' object has no attribute 'stuff' | 10:49 |
fenn | it's not a dict as far as i can see | 10:49 |
kanzure | it needs to be able to init the object | 10:49 |
kanzure | yaml needs to be able to init the object, I mean | 10:49 |
fenn | foo.__init__() seems to work | 10:50 |
kanzure | how about foo() ? | 10:50 |
fenn | somehow yaml manages to extract the stuff attr when i do yaml.dump(foo) | 10:50 |
fenn | foo = yaml.load('foo: ... | 10:50 |
fenn | not a class | 10:50 |
fenn | dummy is the class | 10:51 |
fenn | oh heh | 10:51 |
fenn | it's a dict with 'foo':dummy() | 10:51 |
fenn | foo['foo'].stuff | 10:51 |
kanzure | did you forget how yaml.load() works again? | 10:52 |
fenn | i guess i didnt need the foo: in the yaml.load line | 10:52 |
kanzure | what about: !!python/object/new:module.Class | 10:56 |
kanzure | class Fake(yaml.YAMLObject): yaml_tag="!fake" @classmethod def from_yaml(cls, loader, node) .. | 11:05 |
kanzure | the test: yaml.load("blah: !fake test\n---\nhello: !test") | 11:06 |
kanzure | how is "class dummy_supertest(yaml.YAMLObject):" invalid syntax? | 11:10 |
-!- fenn [n=fenn@cpe-72-177-121-73.austin.res.rr.com] has quit [Read error: 110 (Connection timed out)] | 11:10 | |
kanzure | yaml.add_multi_constructor('!', your_constructor_function) | 11:10 |
-!- fenn [n=fenn@146.6.88.5] has joined #hplusroadmap | 11:14 | |
kanzure | yaml.add_multi_constructor('!', your_constructor_function) | 11:14 |
fenn | looks like the modem crashed. | 11:14 |
kanzure | 11:10 < xitology> the constructor function accepts three parameters: the loader, the tag suffix and the node and should return a python object corresponding to the node | 11:14 |
fenn | i dont see how to give it a tag suffix | 11:15 |
kanzure | "the tag suffix" is anything coming after the ! | 11:15 |
fenn | oh, multi_constructor is a function i'm supposed to pass to add_multi_constructor | 11:15 |
-!- fenn_ [n=fenn@cpe-72-177-121-73.austin.res.rr.com] has joined #hplusroadmap | 11:21 | |
fenn | how nice of you to join us | 11:21 |
-!- xp_prg [n=xp_prg3@c-76-21-115-162.hsd1.ca.comcast.net] has quit ["This computer has gone to sleep"] | 11:30 | |
-!- fenn [n=fenn@146.6.88.5] has quit ["leaving"] | 11:36 | |
-!- fenn_ is now known as fenn | 11:36 | |
CIA-43 | skdb: * ra7e2bae02eb3 /packages/lego/data.yaml: convert rot to vector; move height/diam attrs to geometry object | 11:52 |
CIA-43 | skdb: * r8b9bf46e87a1 /packages/screw/template.yaml: move thread interfaces to the proper place | 11:52 |
kanzure | wait, why isn't "thread: !thread" under "interfaces:" in screw/template.yaml? | 11:55 |
kanzure | oh nevermind | 11:56 |
kanzure | the way it is now is better actually | 11:56 |
fenn | not sure why you said i should use add_multi_constructor vs just add_constructor | 12:03 |
fenn | i think multi_constructor will take precedence over any tag with the prefix (so any !foo tag, basically) | 12:03 |
fenn | hmm constructor doesnt accept a tag suffix, nevermind | 12:04 |
-!- kardan| [n=kardan@p54BE4D9D.dip.t-dialin.net] has joined #hplusroadmap | 12:23 | |
kanzure | maybe if I drank less water, I wouldn't be able to sweat so much | 12:31 |
CIA-43 | skdb: kanzure * rf9824b0f28c6 /core/skdb.py: some helper methods for loading packages | 12:35 |
kanzure | fenn: why are you working on add_multi_constructor? | 12:37 |
fenn | i'm not, i'm writing unit tests | 12:37 |
kanzure | for what? | 12:37 |
fenn | tag_hack | 12:37 |
kanzure | why are you working on tag_hack is what I meant to ask :) | 12:37 |
fenn | wtf does that mean | 12:37 |
kanzure | that means that I got what I asked wrong | 12:37 |
kanzure | tag_hack is so that we don't have to include broken tags and be screwed for it | 12:38 |
-!- El_Matarife [n=El_Matar@adsl-68-88-200-16.dsl.rcsntx.swbell.net] has quit [Read error: 110 (Connection timed out)] | 12:41 | |
-!- kardan| [n=kardan@p54BE4D9D.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)] | 12:42 | |
CIA-43 | skdb: * r8868718a6656 /core/ (skdb.py yamlcrap.py): add tag_hack for loading yaml tags that we dont have classes for yet" | 12:46 |
CIA-43 | skdb: * r5f964b41611c /core/skdb.py: Merge branch 'master' of ssh://adl.serveftp.org/var/www/skdb | 12:46 |
fenn | i guess merge was smart enough to figure it out | 12:47 |
-!- kardan| [n=kardan@p54BE4D9D.dip.t-dialin.net] has joined #hplusroadmap | 12:47 | |
fenn | what is "#replace self's information with the loaded_package information" supposed to do? | 12:55 |
fenn | and what is assert not (os.listdir(settings.paths["SKDB_PACKAGE_DIR"]).count(path) | 12:57 |
fenn | shouldnt open_package just return skdb.load(package/data.yaml)? | 12:58 |
-!- El_Matarife [n=El_Matar@38.96.203.34] has joined #hplusroadmap | 13:00 | |
kanzure | assert not is the same as assert (not False) | 13:03 |
kanzure | hey El_Matarife | 13:08 |
kanzure | how did the meeting go? | 13:08 |
kanzure | fenn: "#replace self's information" meant that the idea was to load up the current object .. er.. like, if you say Package("screw"), it should load up the 'screw' package, and put it into *that* object | 13:09 |
kanzure | into *that* "self" | 13:09 |
kanzure | so anyway once you fix all your broken "improvements" you might be able to run it and see what it does | 13:24 |
kanzure | :p | 13:24 |
El_Matarife | It went pretty well | 13:36 |
El_Matarife | Checked out a lot of different machines, there was some workshops but I missed literally every single one but "How to make money on your CNC Machine" | 13:36 |
-!- drazak_ is now known as draz|lab | 13:43 | |
CIA-43 | skdb: * r9ab3a3a57442 /core/settings.py: make_pretty | 13:58 |
kanzure | fenn: I need you to push. | 14:06 |
CIA-43 | skdb: * r079fd512e220 /core/skdb.py: blarg | 14:07 |
CIA-43 | skdb: * r17f5fd8399dd /core/yamlcrap.py: need this too | 14:07 |
flazmot | its so funny tom barbalet has an article in h+ | 14:07 |
kanzure | hplusmagazine | 14:08 |
kanzure | h+ is hard to search for :/ | 14:08 |
--- Log closed Mon Aug 03 14:35:18 2009 | ||
--- Log opened Mon Aug 03 14:35:21 2009 | ||
-!- kanzure_ [i=bryan@dhcp-84-36.me.utexas.edu] has joined #hplusroadmap | 14:35 | |
