| --- Day changed Wed Apr 09 2008 | ||
| -!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has quit ["Leaving."] | 00:56 | |
| epitron | http://funtarded.com/pics/show/1660 | 05:01 |
|---|---|---|
| -!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has joined #hplusroadmap | 07:34 | |
| kanzure | http://pdbwiki.org | 07:37 |
| -!- Enki-2 [n=weechat@c-71-234-190-248.hsd1.ct.comcast.net] has quit [Read error: 104 (Connection reset by peer)] | 08:02 | |
| -!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has quit ["Leaving."] | 08:07 | |
| -!- Enki-2 [n=weechat@c-71-234-190-248.hsd1.ct.comcast.net] has joined #hplusroadmap | 08:08 | |
| -!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has joined #hplusroadmap | 17:57 | |
| kanzure | fenn: I was figuring today that all we really need to do skdb is just standardized fileformats and IO to programs/packages in a specific way. http://heybryan.org/mediawiki/index.php/Skdb has the specific notes from today at the bottom. | 17:59 |
| kanzure | epitron: my bito.ponzo.net login is not working | 18:04 |
| kanzure | looks like I have hosting at http://biowebhost.kr/ now | 18:07 |
| epitron | kanzure: did it work before? | 18:08 |
| kanzure | no | 18:09 |
| kanzure | didn't check :) | 18:09 |
| kanzure | but today's my data dumping day | 18:09 |
| epitron | so it's still got that "changeme" password | 18:09 |
| epitron | whatever it was | 18:09 |
| kanzure | yes | 18:10 |
| kanzure | I know it too | 18:10 |
| epitron | and it didn't work? | 18:10 |
| kanzure | wait | 18:10 |
| kanzure | do you have an ftpd running? | 18:10 |
| epitron | yep | 18:10 |
| kanzure | then it should have worked | 18:10 |
| epitron | but it's not using the same user database :) | 18:10 |
| kanzure | aha | 18:10 |
| kanzure | okay | 18:10 |
| epitron | you gotta use ssh/scp/sftp | 18:10 |
| kanzure | I'll use SCP | 18:10 |
| * fenn looks | 18:12 | |
| epitron | kanzure: it would be nice if you changed the changeme password too.. i don't remember it being that secure :) | 18:13 |
| kanzure | it's not | 18:13 |
| kanzure | I'll change it | 18:13 |
| kanzure | hehe, bio.cc has me hosted now | 18:13 |
| kanzure | This is the same guy that has all of the bio*.org domain names | 18:13 |
| kanzure | He set up something like 10,000 wikis on 10,000 domains ... | 18:14 |
| epitron | ok, that'd retarded | 18:14 |
| kanzure | sort of | 18:14 |
| kanzure | http://biodatabase.org/ | 18:14 |
| kanzure | the content is good :) | 18:14 |
| epitron | way to fragment the universe | 18:14 |
| kanzure | haha | 18:14 |
| kanzure | yep | 18:14 |
| fenn | agree w/epitron | 18:14 |
| kanzure | fenn: check my file format notes above | 18:15 |
| fenn | i'm reading, and not understanding | 18:15 |
| kanzure | Have you ever hard coded usage requirements into your program? | 18:15 |
| fenn | could you give an example of a program we'd interface with? | 18:15 |
| kanzure | a chair simulator | 18:15 |
| fenn | the 'interface' would be some test layer right? | 18:16 |
| kanzure | # chairSimulator -3dinputmesh my.mesh | 18:16 |
| fenn | and unit evaluation hooks | 18:16 |
| kanzure | yeah, I mean passing parameters and so on | 18:16 |
| kanzure | yep | 18:16 |
| fenn | so uh, why does it need to download anything? | 18:16 |
| kanzure | well, you don't have all of the programs locally to interpret the files do you ? | 18:16 |
| fenn | cant we just use apt? | 18:16 |
| kanzure | and grabbing all of them at once is bulky, unless you have the space, in which case go right ahead | 18:16 |
| kanzure | huh? | 18:16 |
| fenn | to manage software | 18:16 |
| kanzure | right right | 18:17 |
| kanzure | but think about what other topics we've been talking about | 18:17 |
| kanzure | there are specific file formats that software work with | 18:17 |
| kanzure | this is where we specify special features like units and so on | 18:17 |
| fenn | that will be part of the package source | 18:17 |
| kanzure | correct | 18:17 |
| kanzure | but not only that | 18:17 |
| kanzure | but we should explicitly specify what meta-data it is expecting | 18:18 |
| kanzure | meta-data like "yes, I can use a 3ds file" | 18:18 |
| kanzure | all of the command line parameters are hardcoded | 18:18 |
| kanzure | and then written up in man pages | 18:18 |
| kanzure | this is not good ;) | 18:18 |
| fenn | we should at least try to settle on a small number of formats | 18:18 |
| kanzure | (I'm specifically talking about IO-related parameters, there are some flags that you can throw to gcc that will make it do other behaviors, this is different) | 18:19 |
| kanzure | perhaps | 18:19 |
| fenn | too bad all 3d file formats suck balls | 18:19 |
| kanzure | I'm all for integrating formats, and so we'll probably have some programs to help do that | 18:19 |
| kanzure | (converters) | 18:19 |
| fenn | IGES is okay-ish | 18:19 |
| fenn | and widely used in CAD | 18:20 |
| kanzure | I've played with tons of formats - STL, msh, VRML, VRML2, blender, 3ds, obj (yes, ambiguous), max, geom, and tons of others | 18:20 |
| fenn | ok, so the loungechair package has a 3d model of a lounge chair | 18:20 |
| kanzure | yes | 18:21 |
| fenn | it provides: what? | 18:21 |
| kanzure | and so a user should be able to download that 3d model via apt/autogenix | 18:21 |
| kanzure | or really they could just say apt-get all information related to loungechair-pack | 18:21 |
| fenn | er, maybe provides is not the right place, since that should be functionality (and a 3d model isnt a functionality of a chair) | 18:21 |
| kanzure | the 3d model provides information for simulation, I'd expect | 18:21 |
| kanzure | right | 18:21 |
| kanzure | it's more like a functional unit for a simulation | 18:21 |
| kanzure | while the loungechair package should be ... instructions to build the chair | 18:21 |
| fenn | the package should have everything relating to that specific lounge chair | 18:22 |
| kanzure | such as lathing, chopping wood, whatever (obviously chopping wood should be another package of information for programming) | 18:22 |
| kanzure | right | 18:22 |
| fenn | CAM files, maybe | 18:22 |
| kanzure | ahuh | 18:22 |
| kanzure | I don't want to care too much about extensions | 18:22 |
| kanzure | extensions are useful, but you lose lots of information | 18:22 |
| kanzure | metadata on the file is pretty darn useful | 18:23 |
| kanzure | just like the DTD statements on (valid) XHTML pages | 18:23 |
| fenn | CAM is like machine language instructions for CNC'ing out the pieces | 18:23 |
| kanzure | of course, it's all the same for XHTML, but nobody seems to care? | 18:23 |
| kanzure | ah, neat | 18:23 |
| fenn | more precisely, CAM is the program that generates the machine language | 18:23 |
| fenn | from a cad file | 18:23 |
| kanzure | epitron: it works, thanks | 18:24 |
| fenn | explaining human by-hand operations is going to be hard | 18:24 |
| kanzure | http://bito.ponzo.net/~kanzure does not work, though | 18:24 |
| kanzure | fenn: yes; I suppose this will be accounted for if we ever can train neural nets to do hand movements with constructed hands or something? | 18:25 |
| kanzure | that's lots of complexity that frankly I think will come about on its own | 18:25 |
| epitron | kanzure: static.heybryan.org | 18:25 |
| kanzure | the point is that these specs can support such a development if that comes about | 18:25 |
| kanzure | epitron: ah, right | 18:25 |
| kanzure | I wonder if I switched DNS hosts or not | 18:25 |
| epitron | you did not :) | 18:25 |
| epitron | you switched and switched back 5 minutes later | 18:26 |
| kanzure | ns1.everydns.net ? | 18:26 |
| epitron | hold on | 18:26 |
| epitron | i'm making sure it works | 18:26 |
| epitron | nope | 18:26 |
| epitron | looks like it's down again | 18:26 |
| epitron | you have some kind of power | 18:26 |
| kanzure | hm | 18:26 |
| kanzure | you mean, everydns is down? | 18:26 |
| epitron | it's not down | 18:27 |
| epitron | but it's not letting me use it as a resolver | 18:27 |
| fenn | presumably there already exists some type of parts-assembly language that we can take advantage of (it's probably proprietary or standards-encrusted) | 18:27 |
| epitron | my other everydns hosted domains are working | 18:27 |
| kanzure | fenn: the package specification file format gets to update, and when apt/autogenix grabs that file from the wiki and realizes it doesn't know how to interpret version 5.x or something, | 18:27 |
