ExplanationThe Code of Federal Regulations is a codification of the general and permanent rules published in ...
07:32 < fenn> maybe it's a schema for legal documents
07:34 < fenn> ugh navigating legalxml.org is like trying to get to the ISO standards documents
07:35 -!- cluckj [~cluckj@cpe-24-92-63-104.nycap.res.rr.com] has joined ##hplusroadmap
07:36 < fenn> not that it helps anything, but i guess it's this: http://www.gpo.gov/fdsys/bulkdata/CFR/resources/CFRMergedXML.xsd
07:37 < fenn> i have no opinions on xml to json conversion. maybe parse it into a tree with your favorite parser and dump with your favorite dumper?
07:37 < FourFire> kanzure, my point is asking being, that depending on the type of dog noise, it could very well be normal for crows to make it
07:37 < dcary> fenn: Yeah, even though almost all my circuits run at 5 V or less, I don't think I've ever used less than 6 mil spacing (about 0.15 mm spacing).
07:37 < dcary> fenn: Yeah, it's pretty linear -- see p. 39 of http://www.the-bao.de/divers/ipc2221.pdf for details.
07:41 < fenn> interesting document
07:43 < fenn> a common problem i have is determining the name of a particular common connector
07:45 < fenn> there are a relatively small number of IC packages, but it seems like an almost infinite number of connectors and very easy to order the wrong one
07:49 < dcary> fenn: It seems that most PCB fabs now claim they can make boards with 4 mil trace / 4 mil space (which is 0.1 mm trace / 0.1 mm space). Picking a random manufacturer from my list, Cirexx claims they can do flex circuits with 0.003 inch trace/gap.
07:51 -!- ebowden [~ebowden@CPE-60-231-177-134.lns4.dav.bigpond.net.au] has quit [Remote host closed the connection]
07:51 < gradstudentbot> His lab is so awkward.
07:53 < fenn> if the speed of sound in water is 1484m/s then one wavelength at 10MHz is 0.1484mm so i guess it's possible that we'll need that level of spacing
07:54 < fenn> i say "if" because it changes
07:55 < fenn> "In salt water that is free of air bubbles or suspended sediment, sound travels at about 1560 m/s."
07:58 < cluckj> if you're using a small enough tank size, the wavelength changes at (maybe) insignificant digits
08:00 < fenn> anyway it's a lot smaller than radio waves of the same frequency
08:02 < dcary> fenn: Do we really need so many different kinds of connectors? Bits are bits, right? https://xkcd.com/927/
08:02 < fenn> uh what "Floating point numbers are not allowed in canonical JSON."
08:04 < fenn> i wonder what that grammar representation format is called
08:04 < fenn> int: digit digit1-9 digits - digit1-9 - digit1-9 digits
08:04 < fenn> that didn't paste correctly
08:07 < fenn> dcary: not all connectors are used for transferring bits
08:08 -!- marciogm [~marciogm@179.126.71.80] has joined ##hplusroadmap
08:08 < fenn> i do wish we had stopped at mini USB 2.0 though
08:09 < juri_> mini usb 3 is weird. :)
08:09 < fenn> also completely pointless
08:09 < fenn> just use sata
08:10 < fenn> or ethernet. we should have just invented a tiny power over ethernet connector for TTL voltages
08:10 < fenn> wtf is USB, an abomination
08:11 < dcary> fenn: Yeah, the format at http://wiki.laptop.org/go/Canonical_JSON is apparently some informal variant of Extended Backus–Naur Form (EBNF). What is the name for the kinds of pretty grammar diagrams used at http://json.org/ ?
08:11 < fenn> railroad diagrams?
08:11 < fenn> i meant the format used to specify EBNF grammars in json
08:12 < fenn> there was this "kwalify" thing for validating yaml files, but it was kinda wonky
08:13 < fenn> can we just call a slash a slash, "solidus" pff
08:15 < dcary> fenn: Thanks. I learned a new term today, for something I've been looking at for decades: "railroad diagram" http://en.wikipedia.org/wiki/railroad_diagram
08:16 < fenn> Some of the popularity of the JSON data interchange format is due to its representation in railroad diagrams.[citation needed] CITATION NEEDED
08:17 < dcary> fenn: people use cables to carry power (generally in only one direction) and bits (often in both directions). What else are connectors used for?
08:18 < mosasaur> I know all this talk about standards and data formats and circuit boards and chips makes you feel special and mathy engineeric but to me it's just like the discussion is stranded into some morassy lowland where it's impossible to spread my wings.
08:19 < fenn> analog signals, modulated power, AC power, DC power, blood of the innocent dinosaurs
08:19 <@ParahSailin> one might reasonably expect glob.glob in python to attempt to give lexicographically ordered output
08:19 <@ParahSailin> one would be wrong
08:21 < fenn> somehow we've managed to stick to 120V 60Hz sinusoidal for 100 years
08:21 < fenn> at least in the US
08:21 < cluckj> thanks technological somnambulism
08:22 < fenn> it's because houses were never designed to be taken apart
08:23 < fenn> tesla wished he could have switched to 1KHz
08:23 < fenn> [CITATION NEEDED]
08:24 < cluckj> lol
08:24 < FourFire> fenn I've read your citation
08:24 * cluckj drains FourFire of blood and summons nikola's ghost
08:24 < FourFire> he regretted making AC at 60Hz because he made that better
08:25 < cluckj> OH GREAT GHOST PLEASE REMEDY THIS INTERNET CITATION ATTEMPT
08:26 -!- nsh [~nsh@wikipedia/nsh] has quit [Ping timeout: 240 seconds]
08:27 -!- nsh [~nsh@wikipedia/nsh] has joined ##hplusroadmap
08:28 < kanzure> fenn: instead of parsing xml, i was thinking something like "take any word that has a certain frequency less than some threshold i pick, such that sentences/phrases say unique technology names, and the rest is skipped"
08:45 < dcary> Yeah, I heard that all the early designs for the international space station used 20 kilohertz power distribution because it is technically superior to other alternatives considered -- DC, 50 Hz, 60 Hz, 400 Hz. The rumor is that the ISS switched to DC power distribution for "political reasons".
08:46 < mosasaur> With the coming solar power era maybe we'd be better off with low voltage DC.
08:51 -!- nmz787_i [~nmccorkx@134.134.139.70] has joined ##hplusroadmap
08:53 < fenn> i get 20KHz and 50/60Hz, but why 400Hz?
08:53 < fenn> oh. "Every time you fly in commercial airliners, the 400HZ power produced by the alternators on each engine"
08:56 < fenn> some day we'll have a wall with a matrix of plugs sorted by V and F
08:58 < fenn> and auto-retracting cables
08:59 < kanzure> and the wall will laugh at you
08:59 < mosasaur> devices will become smaller and use less power and will be closer to our neurons
09:00 < fenn> we'll exchange data as puffs of coded DNA, carried by radiation pressure from lasers in our eyes
09:01 < mosasaur> wait, are you talking about sex?
09:01 < fenn> woah, dude, don't be gay
09:02 < fenn> it's only sex if your gametes touch
09:02 < kanzure> not according to nevada law
09:03 < fenn> i defer to bill clinton
09:03 < fenn> excuse me, the 42nd president of the united states of america
09:03 < cluckj> lol
09:05 < fenn> what does "is" mean anyway
09:05 < kanzure> if eleitl is asking for 2 spikerboxes from me, does that mean he probably doesn't have that equipment already
09:05 < fenn> "if 'is' means is and never has been, that is not--that is one thing. If it means there is none, that was a completely true statement..."
09:06 < kanzure> i find it hard to believe that he legitimately wants an introductory kit itself
09:09 < fenn> does the spikerbox actually do anything or is it just a glorified audio connector?
09:09 < mosasaur> fenn: let's just stick to laser encoded data streams, I mean it even works for high bandwidth satellite communication.
09:09 < kanzure> comes with some silly ipad app i think
09:10 < fenn> mosasaur: never underestimate the bandwidth of a station wagon full of sperm hurtling down the jupiter brain
09:14 -!- senior [senior@m90-129-71-30.cust.tele2.se] has joined ##hplusroadmap
09:15 < fenn> hoela senior
09:15 -!- senior [senior@m90-129-71-30.cust.tele2.se] has left ##hplusroadmap []
09:15 < kanzure> you overwhelmed him
09:16 < fenn> should i have said "Hej då!"
09:16 < kanzure> yes
09:16 < mosasaur> everything moves at light speed anyway, it's just that some things have more of a time like motion vector.
09:17 < gradstudentbot> If I write this paper, then maybe I can use that as my thesis?
09:17 < fenn> temporal wormholes and time-like mazes, a thesis submitted by gradstudentbot to fulfill the requirements of the advanced institute of internet studies
09:17 < gradstudentbot> I don't think my PI remembers me.
09:19 < fenn> huh "temporal wormhole" isn't a trope?
09:20 < kanzure> only an arxiv.org trope
09:20 < kanzure> were you around when arxiv.org announced their new "business model"?
09:21 < fenn> -_-
09:21 < fenn> in the future, just don't tell me horribly depressing things
09:22 -!- nmz787_i [~nmccorkx@134.134.139.70] has quit [Quit: Leaving.]
09:22 < kanzure> it's not as bad as you think
09:22 < fenn> haven't we learned yet not to build centralized systems?
09:22 < kanzure> parent university didn't want to fund it on their lonesome, so now some additional institutions are agreeing to chip in
09:23 < kanzure> i don't think the particle physics people have ever not built a centralized system
09:23 < kanzure> (http was client-server, no "he invented the interwebs!")
09:24 < kanzure> *no "but he invented
09:24 < fenn> i never really understood the original use case for http
09:24 < fenn> was a web server supposed to be like a wiki?
09:24 < kanzure> maybe it was a reaction to gopher stuff
09:26 < kanzure> i should make gwern analyze the commerce control list, he likes that sort of mindless data handling, right?
09:27 < fenn> http://info.cern.ch/hypertext/WWW/Proposal.html where did the "stored information" come from?
09:28 < fenn> gosh i still have this problem all the time. "once you found out that the name of Joe Bloggs is listed in an incomplete description of some on-line software, it is not straightforward to find his current electronic mail address."
09:30 < kanzure> this is why you must practice the art of stalk and lurk
09:30 < fenn> obviously what happened was that Time Spammers invaded CERN in 1988.1 and inserted their confounding protocol into the HTTP spec
09:31 < kanzure> makes sense, time travelers probably prefer particle accelerator facilities
09:32 < fenn> they use the temporal monopoles generated by bogon-bogon collisions
09:32 < xmj> kanzure: i'm curious. do you know what gwern actually does with his life?
09:33 < kanzure> evidence suggests reading and writing, and not much else.
09:33 < kanzure> i haven't seen him do much other than that
09:33 < fenn> "I am a freelance writer & researcher. I have worked for or published in MIRI3 (formerly SIAI), CFAR, A Global Village, Cool Tools, Quantimodo, New World Encyclopedia, Bitcoin Weekly, Mobify, Bellroy and private clients"
09:34 -!- augur [~augur@216-164-48-148.c3-0.slvr-ubr1.lnh-slvr.md.cable.rcn.com] has quit [Quit: Leaving...]
09:35 < kanzure> eventually he will be doxed
09:36 < fenn> "I’ve always preferred to work on existing applications and libraries than to go write my own."
09:36 < kanzure> ah, so double NIH
09:36 < fenn> an interesting viewpoint
09:36 < kanzure> it's because i can't remember which way NIH is supposed to work
09:37 < kanzure> am i supposed to prefer my own stuff, or the stuff of others? and if i do either one, which one confers NIH syndrome upon me?
09:37 < fenn> put billions of taxpayer dollars in, get patents out
09:37 < JayDugger> NIH syndrome?
09:37 < kanzure> .wik NIH syndrome
09:37 < yoleaux> "Not invented here (NIH) is the philosophy of social, corporate, or institutional cultures that avoid using or buying already existing products, research, standards, or knowledge because of their external origins and costs." — http://en.wikipedia.org/wiki/NIH_syndrome
09:38 < fenn> NIH syndrome was discredited when it was revealed that the originator of the concept had a conflict of interest
09:38 < cluckj> lol
09:38 < JayDugger> Nat'l Institutes of... never mind
09:38 < xmj> NIH is when you reinvent the wheel because you didn't do it yourself.
09:38 < JayDugger> .wik goo
09:38 < yoleaux> "Disambiguation: Goo" — http://en.wikipedia.org/wiki/Goo
09:38 < JayDugger> That's new.
09:38 < kanzure> huh? "NIH is when you reinvent the wheel because you didn't do it yourself."
09:38 < JayDugger> It might have been the National Institutes of Health.
09:39 < fenn> NIH is when you didn't invent the wheel because somebody else invented it first
09:39 < gradstudentbot> So, there's this really good conference in Spain that I want to attend.
09:39 < kanzure> gradstudentbot: have you submitted the abstract to the paper?
09:39 < gradstudentbot> Are you ever going to publish that?
09:39 < JayDugger> .wik help
09:39 < yoleaux> "Disambiguation: Help" — http://en.wikipedia.org/wiki/Help
09:39 < fenn> .help
09:39 < yoleaux> fenn: I'm yoleaux. Type .commands to see what I can do, or see http://dpk.io/yoleaux for a quick guide.
09:40 < fenn> huh
09:40 < fenn> .commands
09:40 < yoleaux> Commands are divided into categories: services, general, api, demos, admin. Use .commands to get a list of the commands in each.
09:40 < fenn> did i seriously not try that
09:40 < gradstudentbot> Does this look contaminated to you?
09:41 < JayDugger> Found it. Thanks.
09:41 < JayDugger> .commands services
09:41 < yoleaux> Commands in services: acronym, add-command, command-help, del-command, dety, geo, leo, moon, ngrams, nokiageo, o, r2r, roll, rot13, shipping, suggest, swhack, thesaurus, title, tw, twho, weather, yi. Use .help to get information about them.
09:42 < mosasaur> .moon
09:42 < yoleaux> