--- Day changed Tue Oct 14 2008 | ||
xp_prg | wow cool, what does it do? | 00:00 |
---|---|---|
kanzure | You haven't been listening? I thought we told you about it earlier today. | 00:05 |
xp_prg | no | 00:06 |
kanzure | http://synbioss.sf.net/ | 00:06 |
kanzure | The web interface does what we want to do with 'biobench' except without the AJAX interface. | 00:06 |
kanzure | Hell, I was talking with you on the mailing list about this .. | 00:06 |
kanzure | wtf? | 00:06 |
fenn | ha | 00:10 |
xp_prg | wow that is cool software! | 00:11 |
kanzure | I wonder where 'unum' is found. | 00:13 |
kanzure | .. any hints? | 00:17 |
bkero | Like E Pluribus unum? | 00:17 |
kanzure | No, it's a python module. | 00:17 |
kanzure | For synbioss. | 00:17 |
xp_prg | kanzure why do we want to re-create synbioss? | 00:18 |
kanzure | We're not .. | 00:18 |
fenn | prolly bad path or not installed from src yet | 00:18 |
kanzure | the module, fenn? | 00:18 |
fenn | any files called unum in the synbioss release? | 00:18 |
kanzure | xp_prg: AJAX interface + ikiwiki + git backend for the distribution of those files | 00:18 |
kanzure | fenn: no. | 00:18 |
kanzure | "Ecology: The Ascendent Perspective" | 00:18 |
kanzure | oops | 00:18 |
kanzure | bad paste | 00:18 |
kanzure | http://synbioss.sourceforge.net/ | 00:19 |
kanzure | http://sourceforge.net/project/showfiles.php?group_id=203655&package_id=242861&release_id=625492 | 00:19 |
fenn | anyway, it appears to be here: http://sourceforge.net/projects/unum/ http://home.scarlet.be/be052320/Unum.html | 00:19 |
kanzure | this being the 'desktop simulator' | 00:19 |
fenn | and sounds like a great idea to me :) | 00:19 |
kanzure | hm? | 00:19 |
kanzure | 'Unum stands for 'unit-numbers'. It is a Python module that allows defining and manipulating true quantities, i.e. numbers with units like volts, hours, meter-per-second. Consistency between units is checked for every operation involving quantities.' | 00:19 |
fenn | a Python module that allows defining and manipulating true quantities, i.e. numbers with units like volts, hours, meter-per-second. Consistency between units is checked at each expression evaluation; unit conversion and unit output formatting are performed automatically when needed. | 00:19 |
kanzure | geeze | 00:19 |
kanzure | hrm | 00:19 |
xp_prg | how will our project biobench be different then there as far as logic is concerned? | 00:20 |
kanzure | wonder how they're using it. | 00:20 |
kanzure | xp_prg: logic? not at all .. | 00:20 |
kanzure | their code easily serves as backend | 00:20 |
kanzure | but notice that their wiki is kind of wrong | 00:20 |
kanzure | it's not git and it's not ikiwiki | 00:20 |
kanzure | it's on some randomass mediawiki | 00:20 |
kanzure | whereas those SBML files belong in a repository .. | 00:20 |
xp_prg | hmm... ok, so basically all we are doing is adding an ajax front end to their program? | 00:20 |
kanzure | No. | 00:21 |
kanzure | There's also the backend ikiwiki + git stuff. | 00:21 |
xp_prg | but how does that affect the logic? | 00:21 |
kanzure | What? | 00:21 |
xp_prg | the algoirthms | 00:21 |
kanzure | It won't. But the codebase has to be manipulated a bit. | 00:21 |
xp_prg | are we going to use different algoirthms then sybioss or the same ones? | 00:21 |
kanzure | Same ones. | 00:22 |
kanzure | Fundamentally they're the same thing, but it's the architecture that matters | 00:22 |
xp_prg | so what is the real difference besides the ajax front end then? | 00:22 |
kanzure | ikiwiki + git | 00:22 |
xp_prg | which offers what exactly? | 00:22 |
xp_prg | by the way I program python all day :) | 00:22 |
kanzure | Do you know what a repository is? | 00:24 |
xp_prg | yes | 00:24 |
fenn | it offers the ability to edit and massage plain text files without going through some horrible html interface | 00:24 |
fenn | and also easy forking of the whole project | 00:24 |
kanzure | "Git is a free distributed revision control, or software source code management project with an emphasis on being fast." "A wiki is a page or collection of files designed to enable anyone who accesses it to contribute or modify content.[1][2] Wikis are often used to create collaborative websites and to power community websites." | 00:25 |
fenn | and still panders to people who simply must have a pretty html front end | 00:25 |
kanzure | or pretty ajax frontend | 00:25 |
xp_prg | ok | 00:25 |
* fenn would prefer plain html | 00:25 | |
kanzure | fenn: I'd disagree. AJAX, or no HTML at all .. just go CLI at that point. | 00:26 |
fenn | i've seen horrible crimes committed with AJAX | 00:26 |
fenn | HTML on the other hand is hard to do evil with | 00:26 |
kanzure | are you kidding :( | 00:26 |
kanzure | it's all evil | 00:26 |
fenn | there's only so many ways they can screw up though | 00:26 |
fenn | so it's a lesser evil :) | 00:27 |
kanzure | I'm surprised how exactly synbioss matches what I was thinking of | 00:27 |
fenn | we are shooting for interoperability, that's the whole point | 00:27 |
kanzure | it's spot-on .. 'unum' is freaky too | 00:28 |