| kanzure | it can then go fetch a script that can do it; on top of that, this same structure works when trying to crawl through the database to put parts together | 18:27 |
| kanzure | fenn: yeah. I was figuring that there *must* be a 'parameter wrapping library' of the type that I am talking about | 18:28 |
| fenn | fetch a script that downloads a new version of autogenix? | 18:28 |
| kanzure | where we can set up a centralized server for the equivalent of 'DTDs' or file-format-meta-information if that makes any sense | 18:28 |
| kanzure | fenn: in that case, yes, since the package format changed. But if it was just any other file (i.e., the package might link to a .HTML, and you don't have something that can do HTML for some reason) -- then go grab a type of program that can deal with it, such as a viewer. | 18:28 |
| fenn | uh.. parameter wrapping library.. cant we just use XML? | 18:29 |
| kanzure | probably :) | 18:29 |
| kanzure | then we just need a program to make the XML files | 18:29 |
| fenn | hurr | 18:29 |
| epitron | kanzure: hmm... check if your everydns.net account is still setup | 18:29 |
| kanzure | doing it by hand is lots of overhead | 18:29 |
| fenn | no we'd be writing the xml files by hand | 18:29 |
| fenn | because that's the source code | 18:29 |
| kanzure | nono | 18:29 |
| fenn | not in XML of course, in slip-xml | 18:29 |
| kanzure | okay, if we were doing all of the packages in python, sure | 18:29 |
| kanzure | I was about to say, the XML files would be like an extra step in your makefile process | 18:30 |
| epitron | kanzure: everydns.net is up, and it's answering DNS lookup requests, AND it's resolving all of MY domain names, but *.heybryan.org times out | 18:30 |
| fenn | this way all the inherited properties automatically update without any changes to the source | 18:30 |
| kanzure | epitron: because I didn't switch. I thought you were telling me that everydns.net was down, so I was going to wait until ... forget it. Hold on. | 18:30 |
| fenn | slip-xml is just xml written in a different notation, it's not actually python | 18:30 |
| kanzure | hm | 18:30 |
| kanzure | okay | 18:30 |
| epitron | kanzure: no, that's not true | 18:30 |
| kanzure | epitron: huh? | 18:30 |
| epitron | it's because of something internal to everydns | 18:30 |
| epitron | i could look up www.microsoft.com on my own DNS even though ms isn't pointing at it | 18:31 |
| fenn | although, just doing it in python is an interesting idea | 18:31 |
| epitron | i could even have my dns give a different IP | 18:31 |
| kanzure | epitron: I am still on a different DNS | 18:31 |
| epitron | that has nothing to do with this problem | 18:31 |
| kanzure | fenn: I was wondering about inline possibilities | 18:31 |
| epitron | this problem is that everydns doesn't show your heybryan.org records | 18:31 |
| epitron | so that if you DO switch | 18:31 |
| fenn | inline what? | 18:31 |
| epitron | you won't et anything | 18:31 |
| kanzure | inline parameter specifications | 18:31 |
| fenn | please elaborate | 18:31 |
| kanzure | myLibraryObject->addParameter("hey", objectTypeID, "natural language description for humans") where objectTypeID would refer to, say, XHTML 3.2 | 18:32 |
| kanzure | in python this may make sense | 18:32 |
| fenn | sure | 18:32 |
| kanzure | but hardcoded in a C program? heh' kind of silly, perhaps | 18:32 |
| epitron | ... | 18:32 |
| epitron | i'm at work guys | 18:32 |
| epitron | i don't have a lot of time | 18:32 |
| epitron | you wanna focus? | 18:32 |
| epitron | i just need kanz to check his everydns account | 18:32 |
| kanzure | epitron: oh, you say it does not have my records? | 18:32 |
| kanzure | okay | 18:32 |
| epitron | and make sure it's still setup | 18:32 |
| kanzure | yes, it is setup | 18:33 |
| fenn | object.parameters.add("hey", type, '''docstring''') | 18:33 |
| epitron | ok, so something is messed up about it | 18:33 |
| fenn | i'd rather something more structured though | 18:33 |
| kanzure | fenn: sure | 18:33 |
| kanzure | fenn: I think that 'type' might be too simple | 18:33 |
| epitron | kanzure: can you gimme the login/pass? | 18:33 |
| kanzure | I guess it could be an ID to the file format | 18:33 |