-!- Irssi: #hplusroadmap: Total of 24 nicks [0 ops, 0 halfops, 0 voices, 24 normal] | 14:35 | |
-!- bkero_ [n=bkero@osuosl/staff/bkero] has joined #hplusroadmap | 14:35 | |
-!- Irssi: Join to #hplusroadmap was synced in 65 secs | 14:36 | |
-!- kanzure [i=bryan@dhcp-84-36.me.utexas.edu] has quit [Read error: 110 (Connection timed out)] | 14:46 | |
-!- bkero [n=bkero@osuosl/staff/bkero] has quit [Read error: 110 (Connection timed out)] | 14:47 | |
-!- bkero_ is now known as bkero | 14:48 | |
* ybit wants a few recommended torrents, maybe some anime, maybe some music. doesn't matter | 14:52 | |
draz|lab | man | 14:55 |
draz|lab | I've read like 40 papers today | 14:55 |
draz|lab | my brain is fried | 14:55 |
kanzure_ | http://kbyanc.blogspot.com/2007/07/python-aggregating-function-arguments.html | 15:06 |
kanzure_ | draz|lab: 40papers/day is nothing. faster! | 15:07 |
fenn | def test(**kw): print kw | 15:16 |
fenn | test(arg1='huh', arg2='wut') | 15:16 |
fenn | {'arg1': 'huh', 'arg2': 'wut'} | 15:16 |
fenn | i dont think there's a way to make it automatically parse the dict | 15:17 |
fenn | if you want to pass in a dict | 15:17 |
fenn | def test(**kw): kw['stuff'].update(locals()); print locals() | 15:19 |
fenn | and then i think you have to do del(kw) | 15:19 |
CIA-43 | skdb: * r38dc30326728 /core/skdb.py: meddling | 15:22 |
draz|lab | kanzure_: lol | 15:23 |
fenn | **kw just catches any left over named arguments | 15:25 |
kanzure_ | why do you have to del(kw) | 15:27 |
fenn | well maybe you dont, but i think it's a recursive dict otherwise? | 15:31 |
CIA-43 | skdb: kanzure * r78db5ec1a81c /core/skdb.py: fixed open_package | 15:31 |
CIA-43 | skdb: kanzure * r0c9d8ec29b96 /core/skdb.py: merged skdb.py - move things around | 15:31 |
kanzure_ | still getting the load error for pack = open_package("lego") | 15:36 |
kanzure_ | something about my updated metadata being totally wrong | 15:37 |
kanzure_ | which is just an appended "!package" at the top of the file | 15:37 |
fenn | you know i always meant for skdb.load(open('metadata.yaml')) to work | 15:44 |
CIA-43 | skdb: kanzure * rfe64290fb37e /core/yamlcrap.py: yamlcrap- let classes define whether or not their tag is for a mapping or a scalar | 15:47 |
CIA-43 | skdb: kanzure * red7dbbb44917 /core/skdb.py: remove Package().open() | 15:49 |
fenn | huh? why did you just delete that? | 15:53 |
-!- genehacker [n=noko@pool-173-57-41-223.dllstx.fios.verizon.net] has joined #hplusroadmap | 15:59 | |
genehacker | http://www.exo.net/~jyu/activities/column%20chromatography.pdf | 16:04 |
genehacker | simple column chromatography | 16:04 |
-!- El_Matarife [n=El_Matar@38.96.203.34] has quit [Read error: 104 (Connection reset by peer)] | 16:17 | |
-!- kardan| [n=kardan@p54BE4D9D.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)] | 16:19 | |
-!- kardan| [n=kardan@p54BE4D9D.dip.t-dialin.net] has joined #hplusroadmap | 16:19 | |
-!- draz|lab is now known as draz|exhausted | 16:24 | |
-!- El_Matarife [n=El_Matar@38.96.203.34] has joined #hplusroadmap | 16:25 | |
CIA-43 | skdb: * rfb2416ce4c20 /core/yamlcrap.py: do default yaml dump if no yaml_repr | 17:06 |
CIA-43 | skdb: * ra879bd1d03b6 /core/ (skdb.py yamlcrap.py): Merge branch 'master' of ssh://adl.serveftp.org/var/www/skdb | 17:06 |
CIA-43 | skdb: * radf6c574265a /core/yamlcrap.py: populate attributes as if it were a typical yaml.load call | 17:06 |
fenn | hm | 17:06 |
-!- genehacker [n=noko@pool-173-57-41-223.dllstx.fios.verizon.net] has quit [Read error: 113 (No route to host)] | 17:17 | |
fenn | weird | 17:30 |
fenn | class Foo(skdb.FennObject): value=123 | 17:31 |
fenn | yaml.dump(Foo) | 17:31 |
fenn | TypeError: can't pickle int objects | 17:31 |
kanzure_ | why are you dumping a class | 17:32 |
kanzure_ | >> Either I will build my tools, or my tools will build my | 17:33 |
kanzure_ | >> tools for me. | 17:33 |
kanzure_ | > | 17:33 |
kanzure_ | > Good luck. Ping me when you have your self-replicating nanofactory. Better: | 17:33 |
kanzure_ | > send me one. | 17:33 |
kanzure_ | Maybe I'll mail him some semen. | 17:33 |
-!- roy [n=roy@ip68-108-88-202.lv.lv.cox.net] has joined #hplusroadmap | 17:36 | |
-!- roy is now known as flazmot_ | 17:36 | |
-!- flazmot_ is now known as flazmot__ | 17:36 | |
-!- wrldpc2 [n=benny@pool-173-48-253-182.bstnma.fios.verizon.net] has quit [] | 17:40 | |
kanzure_ | fenn: did you fix it | 17:55 |
kanzure_ | your TypeError | 17:55 |
fenn | no | 17:56 |
-!- ybit [n=ybit@unaffiliated/ybit] has quit [Read error: 113 (No route to host)] | 17:56 | |
draz|exhausted | did you fix it kanzure_ | 17:57 |
draz|exhausted | and by it I mean pyscholar | 17:57 |
kanzure_ | fenn: did you try making an instance of the class and doing a dump on that? | 17:57 |
-!- You're now known as kanzure | 17:58 | |
-!- flazmot [n=roy@ip68-108-88-202.lv.lv.cox.net] has quit ["Lost terminal"] | 17:58 | |
-!- El_Matarife [n=El_Matar@38.96.203.34] has quit [] | 18:00 | |
fenn | yes, works fine | 18:02 |
fenn | also class Foo(): pass; dump(Foo) works | 18:02 |
fenn | erm class Foo: v=123; dump(Foo) | 18:03 |
-!- indigo [n=indigo@theross.com] has joined #hplusroadmap | 18:08 | |
kanzure | hello indigo | 18:08 |
CIA-43 | skdb: fenn * r246e04a72cbd /core/yamlcrap.py: remove unnecessary yaml loading code | 18:09 |
CIA-43 | skdb: fenn * rb3d1efdebbe0 /packages/ (lego/metadata.yaml screw/metadata.yaml): add tag for packages | 18:09 |
CIA-43 | skdb: fenn * rec891b6b03e2 /packages/screw/screw.py: pymates won't load without occ installed | 18:09 |
indigo | hi. | 18:09 |
-!- Ian_Daniher [n=it@69.61.230.246] has joined #hplusroadmap | 18:10 | |
kanzure | hello Ian_Daniher | 18:12 |
kanzure | indigo: we would very much like you to become a package maintainer | 18:14 |
kanzure | maybe you could look at skdb/packages/lego/metadata.yaml for the time being | 18:14 |
indigo | well, let's back up | 18:15 |
fenn | what package would he be maintaining? | 18:15 |
indigo | i'm still trying to figure out exactly what you are trying to accomplish | 18:15 |
kanzure | fenn: I found indigo in #reprap | 18:15 |
kanzure | and told him about the idea of apt-get for hardware projects | 18:15 |