kanzure | although obvious. | 00:28 |
fenn | sigh i guess i'll look at synbioss's stupid pd | 00:28 |
fenn | 00:28 | |
kanzure | link? | 00:28 |
fenn | its just the tutorial, closest i could find to a screenshot | 00:28 |
kanzure | another thing to poke at is the BioModelWiki thing going on there .. don't know how they're storing, generating, or doing anything at all with the downloadable models there | 00:32 |
kanzure | looks like it's just SBML created from looking through the literature | 00:32 |
kanzure | which I'm fairly surprised about .. wondering what sort of tie-ins with biobricks they have there | 00:32 |
kanzure | http://ask.slashdot.org/askslashdot/03/04/25/2054223.shtml?tid=126 open source inventory control software | 00:34 |
fenn | i thought it was the same people | 00:34 |
xp_prg | what is SBML ? | 00:34 |
fenn | synbio modeling language? | 00:35 |
kanzure | systems biology markup/modeling language | 00:36 |
* kanzure sits on the sbml mailing lists | 00:36 | |
xp_prg | wow cool | 00:37 |
* xp_prg feels excited now that we have a backed already written! | 00:38 | |
fenn | way to go team! | 00:38 |
kanzure | hah | 00:38 |
fenn | beer's on me dudes | 00:38 |
kanzure | get back to work. | 00:38 |
fenn | i want to celebrate your progress | 00:38 |
kanzure | a few things; 1) ikiwiki+git+ajax stuff; 2) inventory management, automated procurement for bio materials and eventually equipment / instructions; 3) pulling from giant databases like neurocommons for helping the user to design ridiculous things that they probably don't have the expertise to be designing anyway .. | 00:42 |
kanzure | but .repo format and the formats used by synbioss should probably be done before #2 | 00:42 |
kanzure | s/done/synchronized/ | 00:42 |
fenn | what is .repo exactly? | 00:43 |
kanzure | a load of shit | 00:44 |
kanzure | have you seen the files? | 00:44 |
fenn | i mean what you expect to turn it into | 00:44 |
kanzure | the .tar + yaml metadata | 00:44 |
kanzure | or no .tar, either way | 00:44 |
fenn | duh i must be missing something here | 00:44 |
kanzure | either everything in the same file, like the current .repo files (which I've discovered happens to sucks immensely) or in a tar | 00:44 |
fenn | yeah files were invented for a reason | 00:45 |
kanzure | does the duh signify you understand? because the second part of your message suggests you don't | 00:45 |
kanzure | s/sucks/suck/ | 00:45 |
fenn | i just feel stupid, writing my resume :( | 00:45 |
kanzure | "well, I went to school. does that count?" | 00:45 |
fenn | "should i say tcpdump _and_ tcp/ip knowledge or is the second implied by the first?" | 00:45 |
kanzure | tcpdump, then cite ISO standards for TCP/IP if you must | 00:46 |
fenn | ah can wrassle a whole herd o' com pee yoo tars | 00:46 |
kanzure | o' com pee yoo tars? | 00:46 |
kanzure | hrm | 00:46 |
kanzure | deep south? | 00:46 |
fenn | ken-tucky | 00:46 |
fenn | or southern indiana, if you prefer | 00:47 |
fenn | so .repo is basically skdb projects | 00:48 |
kanzure | except it sucks immensely. it's missing a few important things. | 00:49 |
kanzure | first of all, all artifacts/subcomponents are _not_ their own repo files which is a failure | 00:49 |
kanzure | also, there's no metadata specifying their interconnectivity really | 00:49 |
kanzure | there's some information by having an artifact as a child of the overall system | 00:49 |
fenn | well, if you hate the ADL stuff then dont try to polish a turd | 00:49 |
kanzure | and then there's stuff like "ParentArtifact=" and "InputArtifact=" and "OutputArtifact=" | 00:49 |
kanzure | I really like that it has all of this data already entered | 00:50 |
fenn | but it's useless | 00:50 |
kanzure | I'm not entirely sure of that yet | 00:50 |
kanzure | the format is so terrible that I can't determine that entirely | 00:50 |
fenn | hm. | 00:50 |
kanzure | there's input/output artifacts from each artifact, which is *sort of* like interconnectivity | 00:50 |
kanzure | http://heybryan.org/~bbishop/docs/repo/CFGs/72.png <-- it *looks* useful | 00:51 |
fenn | you dont need 300 different consumer products to show a proof of concept | 00:51 |
fenn | shit why did i click on that | 00:51 |
kanzure | oh crap, you don't like big images, that's right | 00:51 |
kanzure | have you even seen a part of it yet? or did you use imagemagick to size it down | 00:52 |
fenn | i've seen it and i think neato/fdm looks way better | 00:52 |
fenn | and it's not godfuckingawful huge | 00:52 |
kanzure | I ran it through both of those and it looked the same | 00:52 |
kanzure | I'll try again Real Soon | 00:52 |
kanzure | my next run through or something. | 00:52 |
fenn | what's generating the .dot files/ | 00:52 |
kanzure | (btw, I have successfully extracted the files from within the .repo files .. but the data format looks peculiar, I think it's encoded) | 00:52 |
kanzure | well, I exported all of the .repo files into a mysql database | 00:53 |
kanzure | with about 9 different tables to represent the different types of elements | 00:53 |