| kanzure | epitron: no | 18:33 |
| epitron | maybe you setup some detail | 18:33 |
| epitron | well, change your pass first | 18:33 |
| epitron | :) | 18:33 |
| kanzure | HostTypeValueMXTTLDelete | 18:34 |
| kanzure | heybryan.orgA70.113.54.1123600[delete] | 18:34 |
| kanzure | heybryan.orgMXheybryan.org03600[delete] | 18:34 |
| kanzure | static.heybryan.orgCNAMEbito.ponzo.net3600[delete] | 18:34 |
| fenn | object.parameters.add("hey", type, '''docstring''', version) :) | 18:34 |
| kanzure | but you said more structured | 18:34 |
| epitron | that's not the problem | 18:34 |
| kanzure | epitron: that's all that I changed :) | 18:34 |
| epitron | the problem is that it doesn't seem to even know that heybryan.org is one of its zones | 18:34 |
| kanzure | hm | 18:34 |
| fenn | oh i mean like, object.type = "furniture" | 18:34 |
| kanzure | "Currently: no one has access to heybryan.org's zone data" | 18:34 |
| epitron | i'm thinking you might've clicked the wrong thing with you crated it | 18:34 |
| fenn | object.docstring = ''' here is a long winded description in natural language''' | 18:34 |
| kanzure | fenn: that might be too generalized. I'd like the object type ID to correlate one-to-one with the wiki/db. | 18:34 |
| kanzure | epitron: okay, try now | 18:35 |
| fenn | what's more specific? type = loungechair? | 18:35 |
| epitron | trying | 18:35 |
| fenn | type = loungechair5547 | 18:35 |
| epitron | timeout | 18:35 |
| kanzure | epitron: what about http://heybryan.org/ ? | 18:35 |
| epitron | kanzure: in the left column, is it under "primary domains" or one of the others? | 18:35 |
| kanzure | primary | 18:36 |
| kanzure | fenn: I suppose. First come first serve naming basis? | 18:36 |
| epitron | that's really weird | 18:36 |
| fenn | ID = 4206f054bb8070f0e4456aac51592e7e | 18:36 |
| kanzure | hm | 18:36 |
| kanzure | if you make it complex, | 18:36 |
| kanzure | then you might force them to use a way to select file formats from the database | 18:36 |
| kanzure | so that they use standardized tools to interpret files and so on | 18:37 |
| kanzure | instead of doing it themselves | 18:37 |
| kanzure | that might be interesting? | 18:37 |
| kanzure | but it might get OOPy, like Java. | 18:37 |
| kanzure | (Java's library, I mean) | 18:37 |
| fenn | uh.. i'd prefer ease of hackability | 18:37 |
| kanzure | me too | 18:37 |
| fenn | since we're stupid | 18:37 |
| kanzure | epitron: I need to run for a few minutes | 18:37 |
| fenn | and people who take advantage of non-public interfaces should expect to get bitten | 18:37 |
| fenn | the file.pack or filepack-information.html is exactly analogous to the "control" file in a debian package | 18:40 |
| fenn | and that's what i was writing earlier, the valid types of statements you could make in that file | 18:41 |
| fenn | but it got sorta out of hand | 18:41 |
| epitron | hahaha | 18:49 |
| epitron | this project you guys are working on sounds insane | 18:49 |
| epitron | how are you going to do this with the horrible software dev tools we have today, and without the collaboration of major players in various fields? | 18:51 |
| epitron | why don't you just make something that's not bound to fail completely | 18:51 |
| epitron | like.. a self-replicating gnat | 18:51 |
| epitron | build UP to automated-knowledge-system | 18:52 |
| epitron | danny hillis has been trying to get this thing going too -- but at a much more tractable level => http://freebase.com | 18:52 |
| epitron | people would build their collaborative project things on top of freebase, and a pile of other things | 18:53 |
| epitron | and hopefully some new programming environment, like intentional software | 18:53 |
| epitron | although intentsoft still creates "dead" programs | 18:54 |
| epitron | so it's kinda doomed as well | 18:54 |
| fenn | epitron: if we could make a self replicating gnat, we would | 18:54 |
| fenn | as it is, even reprap is fumbling | 18:55 |
| fenn | i'm not trying to make a database of all the world's knowledge | 18:55 |
| epitron | okay.. | 18:56 |
| fenn | a specific subset of DIY projects, namely the ones that are useful for making other DIY projects | 18:56 |
| epitron | ok | 18:56 |