fenn | a useful data format to store open source hardware | 18:15 |
kanzure | but also some algorithms for generating assembly instructions and such | 18:16 |
fenn | yeah but that stuff's hard so don't get your hopes up | 18:16 |
kanzure | instructions are not trajectories | 18:16 |
kanzure | anywho. instructables but better :) | 18:17 |
fenn | sorry you meant "human readable instructions" not "robot instructions" | 18:17 |
kanzure | yes | 18:17 |
-!- ybit [n=ybit@unaffiliated/ybit] has joined #hplusroadmap | 18:17 | |
kanzure | hello ybit | 18:18 |
ybit | hiya | 18:18 |
ybit | had to shut down computer, grandfather is coming to live with me. it's the reason we had to go to the hopital the other day | 18:18 |
* ybit turns 24 tomorrow, what to do, what to do... | 18:19 | |
kanzure | gah if drazak wasn't in here I'd still be the youngest | 18:19 |
* ybit was planning on again scuba diving, but unfortunately, that won't happen due to gramps needing some care | 18:19 | |
kanzure | er not that I expected to get older and the rest of you stay the same or anything | 18:19 |
draz|exhausted | lol | 18:20 |
ybit | draz|exhausted: is what 17? | 18:20 |
draz|exhausted | 16 until next month | 18:20 |
kanzure | 11 | 18:20 |
draz|exhausted | yet I know the most biology | 18:20 |
draz|exhausted | sad | 18:20 |
kanzure | uh, what? | 18:20 |
draz|exhausted | :P | 18:20 |
-!- ybit is now known as ybit|oldFart | 18:21 | |
fenn | indigo: so does skdb sound interesting to you? | 18:21 |
* ybit|oldFart is likely to croke over any second now | 18:22 | |
draz|exhausted | rofl ybit|oldFart | 18:22 |
-!- ybit|oldFart is now known as ybit | 18:22 | |
ybit | indigo: if not, you're lame. just wanted to let you know | 18:24 |
fenn | no pressure or anything :P | 18:24 |
* kanzure picks up a particularly jagged stone | 18:25 | |
ybit | :P | 18:25 |
indigo | well, the concept sounds interesting, but i don't know anything about the details | 18:25 |
draz|exhausted | nobody does | 18:25 |
fenn | pff | 18:26 |
kanzure | draz|exhausted: huh? | 18:26 |
kanzure | where have you been drazak? | 18:26 |
draz|exhausted | :P | 18:26 |
draz|exhausted | I kid | 18:26 |
kanzure | git + django (an unnecessary frontend for viewing) + python + OpenCASCADE/pythonOCC + "packages" of hardware | 18:26 |
indigo | when you say "hardware" what do you mean? | 18:26 |
fenn | machine tools, circuit boards, chemicals, dna sequences... | 18:27 |
kanzure | space stations, wearables, lasers, | 18:28 |
draz|exhausted | physical crap | 18:28 |
kanzure | crap? | 18:28 |
indigo | sounds very broad | 18:28 |
draz|exhausted | indigo: think openwetware | 18:28 |
draz|exhausted | indigo: but with a package system | 18:28 |
kanzure | draz|exhausted: openwetware is a wiki for academics | 18:28 |
kanzure | this is not openwetware.org at all | 18:28 |
draz|exhausted | yeah, with protocals | 18:28 |
kanzure | sigh | 18:28 |
kanzure | draz|exhausted: shut the hell up. | 18:28 |
fenn | unfortunately(?) there are no examples to point at and say "this is what we're doing" because nobody's ever done it before | 18:29 |
fenn | debian comes pretty close | 18:29 |
kanzure | so people have to actually "listen" | 18:29 |
kanzure | and people aren't good at listening | 18:29 |
kanzure | unless they know "apt-get" and friends already :) | 18:29 |
kanzure | "early adopters for a technology that doesn't exist and that we have to create" | 18:30 |
kanzure | btw, openwetware.org already has protocols on their wiki, but they are in plaintext and impossible to computationally parse without spending many thousands of man-hours manually editing their cruft. | 18:31 |
fenn | indigo: a package maintainer would know the specs on their hardware and put it into a format like this: http://adl.serveftp.org/git/gitweb.cgi?p=skdb.git;a=tree;f=packages/screw | 18:32 |
fenn | other people can make modifications and post them on their own repository or some public-writable repo; it's the maintainer's job to make sure that the changes make sense and play nice with the rest of the system | 18:32 |
kanzure | er some of the changes can be self-monitored (like if a measurement is in the wrong units) | 18:33 |
indigo | sounds vauge | 18:33 |
indigo | how is this better than putting the design files in a rcs? | 18:34 |
kanzure | this complements that | 18:34 |
fenn | it is an rcs | 18:34 |
indigo | but i already have an rcs | 18:34 |
fenn | you can merge (or not, i dont care.) the important part is that you provide metadata | 18:35 |
kanzure | we shouldn't have to reinvent the wheel every time we come across CAD information and hardware projects | 18:35 |
kanzure | this is a way to "package" these projects up in a standard way | 18:35 |
indigo | do we reinvent the wheel? | 18:36 |
kanzure | how long have you spent trying to find CAD models before heh | 18:36 |
fenn | every time you find a circuit schematic in pdf format you're reinventing the wheel | 18:36 |
kanzure | posting a fuzzy jpeg of a part is not sufficient. it's 2009 ffs | 18:37 |
indigo | well the solution to that is simple. post the design files. | 18:37 |
fenn | aha, but how do you know what format it is, and how do you know what is in the file? | 18:37 |
fenn | and even then you're often left with just some oddly shaped blob and no idea how to use it | 18:37 |
indigo | well, usually there is a web page | 18:38 |
fenn | one might think so | 18:38 |
fenn | but he'd be wrong most of the time (where most is defined by the number of design files posted on the intarwebs) | 18:38 |
indigo | i think there is a reason there is not a common metadata schema for all hardware | 18:38 |
fenn | me too | 18:38 |
indigo | it's the same reason computers can't understand natural language | 18:38 |
fenn | probably not the same as your reason though | 18:38 |
kanzure | this is engineering, not natural language stuff | 18:39 |
fenn | rdf has been around for decades now | 18:39 |
indigo | can you define a schema for *all* engineering? | 18:39 |
indigo | because that's what it sounds like you are doing | 18:39 |