fenn | i ask this because the graphs would be a lot better if the links were colorized by flow type | 00:53 |
kanzure | instead of labeled by flowtype? | 00:53 |
fenn | yeah, the labels just get in the way | 00:54 |
fenn | not that i know what you're going to do with them anyway | 00:55 |
kanzure | http://dhcp-84-253.me.utexas.edu/phpmyadmin/ if you want to poke at the database | 00:55 |
kanzure | hrm | 00:55 |
kanzure | I was thinking this should be obvious | 00:55 |
kanzure | wouldn't it be a good idea to generate the YAML files that we wanted? | 00:55 |
kanzure | where the 'units' are going to be very basic such that it's just those flow types that the different parts (apparently) accept | 00:55 |
kanzure | each part being a new yaml file | 00:56 |
kanzure | and I guess I want to do it as a system of directories where I'll just tar them up in the end | 00:56 |
kanzure | or tar+gzip is ok too | 00:56 |
fenn | as long as you dont start calling 'flow types' units | 00:56 |
kanzure | 'allowables'. | 00:57 |
kanzure | but how does this mesh with our previous usage of 'units' ontology? | 00:57 |
fenn | it's horribly imprecise and has no numerical quantity | 00:57 |
fenn | like, i cant even guess at what they're talking about for most of the flows | 00:58 |
kanzure | sometimes it's not even a real flow .. sometimes it's supposed to represent an object leaning against another one | 00:58 |
fenn | you mean most of the time | 00:58 |
kanzure | heh | 00:58 |
kanzure | I haven't ran the stats. | 00:58 |
fenn | 'force flow' | 00:58 |
fenn | the force is strong in this bolt | 00:59 |
kanzure | i wish it told me such. | 00:59 |
kanzure | okay, so originally I was thinking it would just be a simple variable in the yaml file as one of the first few things that are specified .. | 01:00 |
kanzure | !!python/thingy/that/I/forget/but/is/necessary/for/specifying/general/type/of/yaml-doc | 01:01 |
kanzure | inputs: some general specification here | 01:01 |
kanzure | outputs: some general specification here | 01:01 |
kanzure | author: | 01:01 |
kanzure | trulymeta-data-crap here: ... | 01:01 |
kanzure | what's that "some general specification" if not "50 grams" ? | 01:01 |
kanzure | (I don't think it was going to be "50 grams" originally anyway) | 01:01 |
fenn | 50 grams isnt enough info | 01:01 |
fenn | is that, capacity: 50g? or weight: 50g? | 01:02 |
fenn | but it gets worse for other units | 01:02 |
kanzure | terrible problem. | 01:03 |
kanzure | what if parts had to 'negotiate' their interfaces? | 01:04 |
* fenn flees in panic | 01:05 | |
kanzure | you have a better idea? | 01:08 |
kanzure | I mean, there's no way for me to figure it all out upfront | 01:08 |
fenn | standards | 01:08 |
kanzure | because of the increasingly sophisticated interconnections that will be required | 01:08 |
kanzure | right | 01:08 |
kanzure | but right now we don't have those standards | 01:08 |
fenn | we'll just have to figure them out one by one | 01:08 |
fenn | or someone will | 01:08 |
kanzure | and in the mean time | 01:09 |
fenn | we can get by with a small number of standards to play with | 01:09 |
* fenn has a sense of deja vu | 01:09 | |
kanzure | aren't we right back to where we started | 01:10 |
fenn | yep | 01:10 |
fenn | i have NIH syndrome and no motivation, so this tend to lead to pacing in circles | 01:11 |
kanzure | I tend to pace in straight lines personally | 01:11 |
kanzure | pacing, by definition, involves going backwards | 01:11 |
kanzure | or looping etc. | 01:11 |
kanzure | would you kill me if I do 'allowable flow types' as inputs to the parts? | 01:11 |
fenn | no, do whatever ou want | 01:12 |
kanzure | I'll certainly never forgive myself because if it catches on it'll just be al oad of shit | 01:12 |
kanzure | *a load of shit | 01:12 |
fenn | yes it will be a load of shit, and already exists as ADL repo, so i dont see why you want to try to continue that | 01:12 |
kanzure | chicken and egg problem | 01:12 |
kanzure | ADL repo is an almost-egg | 01:13 |
fenn | data is easy to make up | 01:13 |
kanzure | valid data .. | 01:13 |
fenn | you dont need 300 repo's/projects to start with | 01:13 |
kanzure | you can't do much with just 1 seed project | 01:13 |
fenn | we need a word for a thing that a person can sit down and use without having to assemble into other things | 01:13 |
fenn | but this should also cover standalone nuts and bolts and wires | 01:14 |
kanzure | connectables? | 01:14 |
fenn | preferably a word that already exists | 01:14 |
fenn | like 'artifact' or 'product' or something | 01:14 |
fenn | a laptop is a: | 01:14 |
fenn | device? | 01:14 |
kanzure | tool? | 01:15 |
fenn | tool is too general | 01:15 |
fenn | i like device | 01:15 |
kanzure | you're going all abstract on me | 01:15 |
kanzure | for the first two variables of the yaml metadata file. | 01:15 |
kanzure | what the hell do I do .. just use the flow BS? :( | 01:16 |
kanzure | gaahh | 01:16 |
fenn | i what are you trying to do? define the skdb file format? | 01:16 |
kanzure | basically. | 01:16 |
kanzure | okay, screw it | 01:16 |
kanzure | no universal catch-all tool | 01:16 |