| epitron | but restricting the scope of the domain isn't going to restrict the scope of the knowledge system | 18:57 |
| fenn | kanzure: we want functionality loops, not dependency loops... | 18:57 |
| epitron | it's still going to have a automated-knowledge-management-system | 18:57 |
| epitron | it'll just have less knowledge in it :) | 18:57 |
| kanzure | fenn: filepack-information.html is supposed to be something like a paper on the subject of 'file'. Does your comment on filepack-information.html still hold? | 18:57 |
| fenn | x provides x, cycle cycle cycle, now you have 9 million x | 18:57 |
| fenn | x depends x, you have nothing | 18:58 |
| kanzure | epitron: yes, I know about freebase - I sent in my resume | 18:58 |
| epitron | that's an interesting idea | 18:58 |
| epitron | buuut.. that's emergence | 18:58 |
| epitron | people are notoriously bad at predicting emergent behaviour | 18:58 |
| kanzure | that's not emergence | 18:58 |
| epitron | it's basically trial and error | 18:58 |
| fenn | kanzure: no, what i'm talking about explains the dependencies of the technology you are building | 18:58 |
| fenn | not the software | 18:59 |
| kanzure | fenn: okay. | 18:59 |
| epitron | i still don't know what you mean by a functionality loop | 18:59 |
| fenn | but it doesnt describe the details | 18:59 |
| epitron | what's an example? | 18:59 |
| kanzure | okay, have you ever seen a mining facility? | 18:59 |
| fenn | epitron: woodworkers often make tools out of wood | 18:59 |
| kanzure | they have large machinery digging into the ground | 18:59 |
| kanzure | this machinery is made out of metal | 18:59 |
| kanzure | which was in the ground in the first place | 18:59 |
| epitron | oh | 18:59 |
| epitron | so just technology feedback | 19:00 |
| kanzure | that's ambiguous, but I guess | 19:00 |
| epitron | how is functionalty loop better? :) | 19:00 |
| fenn | yeah it's positive feedback | 19:00 |
| kanzure | technology feedback may as well mean "let's give them comments" | 19:00 |
| kanzure | but yes, it's a feedback loop | 19:00 |
| epitron | haha | 19:00 |
| epitron | right | 19:00 |
| epitron | +loop | 19:00 |
| epitron | software is notoriously bad at technology feedback though | 19:01 |
| kanzure | fenn: we might have to investigate freebase more closely | 19:01 |
| epitron | we reinvent the wheel constantly | 19:01 |
| epitron | everything is man-made | 19:01 |
| fenn | well, of course | 19:01 |
| epitron | if you're trying to have software be a part of the loop, your loop is broken | 19:01 |
| fenn | if you have a code generator you are fucking up badly | 19:01 |
| fenn | nobody writes assembly code any more (unless they have to) | 19:02 |
| epitron | if the datastructure for your source code is a text file, while the datastructure for your molecule is some complex efficient high dimensional thing... | 19:02 |
| kanzure | fenn: for example - helium: http://freebase.com/view/guid/9202a8c04000641f80000000045be4e9 -- but it does not look tech-oriented? don't know | 19:02 |
| epitron | you got your design upside-down :) | 19:02 |
| fenn | epitron: the text file is just a serialization format, which is necessary for transmitting information around the net | 19:03 |
| epitron | we send lots of stuff around the net | 19:03 |
| fenn | kanzure: i'm going to guess it's a structured wikipedia, without even looking | 19:03 |
| epitron | that isn't text files :) | 19:03 |
| kanzure | fenn: yes. So it has the structured data for various objects. It does not look project-oriented. Can we make use of it or not? | 19:03 |
| fenn | geez they dont have very many fields | 19:04 |
| kanzure | hehe :) | 19:04 |
| kanzure | they need to open up CRC | 19:04 |
| fenn | what about isotope ratios | 19:04 |
| kanzure | (the chemistry/physics handbook) | 19:04 |
| fenn | sure you can derive it since there's only two isotopes, but not always | 19:04 |
| epitron | anyhow, i'm done. :) | 19:04 |
| epitron | ttyl | 19:04 |
| fenn | ciao chicka mao mao | 19:05 |
| fenn | kanzure: also, their url isnt very structured :\ | 19:05 |
| fenn | 9202a8c04000641f80000000045be4e9 | 19:05 |
| fenn | how about physics/elements/helium | 19:05 |
| kanzure | heh | 19:06 |
| kanzure | I want multiple ontologies | 19:06 |