fenn | no, because by the time we're done technology will have progressed | 18:39 |
fenn | that's why i'm not doing "schemas" | 18:39 |
indigo | yeah, there's that | 18:39 |
indigo | well if you aren't doing schemas, what are you doing? | 18:40 |
indigo | without a schema, how can you do better than google? | 18:40 |
fenn | self defining code | 18:40 |
fenn | see, a schema is this static document that breaks if it doesn't get exactly what it wants | 18:40 |
fenn | but if your schema can be modified, people can add new representations for their problem domain | 18:41 |
indigo | yes, but then you have 2000 fields that mean the same thing | 18:42 |
fenn | that's why we need package maintainers | 18:42 |
indigo | i think i could come up with some examples of places that try to do this in a more limited domain and fail | 18:42 |
indigo | try mouser.com, for example | 18:42 |
indigo | they just do electronic components, not all hardware | 18:42 |
indigo | and the search works, sorta. | 18:42 |
fenn | well i'm sorry they don't insist on clean data | 18:43 |
indigo | but actually, most of the stuff i find by thumbing through the catalog, or by finding something written by someone else | 18:43 |
indigo | the thing is you can't insist on clean data | 18:43 |
indigo | for one, you will never get it | 18:43 |
indigo | two, it's too much work to make it | 18:43 |
fenn | how does debian work then? | 18:44 |
ybit | 18:28 < kanzure> unless they know "apt-get" and friends already :) | 18:44 |
kanzure | octopart.org also sucks | 18:44 |
indigo | debian has a very limited domain, and the "metadata" is very limited | 18:44 |
ybit | i was about to say something until i read "and friends" :) | 18:44 |
fenn | actually octopart is pretty good for electronics | 18:44 |
kanzure | ybit: someday I'll actually bother to install gentoo | 18:44 |
indigo | it's "searching" is just a plain text search of the descriptions | 18:44 |
indigo | you can usually do better searching google | 18:44 |
indigo | unless you already know the program you want, of course. | 18:44 |
kanzure | yes but that's because they ignored the ggetopt standard | 18:45 |
fenn | indigo: but, it works! | 18:45 |
indigo | fenn: it works, because it doesn't do anything. | 18:45 |
fenn | don't tell me that a google search will download all the packages you need to run a program | 18:45 |
indigo | apt does just a few things: | 18:46 |
indigo | - only software | 18:46 |
indigo | - has a version field | 18:46 |
indigo | - has a description field | 18:46 |
indigo | - has dependencies | 18:46 |
kanzure | that's not the dot deb format | 18:46 |
fenn | it's not? | 18:46 |
fenn | oh, right | 18:46 |
indigo | it doesn't go in to what the package does | 18:46 |
fenn | it's the .control format | 18:46 |
indigo | just how to install it. | 18:46 |
fenn | well, it does, sorta, but they chose the crappy version of debtags back in the day | 18:47 |
indigo | if you want to say, find linux games that are about racing and are 3d, you'd do better to search google | 18:47 |
fenn | have you tried synaptic? | 18:47 |
indigo | yes, it's just a graphical frontend to apt. | 18:47 |
fenn | so you know about the category search | 18:47 |
fenn | (am i thinking of synaptic?) | 18:47 |
kanzure | yes | 18:47 |
indigo | that's not a synaptic feature | 18:47 |
indigo | it's an apt feature | 18:48 |
kanzure | indigo: so you're claiming that apt doesn't do anything? | 18:48 |
indigo | and searching by categories isn't usually usefull | 18:48 |
fenn | it's not? | 18:48 |
fenn | would you prefer NLP or something? | 18:48 |
indigo | well, i can't remember the last time i used it. | 18:48 |
kanzure | yeah I don't use it either | 18:48 |
fenn | "computer, fetch me the best video game evar!" | 18:48 |
indigo | develop NLP that works and you will be a very rich person. | 18:48 |
fenn | NLP is a dead swan | 18:49 |
kanzure | a dead black swan? | 18:49 |
indigo | the problem with categories is that someone has to define them, and the things have to fit in them. | 18:49 |
kanzure | doesn't that make it a rotting swan? | 18:49 |
fenn | a black swan that choked on a red herring | 18:49 |
indigo | rarely does it work. | 18:49 |
kanzure | that's not what we're on about though | 18:49 |
indigo | well, what are you about,t hen? | 18:49 |
fenn | categories? yes | 18:49 |
fenn | classifications actually | 18:49 |
indigo | classification is really hard. | 18:50 |
fenn | harder than NLP? | 18:50 |
indigo | i'd say it's about the same problem. | 18:50 |
fenn | no way | 18:50 |
indigo | it's easy if you think about screws, nuts, washers, hinges | 18:50 |
indigo | but, what about a bicycle? | 18:50 |
indigo | what category is that? | 18:51 |
fenn | vehicle | 18:51 |
indigo | but doesn't it also have wheels? | 18:51 |
fenn | so? | 18:51 |
indigo | well, what if i'm looking for things that roll? | 18:51 |
kanzure | actuators? | 18:51 |
fenn | um. "things that roll"? | 18:51 |
kanzure | linear motion converters? | 18:51 |
kanzure | I think that's actually expressable in some units | 18:51 |
indigo | no, i just want something that rolls. | 18:51 |
kanzure | how do you not have this problem with debian? | 18:52 |
fenn | well does a bike roll or do the wheels roll? | 18:52 |
indigo | i do have this problem with debian | 18:52 |
kanzure | dependency resolution is the main problem being solved | 18:52 |
indigo | the thing is, apt doesn't really make it any easier to find software unless you already know what you want. | 18:52 |
kanzure | that's true for the leaf nodes | 18:52 |
indigo | ok, dependencies are one thing that could be solved | 18:53 |
kanzure | once you know what leaf node you want, | 18:53 |
indigo | i will give you that | 18:53 |
kanzure | it can travel back to the roots and you have what you need then. | 18:53 |
kanzure | right | 18:53 |
fenn | we're not trying to solve the "computer please read my mind" problem | 18:54 |
kanzure | I have enough trouble reading my own | 18:54 |
indigo | well, here's something to consider | 18:56 |
indigo | apt already handles dependencies and versioning pretty well | 18:56 |
indigo | why not use apt? | 18:56 |