kanzure | no universal catch-all format, except for the distribution of furtherly detailed formats | 01:16 |
kanzure | so the yaml metadata is specifically for sharing data and dependencies | 01:17 |
kanzure | where dependencies are explicitly defined as other parts in the database | 01:17 |
kanzure | now, as for interconnectivity with flows and so on, that's specific to something else | 01:17 |
fenn | no because there could be a range of parts that saisfy the dependencies | 01:17 |
fenn | hex head bolt, allen head bolt, etc | 01:17 |
kanzure | can we call that 'dynamic dependencies' ? | 01:18 |
fenn | sure | 01:18 |
fenn | feel free to borrow all the computer science terminology | 01:18 |
kanzure | which is basically where the client checks the dependencies, which causes it to select over the database for things that fall within that range | 01:18 |
kanzure | still not helping. | 01:19 |
kanzure | there will have to be scripts that are specifically designed to check if those 'dynamic dependencies' are being satisfied | 01:19 |
kanzure | it'll be a very lengthy O(n) process | 01:19 |
kanzure | so it must be domain-specific .. | 01:20 |
fenn | i dont think its that difficult | 01:20 |
kanzure | how. | 01:20 |
fenn | it must be domain specific | 01:20 |
kanzure | right | 01:20 |
fenn | no sense comparing bolts and insulation | 01:21 |
kanzure | you know how we mentioned that the pre-synbioss stuff would require highly specialized knowledge to make the proper grammar and in the mean time we'd use randomly generated stuff? | 01:21 |
kanzure | isn't this the same? | 01:21 |
fenn | yes but i never liked the randomly generated grammar because i can't think that abstractly | 01:21 |
kanzure | the generation step is very simple: you have N components, and then a constraints format according to BNF format, or whatever format the grammar was to be specified in, and just have it fill that in in one swoop step. | 01:22 |
kanzure | only once that step has been done does the abstraction-thinking-issue crumble. | 01:22 |
kanzure | so maybe the important factor is the tool used to generate the 'grammars' or the highly specialized domain-dependent formulation of how to express dependencies | 01:24 |
kanzure | yikes, that's a mouthful | 01:24 |
fenn | i was just going to use python duck typing | 01:24 |
fenn | a centralized grammar file is a bad idea | 01:25 |
fenn | because: restrictions on what you can do with parts is very closely linked to what the part is | 01:25 |
kanzure | that's why it's domain specific -- so there'd be stuff for synbio, stuff for electronics, for .., etc. compartmentalization. | 01:26 |
kanzure | anyway, duck typing is fine as long as the methods that the yaml metadata point to actually exist. | 01:26 |
kanzure | btw, that sounds like 'negotiation' to me. | 01:26 |
kanzure | radio.skdb __add__ operator to computer.skdb .. one would have to check the other (or something) and if they don't have enough compatabilities between the variables they want to share and constants they want to specify, then failure and sadness. | 01:27 |
kanzure | beep | 01:31 |
fenn | ok, sounds fine. "negotiation" sounded like some silly paper about agent-based systems | 01:31 |
kanzure | ok, so in various langs you use 'hooks' for other functions to throw at various times, like after a timer is done, | 01:32 |
kanzure | so I imagine a 'hook' for the init method would be a good thing to specify in the metadata | 01:32 |
fenn | 'init method'? | 01:32 |
kanzure | and then let one object check its compatability by probing the hook to the init method | 01:32 |
kanzure | yes | 01:32 |
fenn | i dont know what that is supposed to mean | 01:32 |
kanzure | s/to/for/ | 01:32 |
kanzure | well, how would they have negotiations if they don't know what to call | 01:32 |
kanzure | I mean, one script has to be the polymerase and the other has to be the thing to be executed/read | 01:33 |
kanzure | to return variables and information .. right? | 01:33 |
kanzure | how else will interfacing compatability be determined. | 01:33 |
fenn | with attributes | 01:33 |
fenn | polymerase wants a NT-type: ss-DNA sequence-length: >10bp | 01:34 |
fenn | or something like that | 01:34 |
kanzure | you're reverting on me. the difference here is that instead of scanning for the first few "inputs" and "outputs" variables in the yaml metadata files, we now have some selected part, and it has a method that lets it check against another given part whether or not its compatible. so it has to go run the metadata scripts and python or whatever for that part. | 01:35 |
kanzure | guess it's somewhat the same thing. anyway. | 01:35 |
kanzure | it's amazing that any work gets done with APIs | 01:35 |
kanzure | I guess the main way it gets done is via the presumption of upstream v. end-users .. "this is the way it is, use it this way or else" | 01:35 |
kanzure | so the same with parts. there will probably be a central core of important components to mimic in their compatability interfaces. | 01:35 |