| kanzure | so that we can do virtual linking | 19:07 |
| fenn | yeah sure why not | 19:07 |
| kanzure | (pretty simple request, IMO) | 19:07 |
| fenn | they'd all point to the same hash | 19:07 |
| kanzure | first of all, this website seems too strict and structured | 19:07 |
| kanzure | even though structure is what we want, I think just dumping our files into a mediawiki installation with some server daemons would work | 19:07 |
| kanzure | and let people edit and backup copies of the db | 19:07 |
| kanzure | right? | 19:07 |
| kanzure | that's what I've been imagining skdb as | 19:07 |
| fenn | no, you need to validate everything | 19:07 |
| kanzure | ah, that's right | 19:08 |
| fenn | otherwise you cant do automated processing | 19:08 |
| kanzure | that should just be a flag | 19:08 |
| fenn | this is why openfarmtech is ... less than useful | 19:08 |
| kanzure | it should be a flag for confirmation or something | 19:08 |
| kanzure | or maybe a certain namespace can be "open to the public" and the other is "validated" ? | 19:08 |
| kanzure | and admins would import data into the validated-namespace | 19:08 |
| fenn | you can apply different validation codes to each revision | 19:09 |
| fenn | like, 'validated, passed unit tests, assembled irl, tested irl' | 19:09 |
| fenn | validation is like making sure the file format is correct, in w3c speak | 19:10 |
| kanzure | oh | 19:10 |
| kanzure | you don't mean "we've checked the books" | 19:10 |
| kanzure | but that could be another level of validation | 19:10 |
| fenn | oh, i forgot about that | 19:10 |
| kanzure | okay, so fact-checks, format validation, software unit tests, assembled irl, tested irl | 19:10 |
| fenn | on the other hand, i dont have much faith in engineering from first principles | 19:10 |
| fenn | well, i do, but i also demand that it be tested experimentally | 19:11 |
| kanzure | so? fact-check would be for, ex: hydrogen mass | 19:11 |
| fenn | which isotope? :) | 19:11 |
| fenn | book will say something stupid like 1.00123 | 19:11 |
| fenn | but mass of 1H is like .9999 | 19:12 |
| kanzure | "which isotope" would have to be a question answered by the file format specs, i.e. either you can specify isotopes with sub-datasets, or you could have one file per isotope if things start getting ridiculous (and then a zip for hydrogen plus its three isotopes or whatever) | 19:12 |
| fenn | ah its 1.00783 | 19:12 |
| kanzure | deutrium, hydrogen, titrium? I forget | 19:12 |
| fenn | mass of 1H is 1.00783 - i guess that's proton plus electron minus binding energy | 19:14 |
| fenn | units electronmass | 19:15 |
| fenn | Definition: 5.485799110e-4 u = 9.1093819e-31 kg | 19:15 |
| fenn | Unknown unit 'hydrogenmass' | 19:16 |
| fenn | units protonmass Definition: 1.00727646688 u = 1.6726216e-27 kg | 19:16 |
| fenn | units protonmass+electronmass amu * 1.007825 | 19:16 |
| kanzure | :) | 19:16 |
| fenn | wtf why is protonmass 1.00727 amu? | 19:17 |
| fenn | 12C/12 seems kinda silly | 19:18 |
| kanzure | do we want to try to integrate with freebase or not. | 19:22 |
| kanzure | intuition says no | 19:22 |
| fenn | no | 19:22 |
| fenn | freebase is the death throes of academia, imho | 19:22 |
| fenn | cyc is way better anyway | 19:23 |
| kanzure | people keep on talking about cyc, but each time I look at it it just seems like a giant vocabulary association deal | 19:25 |
| kanzure | ".. of everyday common knowledge" | 19:25 |
| kanzure | what the hell does that mean | 19:25 |
| fenn | yes but it is code too, not just data | 19:25 |
| fenn | stuff like, water flows downhill, people age 1 year per year | 19:25 |
| kanzure | oh, btw, I've been trying to meet up with the texai guy | 19:25 |
| fenn | you can use that to do NLP stuff | 19:25 |
| kanzure | he lives in Austin somewhere. | 19:26 |
| kanzure | neurolinguistic programming? natural language processing? | 19:26 |
| fenn | natural language processing | 19:26 |
| kanzure | code too, interesting | 19:26 |
| kanzure | is it worth investigating? | 19:26 |
| kanzure | I think our plan of action is pretty solid as it is | 19:26 |
| fenn | investigate using cyc in skdb? | 19:27 |
| kanzure | but I don't want to ignore a potentially useful tool just because I'm confident ;) | 19:27 |
| kanzure | yeah | 19:27 |
| fenn | it might be very useful, yes | 19:27 |
| fenn | i'll have to look at it again, first time i didnt know any lisp so it was gibberish | 19:27 |
| kanzure | lisp is something like | 19:28 |
| kanzure | ($var $var2 $var3 function) | 19:28 |
| kanzure | right? | 19:28 |
| fenn | hmm.. | 19:28 |
| fenn | no | 19:28 |
| fenn | you cant express lisp in most languages | 19:28 |
| fenn | ((predicate $var) function) or ((function) function) and more | 19:30 |
| fenn | you can define a function that accepts 3 vars though, of course | 19:31 |
| fenn | anyway, lisp has a steep learning curve and turns your brain inside out | 19:31 |
| fenn | kanzure: i just watched the douglas lenat google tech talk on cyc a couple days ago | 19:39 |
| kanzure | good? | 19:40 |
| fenn | it made some sense, a good starting point | 19:40 |
| fenn | Cyclify Austin meets on the second Thursday of each month in Austin, TX. | 19:41 |
| kanzure | ah crap | 19:41 |
| fenn | now you have to go | 19:41 |
| kanzure | that was yesterday | 19:41 |
| fenn | no? | 19:42 |
| fenn | tomorrow | 19:42 |
| fenn | they're headquartered in austin btw | 19:42 |
| kanzure | yes | 19:43 |
| kanzure | alright, I'll send an email | 19:43 |
| fenn | it'll probably be way over your head | 19:44 |
| fenn | but you can still promote your own agenda | 19:45 |
| kanzure | http://cyc.com/ seems to be down | 19:45 |
| fenn | www.cyc.com | 19:46 |
| fenn | argh java | 19:47 |
| fenn | hmm maybe i broke it | 19:48 |
| drazak | mmmm b5 | 20:14 |
| kanzure | http://heybryan.org/mediawiki/index.php/Artificial_dopamine_receptor | 20:16 |
| drazak | yay slow internet | 20:18 |
| kanzure | Is it my connection ? | 20:18 |
| drazak | nope | 20:18 |
| drazak | I'm grabbing a b5 episode from my dad's, and watching amusic video on youtube | 20:19 |
| drazak | I need to get a cool college chem textbook (IE. one that's on textbooktorrents, and then I can better help you guys) | 20:19 |
| fenn | kanzure: one hand-drawn picture would sum that all up nicely | 20:20 |
| drazak | kanzure: I'd only be worried about using distilled water, and making sure that your glassware is very clean, oh and making sure that you don't fuck up too bad, you could make something very flammable | 20:21 |
| kanzure | fenn: sure | 20:21 |
| kanzure | drazak: I have a server full of them. What would yo uwant? | 20:22 |
| fenn | hah textbooktorrents.. i wonder what the universitie think about that | 20:23 |
| drazak | kanzure: ah, someone mentioned it the other day in ##chemistry | 20:23 |
| fenn | no knowledge for you! bad student! | 20:24 |
| kanzure | drazak: yes, that was me | 20:24 |
| drazak | kanzure: it'd be nice to have that one :P | 20:24 |
| kanzure | drazak: I'll give you a few indices of my files, if you set me up an FTP server to drop files on | 20:24 |
| kanzure | I'd give you a login, but I happen to have my whole hard drive backed up on the file server | 20:24 |
| kanzure | (and it's a custom firmware OS that sucks, so there's no real permissions like that) | 20:25 |
| drazak | kanzure: I don't do ftp, you'kk have to use scp or sftp | 20:25 |
| kanzure | that's fine | 20:25 |
| kanzure | lemme get you your indices | 20:25 |
| * drazak finds a window on the appropriate machine | 20:26 | |
| drazak | I thinkm I'm going to have to make a pub/priv keypair for this laptop | 20:26 |
| kanzure | drazak: There are a few subdirs. Take your pick: analytical chem, biochem, ChE, current protocols in food analysis, visual encyclopedia of chem-eng, inorganic chem, methods, mineralogy, nuclear chem, organic chem, organometallic chem, other, photochem, supramolecular chem, theoretical chem | 20:28 |
| kanzure | the root dir for chem has general college stuff | 20:28 |
| drazak | lets see here | 20:29 |
| kanzure | http://heybryan.org/docs/books/ | 20:29 |
| drazak | ana, bio, ChE, inorg, nuc, organ, organomet, theo | 20:29 |
| drazak | that's a good start | 20:29 |
| kanzure | wtf's with this directory output | 20:29 |
| kanzure | I'll fix. | 20:31 |
| fenn | eh why not just setup a small web server on kanzure's end | 20:33 |
| fenn | then you can use nice tools like wget | 20:33 |
| kanzure | fenn: I have all of my files on there, including some rather personal items | 20:33 |
| drazak | fenn: because sftp/scp is encrypted | 20:33 |
| kanzure | fenn: and it's not a permission-based system | 20:34 |
| fenn | they're textbooks, not plans for an interocitor | 20:34 |
| kanzure | well, there's my entire email content on there | 20:34 |
| fenn | and kanzure you can allow/disallow by directory | 20:34 |
| kanzure | fenn: it's on a dedicated "external data storage device" | 20:34 |
| kanzure | with its own HTTP/FTP daemons | 20:34 |
| fenn | or not.. i dont care much | 20:34 |
| fenn | oh like a slu2 | 20:34 |
| kanzure | hm? | 20:35 |
| drazak | yeah, that's what it is, probably | 20:35 |
| fenn | a toaster | 20:35 |
| kanzure | so | 20:35 |
| kanzure | haha, I didn't think of something | 20:35 |
| kanzure | I can let you access a base directory | 20:35 |
| kanzure | and it's too stupid to allow /../ | 20:35 |
| kanzure | it's probably Windows Mobile Edition or something | 20:35 |
| drazak | no | 20:36 |
| drazak | it's proprietary | 20:36 |
| fenn | sometimes i feel like transhumanism is like.. 'every sperm is sacred' | 20:37 |
| fenn | i have this baby rabbit in my room i found today, he'll probably be dead by morning | 20:38 |
| kanzure | 'every sperm is sacred' ?? it's more about personal modification. With me, I want to process more information, or do it better. | 20:39 |
| fenn | man i wish i had cad software like this http://reprap.org/bin/view/Main/AssemblingDarwinMachinery | 20:43 |
| kanzure | maybe I should go re-install proxypass for apache? but then you'd be going through the mini-data-server and back through the linux server, which will be slow | 20:44 |
| Enki-2 | http://www.gnu.org/fun/jokes/unreliable.net.html | 20:53 |
| drazak | kanzure: it won't be slow if they're on the same network | 20:53 |
| kanzure | eh, let me see what I can do | 20:54 |
| Enki-2 | win joke is WIN. | 20:56 |
| kanzure | Enki-2: | 21:02 |
| kanzure | On Wednesday 09 April 2008, Eric Yu wrote: | 21:02 |
| kanzure | > It's much easier to fix a few disabilities. Also, autistic people | 21:02 |
| kanzure | > have more of a disability because people tend to be non-autistic. | 21:02 |
| kanzure | That's like saying linux is a disability. | 21:02 |
| epitron | linux is a disability when you have to work in a windows network | 21:04 |
| epitron | (linux is a bad example beacuse it works very well with windows networks. ... so imagine its' linux from 1995) | 21:05 |
| epitron | kanzure: you might wanna fix your domain.. right now it's pointing at your old AND your new nameserver at the same time | 21:06 |
| epitron | that can't be good | 21:06 |
| epitron | servers might ban you for having inconsistent records | 21:06 |
| epitron | (it could look like a DNS attack) | 21:07 |
| kanzure | ah, okay | 21:08 |
| kanzure | epitron: did windows have networks in 95? | 21:08 |
| epitron | yep | 21:08 |
| epitron | SMB is old | 21:08 |
| epitron | it's from the novell days | 21:08 |
| drazak | from when you ran windows 95 ontop of novell | 21:11 |
| kanzure | epitron: try http://heybryan.org/ please :) | 21:15 |
| kanzure | I don't know if everydns.net is working for it or not | 21:15 |
| kanzure | fenn: okay, I'm attending tomorrow night's Cyc meetup | 21:37 |
| epitron | kanzure: it's working here | 21:43 |
| epitron | oh wait | 21:43 |
| epitron | it was cached | 21:43 |
| epitron | haha | 21:43 |
| epitron | i dunno man | 21:43 |
| epitron | try deleting the zone and recreating it | 21:43 |
| epitron | something probably got fucked up when you were messing with it | 21:43 |
| epitron | maybe you can't delete the A record without it thinking the whole thing has been destroyed, even if you add it again later | 21:44 |
| kanzure | fenn: I think we might have a small server farm willing to host us. | 22:03 |
| kanzure | So what was the next step that we were going to take? Was I going to go write the XML standards for fileformat IO information? | 22:10 |
| kanzure | http://www.koreamed.org/SearchBasic.php?RID=368379&DT=1&QY=openfree+%5BALL%5D <-- but on the other hand, wtf? | 22:18 |
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!