fenn | versioning doesn't work so well for hardware compatibility | 18:57 |
kanzure | actually I looked into repurposing apt | 18:58 |
kanzure | and other managers | 18:59 |
kanzure | because that's the obvious first step right? | 18:59 |
kanzure | it turns out that apt is grossly integrated into the beating heart of debian | 18:59 |
kanzure | and the development team (deity) has long been disbanded | 18:59 |
kanzure | and the original development work happened in '91 or something in some random-ass unlogged IRC channel | 18:59 |
kanzure | that nobody has logs to any more. | 18:59 |
indigo | fink has repurposed apt | 19:01 |
kanzure | http://finkproject.org/ ? | 19:01 |
indigo | so it must not be impossible | 19:01 |
kanzure | what I want is a generic packaging system | 19:01 |
kanzure | xpkg is close, but it's the only one I can find | 19:02 |
indigo | that's the one | 19:02 |
kanzure | smartpm is some nice algorithms, but where's the rest of it :p | 19:02 |
indigo | there's nothing in apt that is specific to software best i can tell | 19:02 |
indigo | it's just files | 19:02 |
indigo | also, many RCS support nested branches | 19:02 |
indigo | so, when you check out X, you get Y and Z as well. | 19:02 |
indigo | then there are simple soultions (my favorite) like makefiles or just plain shell scripts that get the dependencies | 19:03 |
kanzure | what was the early package manager that was a shell script that just downloaded dependencies? | 19:04 |
kanzure | tzsh? | 19:04 |
kanzure | sht? or something? | 19:04 |
fenn | apt doesn't keep track of separate dependency trees | 19:06 |
kanzure | ". For example, the Angel's told us that in their case they called their local law enforcement in Florida who sent a uniformed officer in a squad car to their home. The first thing you can imagine the officer asked was, "What's a domain?".'"" | 19:06 |
fenn | we need to keep track of both build-time and run-time hardware deps | 19:06 |
kanzure | the software isn't particularly hard to write.. | 19:06 |
kanzure | if I can only remember to actually get it done | 19:07 |
kanzure | kword has corporate backing? | 19:14 |
-!- Ian_Daniher [n=it@69.61.230.246] has quit [Read error: 110 (Connection timed out)] | 19:29 | |
kanzure | ok, spider veins are necessary to complete the vegeta look | 19:41 |
kanzure | "Cushing's syndrome (also called hyperadrenocorticism or hypercorticism) is a hormone (endocrine) disorder caused by high levels of cortisol (hypercortisolism) in the blood. This can be caused by taking glucocorticoid drugs, or by tumors that produce cortisol or adrenocorticotropic hormone (ACTH).[1] Cushing's disease refers to one specific cause, a non-cancerous tumor (adenoma) in the pituitary gland that produces large amounts of ACTH, which in turn elevates cortisol. It can usually be cured by surgery." | 19:49 |
draz|exhausted | watching house? | 19:50 |
kanzure | "genetic horoscopes" | 19:51 |
kanzure | yes but not really paying attention | 19:51 |
kanzure | fenn was wondering what cushings was | 19:51 |
-!- indigo [n=indigo@theross.com] has left #hplusroadmap [] | 19:51 | |
kanzure | holy crap | 19:57 |
kanzure | Resolving single cone inputs to visual receptive fields | 19:57 |
draz|exhausted | hmm | 20:02 |
draz|exhausted | lets see if this keyboard is any good | 20:02 |
draz|exhausted | it's complete blank | 20:02 |
draz|exhausted | and it has no markings for home keys | 20:02 |
draz|exhausted | I dunno if I can handle it | 20:02 |
kanzure | "This suggests a greater than normal dependence on cortical regions in which movements are represented in intrinsic coordinates of motion (M1 and somatosensory cortex) and a less than normal dependence on regions in which movements are represented in extrinsic coordinates (premotor and posterior parietal)." | 20:02 |
kanzure | "A stronger than normal association between motor commands and proprioceptive feedback may be a consequence of the fact that M1 and somatosensory cortex are nearby cortical regions and short-range cortical connections are overexpressed in children with ASD." | 20:03 |
draz|exhausted | lets see about this keyboard thing | 20:03 |
kanzure | http://www.slideshare.net/IgnitePhoenix/slicing-brains-very-quickly-todd-huffman | 20:22 |
draz|exhausted | hmmmm this keyboard isn't too bad | 20:24 |
kanzure | wpmage? | 20:29 |
kanzure | I seem to type most quickly on laptop keyboards for some reason | 20:30 |
kanzure | not sure why | 20:30 |
draz|exhausted | eh, I dunno | 20:30 |
draz|exhausted | I'm not quite as fast as I am on a real Model M | 20:31 |
draz|exhausted | but I'm close | 20:31 |
draz|exhausted | so probably 80some | 20:31 |
draz|exhausted | dunno, I'm typing normal speed, which is like 60 or so | 20:31 |
draz|exhausted | I did some fast yping and I think I hit 80some | 20:31 |
-!- genehacker [n=noko@pool-173-57-41-223.dllstx.fios.verizon.net] has joined #hplusroadmap | 20:32 | |
genehacker | internet down | 20:32 |
genehacker | wireless inoperable | 20:33 |
kanzure | draz|exhausted: 80some should be slow for you | 20:42 |
kanzure | 120 is where I max out at | 20:43 |
kanzure | (on qwerty) | 20:43 |
draz|exhausted | I max out around 100 | 20:43 |
kanzure | yeah okay | 20:43 |
kanzure | 100!=80 | 20:43 |
draz|exhausted | I did some fast typing on this keyboard and I hit about 80 | 20:43 |
draz|exhausted | I hit about 100 on a model m | 20:43 |
kanzure | in middle school and high school we had this windows network drive where we had these common apps | 20:44 |
kanzure | one of the apps was this typing program | 20:44 |
kanzure | over the years I tested myself weekly or something | 20:44 |
kanzure | and I got to the point where .. I knew what the coders got wrong about it | 20:44 |
draz|exhausted | lol | 20:44 |
kanzure | so I kind of knew when to type and when not to type | 20:44 |
kanzure | in order to game the fuck out of the system | 20:44 |
draz|exhausted | lol | 20:44 |
genehacker | heh reminds me of high school | 20:50 |
genehacker | we had a network drive for uploading projects | 20:50 |
genehacker | so one kid comes along and installs a game there | 20:50 |