fenn | i am too slow, i can hardly read all this | 01:36 |
kanzure | sorry. I'll tone down my braindump. I think this is all equating to "there's no difference whatsoever and just do it anyway" | 01:36 |
fenn | yaml data is numbers and stuff that gets loaded into python class attributes | 01:36 |
kanzure | you're reverting on me. | 01:37 |
fenn | am not | 01:37 |
kanzure | were were discussing some alternatives besides the first few variables | 01:37 |
kanzure | yeah you are, we were discussing alternatives above | 01:37 |
fenn | i think your 'init method' is the class name | 01:38 |
fenn | radio.skdb start out with !python/skdb.radio | 01:39 |
fenn | then all the attributes of a radio | 01:39 |
fenn | and all the inherited attributes from electronic devices | 01:39 |
wrldpc | oi | 01:43 |
-!- xp_prg is now known as xp_gone | 01:59 | |
kanzure_ | wtf | 03:03 |
kanzure_ | The machine just turned off on me. | 03:03 |
kanzure_ | I was actively using it .. | 03:03 |
fenn | overheat | 03:04 |
fenn | PSU load | 03:04 |
kanzure_ | overheat. | 03:04 |
kanzure_ | it's been doing this increasingly often .. | 03:04 |
fenn | blow out the dust | 03:04 |
kanzure_ | ` | 03:04 |
kanzure_ | `````````````````,~\ | 03:05 |
kanzure_ | uh, dropped keyboard | 03:05 |
bkero | flimsy 2lb keyboards are hard not to drop | 03:06 |
fenn | real steampunkers use solid brass keyboards | 03:07 |
bkero | You mean these? | 03:08 |
bkero | http://steampunkworkshop.com/keyboard.shtml | 03:08 |
bkero | Made from 5lb model M's? | 03:08 |
kanzure_ | bah, my kb is powered on steam | 03:08 |
kanzure_ | it's 200 m tall and it's made out of 34 solid metric tons of brass | 03:09 |
fenn | bkero: that flimsy plastic thing? | 03:09 |
fenn | i dont see where you're supposed to put the coal? | 03:11 |
kanzure_ | why, in the laptop! | 03:11 |
kanzure_ | (the steam is wirelessly transported through the aether) | 03:12 |
fenn | steam doesn't travel in wires anyway | 03:13 |
fenn | what are they teaching kids in school i wonder | 03:13 |
kanzure_ | see /school | 03:13 |
fenn | do you even know how to re-babbitt a stopcock? | 03:13 |
kanzure_ | what the hell? | 03:13 |
kanzure_ | No? | 03:13 |
fenn | trim in an injector? | 03:14 |
kanzure_ | you talk in tongues. | 03:14 |
* fenn shakes head disapprovingly | 03:14 | |
kanzure_ | you'll be happy to know that I sat in on a lecture on 'team leadership' today | 03:14 |
kanzure_ | "Start every meeting with a motivating quote." | 03:15 |
kanzure_ | yay education | 03:15 |
fenn | preferably not something from neverness | 03:15 |
kanzure_ | that's a cold stab. | 03:15 |
kanzure_ | we all know team management talk is bullshit .. | 03:15 |
* fenn focuses on finishing resume before midnight | 03:16 | |
kanzure_ | you're still working on that? | 03:16 |
fenn | fuck i hate this shit so much you dont even know | 03:17 |
kanzure_ | why are you bothering again? | 03:18 |
fenn | because jobs that pay more than $10/hr generally require them | 03:20 |
bkero | http://www.engadget.com/2008/10/13/mr-blurrycam-reveals-the-updated-macbook-pro-899-laptop-model/ | 03:20 |
kanzure_ | so far I've never had to actually use my resume in landing me a >$10/hr job | 03:22 |
fenn | well whoopdeedoo | 03:22 |
bkero | I had to use my resume to land an $8/hour job, a $12/hour job, 2 $22/hour jobs, and a $35/hour job. | 03:23 |
kanzure_ | I mean, these people aren't really looking at resumes. it's weird. | 03:23 |
fenn | you should write a book on your success secrets | 03:23 |
kanzure_ | fenn: secrets? you mean, like, falling assbackwards into luck? | 03:23 |
fenn | ok i'm just going to leave off the whole chronolgy section since it's demeaning | 03:23 |
fenn | kya! letter sent! | 03:58 |
fenn | suppose i should like, post this to me website or something now | 04:00 |
fenn | (as if anyone ever looks at it) | 04:00 |
kanzure_ | i do, i do! | 04:00 |
fenn | ok lemme hack off the email addy | 04:00 |
kanzure_ | ? keep it on? | 04:01 |
willPow3r | who was asking yesterday about using a laptop as a 2nd monitor? | 04:02 |
willPow3r | http://www.maxivista.com/ | 04:02 |
kanzure_ | gene. | 04:02 |
fenn | http://fennetic.net/ben_lipkowitz_resume.pdf | 04:06 |
fenn | the thing is, you can learn like 50% of that in under an hour | 04:07 |
kanzure_ | nobody knows that. | 04:07 |
kanzure_ | I mean, if they did, we'd all be out of a job. | 04:07 |
fenn | *ahem* | 04:07 |
fenn | ok, maybe more like a day or two | 04:08 |
fenn | the ol' 5x what you expect it to take rule | 04:08 |
kanzure_ | how much Japanese? | 04:14 |
bkero | http://www.youtube.com/watch?v=9XQaxIw5y40 | 04:53 |
fenn | enough to confuse some old people at a tea shop in tokyo | 06:30 |
nsh | what's the chatter? | 09:29 |
nsh | kanzure might find this interesting: http://www.biologynews.net/archives/2008/10/13/caltech_biologists_spy_on_the_secret_inner_life_of_a_cell.html | 09:43 |
nsh | wait, wrong article | 09:43 |
nsh | http://www.biologynews.net/archives/2008/10/11/digital_zebrafish_embryo_provides_the_first_complete_developmental_blueprint_of_a_vertebrate.html | 09:43 |
nsh | better article from nature: http://www.nature.com/news/2008/081009/full/news.2008.1162.html | 10:26 |
nsh | journal article: http://www.sciencemag.org/cgi/rapidpdf/1162493.pdf | 10:29 |
nsh | kanzure_, you should also check out WikiPathways | 10:51 |
nsh | ( http://www.nature.com/news/2008/080903/full/455022a.html ) | 10:51 |
kanzure_ | digita zebra fish embryo :) heh | 12:46 |
* kanzure_ hasn't actually checked the links yet. | 12:46 | |
willPow3r_ | this biology class is teaching me how to hate biology | 15:35 |
-!- Irssi: #hplusroadmap: Total of 18 nicks [0 ops, 0 halfops, 0 voices, 18 normal] | 16:24 | |
kanzure_1_ | http://dpatoolkit.sourceforge.net/ design pattern automation toolkit | 17:22 |
fenn | adelaide sounds great for growing algae | 17:49 |
fenn | grows in seawater, cover ponds with plastic sheet to catch the evaporated water | 17:49 |
kanzure_1_ | http://doc.trolltech.com/3.1/qbytearray.html#qCompress | 18:04 |
kanzure_1_ | any ideas on what format that compression format that is? | 18:04 |
fenn | "Few [philosophers] were sufficiently correct that people have forgotten who discovered what they discovered." | 18:04 |
kanzure_1_ | libz yay | 18:06 |
fenn | qCompress merely uses the same compression algorithm as e.g. gzip, | 18:06 |
kanzure_1_ | the internets says it's "not quite gzip" | 18:09 |
fenn | probably missing some file formatting | 18:11 |
fenn | RFC 1952 | 18:12 |
* fenn yawns | 18:13 | |
kanzure_1_ | gunzip is telling me it's not gz.. | 18:17 |
kanzure_1_ | argh | 18:17 |
fenn | of course | 18:17 |
kanzure_1_ | forgot to debase64 it | 18:17 |
fenn | what are you trying to do anyway? | 18:17 |
fenn | did some fool qcompress() data and then stick it in an xml file? | 18:18 |
kanzure_1_ | yep | 18:19 |
fenn | "i've got to shrink these files somehow! i know, i'll compress the data!" | 18:19 |
kanzure_1_ | http://function.device.mst.edu/repositoryEntry/doxygen/d8/d12/class_x_m_l_module.html#07e801857b3ca8e1039f415ab27e0e65 | 18:19 |
kanzure_1_ | gunzip still hates it, even now after debase64ing it. | 18:20 |
kanzure_1_ | yay for qt being closed source | 18:20 |
fenn | it's not closed source | 18:20 |
fenn | it's just unreasonably large | 18:21 |
kanzure_1_ | I can go check the qt sources? | 18:21 |
ybit | yep | 18:21 |
kanzure_1_ | how is this not open. license crap? | 18:22 |
fenn | it's dual licensed | 18:23 |
fenn | so in practice forks are discouraged, although there's no reason one couldnt fork it | 18:23 |
fenn | too bad their docs dont link to the code | 18:24 |
kanzure_1_ | I was just wondering about that | 18:25 |
kanzure_1_ | it's rather useless to say "it compresses!" without telling me the format | 18:25 |
kanzure_1_ | http://lists.trolltech.com/qt-interest/2007-05/thread01239-0.html | 18:26 |
kanzure_1_ | nope, not zip either. | 18:27 |
kanzure_1_ | ldd /usr/lib/libQtCore.so.4.2.3 | grep libz libz.so.1 => /usr/lib/libz.so.1 (0x00002b6e465ad000) | 18:28 |
fenn | libz/gzip/zip are all the same algorithm, just different file formats | 18:28 |
kanzure_1_ | so there's a -qt-zip option you can throw at the configure file before the make. | 18:29 |
kanzure_1_ | *-qt-zlib | 18:29 |
kanzure_1_ | which makes it fairly obvious that it's using zlib. | 18:29 |
nsh | what you trying to recover? | 18:29 |
kanzure_1_ | nsh: http://heybryan.org/~bbishop/docs/repo/CDD/ one of these two zip files. I think the second one has the decoded version. | 18:29 |
kanzure_1_ | (from base64) | 18:29 |
nsh | no, what are they? | 18:30 |
kanzure_1_ | but for some reason it's also 'compressed' using some mystical qt method that isn't zip, isn't gzip, etc. | 18:30 |
kanzure_1_ | the files are 'CDD' files for ConceptDraw http://conceptdraw.com/ | 18:30 |
nsh | i meant | 18:30 |
nsh | WHAT ARE THEY | 18:30 |
kanzure_1_ | black box diagrams in an equivalent of SVG | 18:30 |
nsh | what is the thing that you want in terms of why you are interested in that particular pattern of ones and zeros | 18:30 |
fenn | serious ontological crisis | 18:30 |
fenn | "this is not a pipe" | 18:31 |
nsh | at the interface between your brain and the external bloody universe | 18:31 |
nsh | what do they represent?!? | 18:31 |
kanzure_1_ | black box diagrams | 18:31 |
kanzure_1_ | black box diagrams in an equivalent of SVG | 18:31 |
nsh | hmm | 18:31 |
fenn | nsh: consumer crap, mostly, and some obscure nasa subsystems | 18:31 |
nsh | ok, sounds boring | 18:31 |
kanzure_1_ | it is. | 18:31 |
kanzure_1_ | :( | 18:31 |
nsh | (btw, still can't access heybryan for some reason) | 18:32 |
kanzure_1_ | jaldfjkalfkjal; | 18:32 |
kanzure_1_ | anyone else? | 18:32 |
kanzure_1_ | can somebody whois kanzure_ ? | 18:32 |
kanzure_1_ | I can't because of the stupid alt key config | 18:32 |
kanzure_1_ | http://lists.trolltech.com/qt-interest/2006-07/thread00070-1.html <-- so it's zip but it's not zip | 18:33 |
nsh | kanzure_ is i=bbishop@66.112.232.108 * Bryan Bishop | 18:33 |
nsh | you should consider dyndns type arrangement | 18:34 |
fenn | heybryan.org is up for me | 18:35 |
kanzure_1_ | why did I bother asking for the /whois | 18:35 |
kanzure_1_ | I just uploaded those CDD.zip files | 18:35 |
nsh | oh | 18:36 |
nsh | i may have added a hosts entry for heybryan before | 18:36 |
kanzure_1_ | if I have to go write my own qt method to uncompress this stupid data: http://lists.trolltech.com/qt-interest/2006-08/msg00577.html | 18:38 |
kanzure_1_ | what really annoys me is that once I do uncompress all of this, | 18:38 |
fenn | hmm 5W 1TB disk at newegg.. | 18:38 |
fenn | $115 | 18:38 |
nsh | better | 18:38 |
kanzure_1_ | I'm going to have to go find the automated Windows XP app-clicker thingies | 18:38 |
kanzure_1_ | and then write a script to use the ConceptDraw software to convert to SVG | 18:39 |
kanzure_1_ | that wouldn't be terrible, normally | 18:39 |
kanzure_1_ | except that those are *SVG* with points and geometries drawing boxes and such | 18:39 |
kanzure_1_ | I'd have to write a script to convert from the SVG to something quantified | 18:39 |
kanzure_1_ | and even then, all of this data was hand-entered, making it mostly useless.. | 18:40 |
fenn | all of the data was hand entered anyway | 18:40 |
kanzure_1_ | you'd think so | 18:40 |
kanzure_1_ | but there's some products in the repository that do not have CDD files | 18:40 |
kanzure_1_ | thus the data set is incomplete | 18:40 |
kanzure_1_ | I mean to say that this was hand-entered separately, on top of everything else | 18:40 |
kanzure_1_ | there's only 94 out of the 125 products with these CDD files | 18:41 |
fenn | poor data integration | 18:41 |
kanzure_1_ | this is quickly turning into a nightmare. | 18:41 |
fenn | i dont know why you bother | 18:41 |
kanzure_1_ | do I get to call trolltech and yell at them? that'd be fun. | 18:41 |
kanzure_1_ | neither do I | 18:41 |
fenn | what will you do with a diagram anyway? | 18:42 |
fenn | and why is it not accessible anywhere? | 18:42 |
kanzure_1_ | don't understand the second question. | 18:42 |
fenn | there should be a web interface to get these diagrams | 18:43 |
kanzure_1_ | with the diagram I will convert it into a quantified, discrete format, and that's basically what we would call a very very basic set of interconnections of components in terms of 'flow type'. it's a black box diagram. so.. | 18:43 |
fenn | otherwise why did they bother putting them in the repo (if you can't get them back out) | 18:43 |
kanzure_1_ | hrm | 18:43 |
kanzure_1_ | ConceptDraw. | 18:43 |
fenn | but it's just a black hole | 18:44 |
kanzure_1_ | the entryapp allows you to click a button and save-as the CDD file | 18:44 |
fenn | ok so just do that | 18:44 |
kanzure_1_ | 94 times? | 18:44 |
fenn | yeah | 18:44 |
fenn | sorry if it offends your programmer instincts, but sometimes you have to choose your battles | 18:45 |
fenn | it's disappointing that i cant even find the qt source code | 18:50 |
kanzure_1_ | No, this is ridiculous. There's about 20 clicks to get to each CDD file. | 19:03 |
kanzure_1_ | This is not including the step where I repetetively translate from CDD to SVG .. | 19:04 |
kanzure_1_ | repetitively. | 19:04 |
kanzure_1_ | Not only 20 clicks but a giant pause in between while the repositoryapp "loads from the database" (AKA my file system ..) | 19:04 |
fenn | do you have the source for the data entry app? | 19:04 |
kanzure_1_ | I linked you to it | 19:05 |
kanzure_1_ | yeah | 19:05 |
kanzure_1_ | I'd rather just do a quick hack like http://lists.trolltech.com/qt-interest/2006-08/msg00577.html | 19:07 |
kanzure_1_ | I've spent 30 minutes on this? where does time go. | 19:07 |
fenn | computers are fantabulous devices capable of doing absolutely nothing for hours on end | 19:10 |
kanzure_1_ | s/computers/humans/ | 19:11 |
fenn | no, i'm sitting here in front of a computer giggling | 19:11 |
fenn | reading paul graham on philosophy: "There's a market for writing that sounds impressive and can't be disproven." | 19:11 |
kanzure_1_ | evil giggling? | 19:12 |
fenn | i'm not evil enough, this is more like schoolgirl-ish giggling | 19:13 |
kanzure_1_ | fenn: What was the alternative to dotty that you wanted me to try out for drawing the graphs? | 19:32 |
kanzure_1_ | neato | 19:32 |
fenn | and fdp | 19:36 |
fenn | fdp looks more "square" | 19:36 |
kanzure_1_ | fdp looks exactly the same. how weird. | 19:36 |
fenn | someone should make a map of TLA space | 19:38 |
fenn | see how many are used in each permutation | 19:38 |
fenn | 26x26x26 voxel cube | 19:38 |
kanzure_1_ | TLA? | 19:39 |
fenn | three letter acronym | 19:39 |
fenn | http://auction.meca.edu/wp-content/uploads/2008/09/goochh_artauctionimage.jpg | 19:40 |
fenn | a heretofore unconsidered construction material | 19:40 |
kanzure_1_ | I'm doing all of the graphs incorrectly. | 19:41 |
kanzure_1_ | the labels should be the nodes, the current nodes (the "verbs") should be the edges. | 19:41 |
* kanzure_1_ tests. | 19:42 | |
fenn | that would certainly cut down on the number of edges, but it would be totally pointless since you lose interconnectivity information | 19:43 |
kanzure_1_ | input, artifact, output such that i/o specifies another artifact instantation ID, and such that the input flow and output flow is labeled with a 'flow name' which is correlated to a unique artifact instantation ID. | 19:49 |
kanzure_1_ | gah that's hard to say / think about | 19:50 |
kanzure_1_ | yes, it is true that the flows do not connect to one another. the artifacts do however. and the flows are implicitly connected to the 'artifacts'. thus, the flows connect. | 19:50 |
kanzure_1_ | perhaps the biggest piece of evidence is this: | 19:53 |
fenn | Flows are not like some of the other grammatical modes which express precisely a mode of conception without any reality that corresponds to the conceptual mode, and consequently do not express precisely something in reality by which the intellect could be moved to conceive a thing the way it does, even where that motive is not something in the thing as such. | 19:54 |
kanzure_1_ | in the file where I generate the .dia files, I literally say: "$SubInputArtifact -> $currArtifact [label=flow input flow]" then the next line says "currArtifact -> SubOutputArtifact label outputflow" | 19:54 |
kanzure_1_ | so why not jdo flowinput -> flowoutput | 19:55 |
kanzure_1_ | *just do | 19:55 |
kanzure_1_ | connectivity will be maintained but there will be this ridiculous number of lines going into each of the flows since there's only like 8 types or something | 19:56 |
kanzure_1_ | unless I give each of them a unique ID. Then the problem becomes how to keep the connectivity correct with nodal ID assignments to each of the flows. Ah, the IDs will not be the flow IDs but rather the artifact IDs. This way, connectivity is preserved. | 19:58 |
kanzure_1_ | (there's a way to name a node and hide its name in dia, yes? I seem to recall something like this.) | 19:59 |
fenn | there will be approximately 8*number of artifacts | 19:59 |
fenn | edges | 19:59 |
kanzure_1_ | if you count the final graph, yes, but the dia/dot file will have overlapping arcs/edges. | 20:03 |
kanzure_1_ | http://heybryan.org/~bbishop/docs/repo/graphs/2008-10-14_flowgraph_78.png | 20:20 |
* fenn makes a wanking motion | 20:22 | |
wrldpc | what's the plan? | 20:29 |
kanzure_1_ | Context? | 20:38 |
gene | here's the plan[TM] http://www.geocities.com/SouthBeach/1380/plantxt.html | 20:41 |
fenn | the south beach diet plan | 20:41 |
wrldpc | what's the plane? | 20:42 |
gene | the plan | 20:42 |
wrldpc | oh | 20:42 |
gene | [tm] | 20:42 |
wrldpc | (TM(TM) | 20:42 |
gene | is to pave the earth flat | 20:43 |
gene | So that you can drive to wherever you need to go without turning or slowing down | 20:43 |
fenn | jetpacks for everybody | 20:49 |
gene | Why jetpacks? | 20:52 |
gene | can you drink beer and eat burgers while using a jetpack? I think not | 20:53 |
fenn | you just have a limited imagination | 20:54 |
gene | Besides you can't go that fast in a jet pack | 20:55 |
gene | why not join the future and use a rocket powered hypercar to get from place to place | 20:56 |
gene | besides, rocket powered hypercars use more scarce natural resources and create more pollution than jetpacks | 20:58 |
gene | Don't you want to be able to look at the sky and say "I made the sky what it is today"? | 20:59 |
fenn | i do, but dont see what cars have to do with it | 20:59 |
gene | http://www.geocities.com/SouthBeach/1380/pave.html | 21:00 |
fenn | it wasn't funny to begin with | 21:00 |
gene | it's an old meme | 21:01 |
gene | a very old meme | 21:01 |
kanzure | fenn: Thanks for the heads up on http://www.newegg.com/Product/Product.aspx?Item=N82E16822152102 that 1 TB $115 hdd. | 23:31 |
fenn | i'd probably get one but no money | 23:46 |
fenn | it's not like hard drives are going anywhere | 23:46 |
willPow3r | you're not cool unless you have one. | 23:50 |
willPow3r | sorry. | 23:50 |
kanzure | You're not a man unless you have one. | 23:51 |
willPow3r | well i have 50 | 23:51 |
willPow3r | so thats a testament to how cool and manly i truly am | 23:51 |
willPow3r | well, by irc standards anyway | 23:53 |
willPow3r | which i'm sure will translate to 2nd life | 23:53 |
willPow3r | wow, 76 F in mid october? its so freaking hot here | 23:54 |
kanzure | You're not a man unless 80 degrees is abnormally cold where you live. | 23:55 |
fenn | i always thought colder climates were more manly | 23:55 |
fenn | need fur and blubber and tusks | 23:56 |
willPow3r | i think an african zulu tribesman would beat an inuit in a hand-to-hand match | 23:56 |
fenn | nah zulu get their reputation because of battle tactics | 23:57 |
willPow3r | then again, i think it would be easier to adapt to a warmer climate than to a colder one | 23:57 |
kanzure | The warm clients suck. | 23:57 |
kanzure | Oh, climates? | 23:57 |
fenn | shaka zulu basically reinvented the phalanx: http://en.wikipedia.org/wiki/Shaka#The_famous_.22buffalo_horns.22_formation | 23:58 |
willPow3r | so zulu > eskimo | 23:58 |
willPow3r | at least in physical dominance | 23:59 |
fenn | no, faggot | 23:59 |
willPow3r | but give an eskimo and a zulu an education and i'm sure the eskimo would come out on top | 23:59 |
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!