genehacker | as a "computer class project" | 20:50 |
genehacker | disconnecting from hivemind | 20:52 |
draz|exhausted | lol | 20:52 |
kanzure | joseph jackson is on fast forward radio blog next week | 20:55 |
-!- genehacker2 [n=noko@pool-173-57-41-223.dllstx.fios.verizon.net] has joined #hplusroadmap | 20:58 | |
-!- genehacker [n=noko@pool-173-57-41-223.dllstx.fios.verizon.net] has quit [Read error: 110 (Connection timed out)] | 21:10 | |
-!- genehackerclone [n=noko@pool-173-57-41-223.dllstx.fios.verizon.net] has joined #hplusroadmap | 21:13 | |
-!- genehacker2 [n=noko@pool-173-57-41-223.dllstx.fios.verizon.net] has quit [Read error: 60 (Operation timed out)] | 21:15 | |
-!- embraceunity [n=quassel@74.94.105.238] has joined #hplusroadmap | 21:20 | |
embraceunity | bryan, loved your most recent post on open manufacturing | 21:21 |
QuantumG | what'd he say? | 21:22 |
embraceunity | "Meanwhile, this beautiful, undesigned flower of abundant anarchy | 21:22 |
embraceunity | unfolds before our eyes, springing up out of the manure of capitalism. | 21:22 |
embraceunity | What a time to be alive! | 21:22 |
kanzure | http://heybryan.org/school/English%20III/9-27-06,%20Captivity%20narrative.html | 21:22 |
kanzure | hey embraceunity | 21:22 |
embraceunity | hey | 21:22 |
kanzure | what on earth does "Mein Treffen mit dem Hallenführer " mean and why was I writing in german? | 21:23 |
QuantumG | I stopped reading after noticing that many people on there want to post 5 page essays about nothing. | 21:23 |
genehackerclone | kanzure you must be sleep typing again | 21:23 |
genehackerclone | maybe you're sleep-learning german | 21:23 |
kanzure | QuantumG: that's just paul | 21:23 |
kanzure | nei heiler fullen schazzen! | 21:24 |
genehackerclone | ??? | 21:24 |
genehackerclone | INCOMPATIBLE DATA FORMAT | 21:24 |
embraceunity | time to try out rc5... brb | 21:31 |
-!- embraceunity [n=quassel@74.94.105.238] has quit [Remote closed the connection] | 21:31 | |
genehackerclone | just read about senator proxmire | 21:34 |
genehackerclone | damn | 21:34 |
genehackerclone | we could have space colonies already if it wasn't him | 21:35 |
genehackerclone | *wasn't for him | 21:35 |
fenn | i always wondered what happened | 21:36 |
kanzure | the paper about backtracing individual photocones in the retina wasn't as cool as it could have been | 21:40 |
genehackerclone | some douche on national television said space colonies are stupid is what happened | 21:40 |
genehackerclone | btw, a 40 meter blast wave accelerator could shoot stuff into orbit | 21:41 |
genehackerclone | if that stuff can withstand the 10,000 g | 21:41 |
genehackerclone | force | 21:41 |
kanzure | "was it just that in the past people who were too practically disoriented to live sort of disappeared?" | 21:44 |
genehackerclone | ??? | 21:45 |
kanzure | "proprioception is the sense that indicates whether the body is moving with required effort" | 21:48 |
fenn | with required effort? | 21:48 |
fenn | WORK HARDER! | 21:49 |
genehackerclone | uh | 21:49 |
genehackerclone | it's also the sense of where your limbs are | 21:49 |
genehackerclone | we have muscles just to figure out where are arms are | 21:49 |
fenn | we dont have a sense of where limbs are | 21:50 |
fenn | no optical encoders, ya know | 21:50 |
genehackerclone | nope | 21:50 |
genehackerclone | we got a kludge | 21:50 |
kanzure | http://en.wikipedia.org/wiki/Proprioception | 21:51 |
kanzure | M1 = primary motor cortex | 21:55 |
kanzure | stereotactical coordinate systems? | 21:57 |
genehackerclone | ??? | 21:58 |
genehackerclone | you mean a big matrix in our heads with positions of limbs | 21:58 |
genehackerclone | cool | 21:58 |
kanzure | oh cool there's a name for it: | 21:58 |
kanzure | stereotactic frame | 21:58 |
kanzure | http://www.parkinsons-information-exchange-network-online.com/archive/fig1.jpg | 21:58 |
kanzure | the rat brain in stereotaxic coordinates: http://www.amazon.com/Rat-Brain-Stereotaxic-Coordinates-Coronal/dp/0120884720 | 22:00 |
kanzure | "brain cartographers" heh | 22:00 |
fenn | *cough*bullshit*cough* | 22:01 |
fenn | eh what? that's brain coordinates not limb coordinates | 22:02 |
kanzure | yes sorry I jumped topics on you | 22:02 |
fenn | i was deceived by your clone | 22:02 |
kanzure | http://www.loni.ucla.edu/twiki/bin/view/MouseBIRN/MouseBIRNTools various tools used for parsing 3D brain data for mice | 22:08 |
kanzure | "This 3D Mouse Atlas (v2.0) is a high-resolution,17.9 micron, isotropic 3D dataset. It is composed of reconstructed Nissl stained sectional material from a freshly frozen brain of an adult male C57BL/6J strain mouse. " | 22:08 |
genehackerclone | was that brain frozen cryogenically? | 22:09 |
QuantumG | 17.9 micron | 22:10 |
QuantumG | .. what can ya see with that... grey matter vs white matter? | 22:11 |
QuantumG | you're certainly not going to be mapping any circuits | 22:11 |
genehackerclone | neurons are big | 22:12 |
genehackerclone | 100 microns wide | 22:12 |
* kanzure does a full rip of http://www.nbirn.net/bdr/data/ | 22:12 | |
QuantumG | are dendrites? | 22:13 |
genehackerclone | don't know | 22:13 |
QuantumG | On the average, dendrites measured 15-20 microns in length and 0.1-0.3 micron in diameter. | 22:16 |
genehackerclone | blast | 22:16 |
genehackerclone | so pretty much useless | 22:16 |
QuantumG | even if you had fantastic resolution, it'd still be a guessing game.. "first we need to figure out how to cold start this brain" | 22:17 |
genehackerclone | reminds me of neural network initializing beam robots | 22:18 |
genehackerclone | though you could try just firing neurons randomely | 22:18 |
QuantumG | and that's if you believe the "connectionist" viewpoint taken to a caricature extreme. | 22:19 |
QuantumG | even the most hardcore connectionist acknowledges the importance of levels of biochemicals | 22:20 |
kanzure | QuantumG: nobody ever claimed that having a map of dendrites would be enough to "cold boot" a brain | 22:20 |
genehackerclone | yup | 22:20 |
QuantumG | ya, agreed | 22:21 |
QuantumG | it'd be nice to study | 22:21 |
kanzure | so where was that coming from | 22:21 |
kanzure | sounds strawman though | 22:21 |
genehackerclone | though it's really better if you know how the neurons act then wire them and fire em up | 22:21 |
fenn | 'cold boot' is ridiculous | 22:21 |
fenn | just throw some noise into the system | 22:21 |
QuantumG | nah, not making an argument, just saying there's a lot more data to collect | 22:22 |
fenn | "how to start a resonator resonating" | 22:22 |
genehackerclone | woo almost completed my wearable heat exchanger | 22:23 |
kanzure | ok a backup is on adl:/home/bryan/cache/birn/ | 22:25 |
kanzure | 2.3GB of neuro imagery | 22:25 |
genehackerclone | of mouse brain or you before the accident? | 22:25 |
zvader | i wasn't aware of a difference | 22:26 |
genehackerclone | zvader are you at UT? | 22:29 |
kanzure | do /whois zvader | 22:30 |
genehackerclone | he is at UT | 22:30 |
genehackerclone | @ heckle | 22:30 |
kanzure | ellington's machine? | 22:30 |
genehackerclone | perhaps he works with aptamers | 22:30 |
genehackerclone | yup | 22:30 |
genehackerclone | I do believe so | 22:31 |
genehackerclone | btw how did the meeting with cline go? | 22:32 |
kanzure | http://imaging.mrc-cbu.cam.ac.uk/imaging/MniTalairach MNI v. Talairach coordinate systems for the brain | 22:32 |
kanzure | "The current standard MNI template is the ICBM152, which is the average of 152 normal MRI scans that have been matched to the MNI305 using a 9 parameter affine transform." | 22:33 |
kanzure | International Consortium for Brain Mapping http://www.loni.ucla.edu/ICBM | 22:33 |
kanzure | hm you can download the warp5 atlas: http://www.loni.ucla.edu/ICBM/Downloads/Downloads_452T1.shtml | 22:34 |
genehackerclone | woohoo my heat exchanger is finished enough to test | 22:44 |
kanzure | http://en.wikipedia.org/wiki/Neuronavigation | 22:45 |
fenn | pics or GTFO | 22:49 |
kanzure | there's Talairach, MNI, stereotaxic | 22:50 |
kanzure | http://images.google.com/images?hl=en&client=iceweasel-a&rls=org.debian%3Aen-US%3Aunofficial&um=1&sa=1&q=Talairach+brain&aq=f&oq=&aqi=&start=0 | 22:50 |
fenn | um. i meant the heat exchanger, sorry. | 22:50 |
kanzure | ooh | 22:50 |
kanzure | http://www.journalofvision.org/3/10/1/table02.gif | 22:50 |
kanzure | er, no, those "subjects" must be initials | 22:50 |
kanzure | not regions :( | 22:50 |
kanzure | (from the paper Visual field representations and locations of visual areas V1/2/3 in human visual cortex) | 22:51 |
kanzure | BrainLinks3D looks like what I want: http://www.cs.sunysb.edu/~mueller/research/brainMiner/ | 22:52 |
kanzure | "brainminer - to automatically discover brain functions in different regions" | 22:52 |
-!- El_Matarife [n=El_Matar@adsl-68-88-200-16.dsl.rcsntx.swbell.net] has joined #hplusroadmap | 22:52 | |
kanzure | http://images.google.com/images?hl=en&client=iceweasel-a&rls=org.debian:en-US:unofficial&um=1&q=brain+regions+coordinates&sa=N&start=0&ndsp=18 | 23:00 |
kanzure | is google images working for anyone? | 23:00 |
kanzure | "next page" doesn't seem to work any more | 23:00 |
fenn | yar | 23:01 |
fenn | after watching family guy, when reading emails everyone's voice is either stewie or bryan | 23:08 |
fenn | *sob* | 23:08 |
kanzure | I need a better way to keep track of brain regions that I learn about or whatever. | 23:13 |
kanzure | I've been meaning to do something like this for a very long time | 23:13 |
kanzure | but the problem is that nobody uses 'generic' coordinates or something | 23:13 |
kanzure | it's always some sort of image averaging algorithm from MRI data or CT data | 23:13 |
kanzure | now, I don't really need coordinates per-se to reference big blocks of brains | 23:14 |
kanzure | but it would be useful to give a general location with respect to other regions of the brain, or other connections in the brain (such as the connectome projects) | 23:14 |
-!- embraceunity [n=quassel@74.94.105.238] has joined #hplusroadmap | 23:16 | |
kanzure | hey embraceunity | 23:16 |
embraceunity | hello kanzure | 23:17 |
fenn | back for more eh | 23:17 |
embraceunity | i can't help it, im an addict | 23:17 |
fenn | "flower of abundant anarchy" was emlyn oregan | 23:17 |
embraceunity | ah | 23:18 |
embraceunity | well it sounded beautiful | 23:18 |
embraceunity | in any event | 23:18 |
kanzure | maybe if I try hard enough I can make it seem like I've done some work for tomorrow's meeting | 23:19 |
fenn | somehow i doubt campbell will be impressed that packages load | 23:19 |
kanzure | most of the past week has been me making commits, fenn undoing my commits, fenn rewriting my commits, and then me waiting for fenn to stop so that I can get back to using it | 23:20 |
fenn | the past week being... today. | 23:20 |
kanzure | nah there was some other code that this happened with | 23:20 |
kanzure | last saturday | 23:20 |
kanzure | er, last-last saturday | 23:20 |
fenn | the lego interface thing? | 23:20 |
* fenn checks the logs | 23:21 | |
kanzure | maybe, but also all sorts of file moving and class moving and general rewriting of the architecture | 23:21 |
kanzure | some of which was necessary admittedly | 23:21 |
kanzure | I think campbell will be sufficiently impressed with an interactive python session | 23:22 |
kanzure | where you say "lego = skdb.load('lego')" | 23:22 |
fenn | last last saturday was alcor and then receipts data entry | 23:22 |
fenn | hm really? | 23:23 |
kanzure | maybe it was the sunday then | 23:23 |
kanzure | no not really | 23:23 |
fenn | wah. | 23:23 |
kanzure | trying to raise my hopes | 23:23 |
genehackerclone | damn didn't go to alcor | 23:23 |
genehackerclone | what'd I miss? | 23:23 |
kanzure | nothing | 23:23 |
genehackerclone | oh | 23:23 |
kanzure | they didn't do the training | 23:24 |
genehackerclone | btw how much does cryopreservation cost? | 23:24 |
kanzure | for a student? it's cheap | 23:24 |
genehackerclone | define cheap? | 23:24 |
kanzure | flat rate $30k for neuro/head preservation, but for a student it's significantly less | 23:24 |
fenn | there's some way to re-route life insurance towards cryopreservation | 23:24 |
kanzure | and it's even cheaper with medical coverage | 23:24 |
fenn | supposedly it's like $15/mo | 23:24 |
genehackerclone | hmmmm... | 23:24 |
genehackerclone | I could afford that | 23:24 |
kanzure | apparently Cryonics Institute (in Europe) is ran entirely by two men | 23:25 |
kanzure | these are the two most awesome men in the world right now | 23:25 |
kanzure | because technically their establishment is considered a cemetary | 23:25 |
genehackerclone | I don't trust cryonics institute | 23:25 |
fenn | i thought they were in michigan | 23:25 |
genehackerclone | because they have 2 men | 23:25 |
embraceunity | cryonics isn't that important | 23:25 |
embraceunity | in my book | 23:25 |
embraceunity | but i support your efforts | 23:25 |
kanzure | it's worth playing around with | 23:25 |
genehackerclone | what book? | 23:25 |
fenn | it's sort of like reprap and sexual reassignment surgery | 23:26 |
embraceunity | hahaha | 23:26 |
fenn | "neat idea, but does it really work?" | 23:26 |
genehackerclone | yeah | 23:26 |
embraceunity | damnit | 23:26 |
embraceunity | my reprap is just a hobby | 23:26 |
embraceunity | i knew it | 23:26 |
kanzure | do not view this page: | 23:27 |
kanzure | http://ai.eecs.umich.edu/people/conway/TS/SRSlink.html | 23:27 |
* fenn attempts to click the link | 23:27 | |
embraceunity | fenn: now j00 got virusez... hahaha | 23:27 |
genehackerclone | you have a reprap? | 23:27 |
genehackerclone | does it work? | 23:27 |
embraceunity | no | 23:28 |
genehackerclone | blast | 23:28 |
embraceunity | but im building it | 23:28 |
embraceunity | the hackerspace by me | 23:28 |
embraceunity | has 1 working one and one almost working one | 23:28 |
embraceunity | they promised to lend me a hand | 23:28 |
genehackerclone | I need to test some printable bearings so I can make a printable stepper motor powered by printable fluidic circuitry | 23:28 |
embraceunity | that.... is .... awesome | 23:29 |
fenn | first you need to acquire a printer | 23:29 |
genehackerclone | the stratasys in the ME building when I get back to school | 23:29 |
kanzure | embraceunity: did you ever read the microfluidics paper archive | 23:29 |
fenn | all 5000 pages of it | 23:29 |
kanzure | or at least parts of it | 23:29 |
embraceunity | kanzure: joseph told me about it.... link me | 23:29 |
genehackerclone | kanzure can I transfer some files to you via IRC | 23:29 |
genehackerclone | I got some good stuff about microfluidics | 23:30 |
kanzure | embraceunity: http://adl.serveftp.org/papers/microfluidics/ | 23:30 |
kanzure | genehackerclone: just upload them to the server | 23:30 |
genehackerclone | how do I do that? | 23:30 |
kanzure | take your pick: FTP, SCP, sneakernet | 23:30 |
genehackerclone | the datafile covering how to ftp stuff in my head has been corrupted | 23:30 |
genehackerclone | ftp | 23:30 |
embraceunity | by the way | 23:31 |
embraceunity | i just bought 10 stepper motors for 2 bucks a pop | 23:31 |
kanzure | http://www.velnetsupport.co.uk/parrots/FTP/Filezilla/ | 23:31 |
fenn | genehackerclone: what OS are you on? | 23:31 |
kanzure | filezilla is an ftp extension for firefox | 23:31 |
-!- El_Matarife [n=El_Matar@adsl-68-88-200-16.dsl.rcsntx.swbell.net] has quit ["Nettalk6 - www.ntalk.de"] | 23:31 | |
kanzure | guess you can use that.. | 23:31 |
kanzure | but if you're on ubuntu, there's a better way | 23:31 |
genehackerclone | ubuntu | 23:31 |
fenn | you have anonymous ftp upload enabled? | 23:31 |
genehackerclone | oops winblows right now | 23:31 |
fenn | er. you do realize that ftp sends passwords in the clear don't you? | 23:31 |
genehackerclone | oh wait | 23:31 |
genehackerclone | let me fix that | 23:32 |
kanzure | ? | 23:32 |
genehackerclone | time to get my dad's huge cryptoket | 23:32 |
genehackerclone | it's very hard to crack | 23:32 |
kanzure | I'll probably be asleep | 23:32 |
fenn | just use filebin.a | 23:32 |
fenn | filebin.ca | 23:32 |
genehackerclone | is it secure | 23:33 |
genehackerclone | ok good enough | 23:33 |
genehackerclone | just a bunch of pictures | 23:33 |
* fenn expects weird dreams tonight | 23:34 | |
fenn | i'll tweet about the in the morning!~ | 23:34 |
embraceunity | by the way guys, i just visited factor e farm today... marcin kicked out Jeremy and Ben... | 23:35 |
embraceunity | if anyone is interested | 23:35 |
embraceunity | i bet they are looking for new digs | 23:35 |
fenn | who are jeremy and ben and what's their big deal? | 23:36 |
embraceunity | jeremy is a total diyh+ er | 23:36 |
embraceunity | ben is more into agriculture | 23:36 |
fenn | and how the hell do you get kicked out of a mud hut in the middle of nowhere? | 23:36 |
embraceunity | haha | 23:36 |
genehackerclone | it's too big for filebin | 23:36 |
embraceunity | well Marcin has kicked out a lot of people over time | 23:36 |
genehackerclone | 57 mb | 23:36 |
embraceunity | i love him, but I think he is a demanding guy | 23:36 |
fenn | really.. i thought he was building an army or something | 23:37 |
genehackerclone | huh? | 23:37 |
embraceunity | he probably is building an army | 23:37 |
embraceunity | but it currently consists of him and a 50 something year old austrian woman named Inga | 23:37 |
fenn | i suppose i should go lay on my pallet and pretend to sleep | 23:38 |
genehackerclone | upload commencing | 23:40 |
genehackerclone | http://filebin.ca/ordqrh | 23:41 |
genehackerclone | get it while it's hot | 23:41 |
genehackerclone | upload 2 commencing | 23:42 |
genehackerclone | http://filebin.ca/ayxuh | 23:42 |
genehackerclone | done | 23:42 |
genehackerclone | any downloaders? | 23:44 |
genehackerclone | it has info on microfluidic chemical factories... | 23:45 |
embraceunity | ok i'll put it on bittorrent | 23:51 |
embraceunity | haha | 23:51 |
genehackerclone | just don't link it to me | 23:53 |
genehackerclone | you can get all the wuffie you want | 23:54 |
embraceunity | there is some network like Tor for bittorrent | 23:56 |
embraceunity | i can't remember the name | 23:56 |
embraceunity | I2P that's it | 23:57 |
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!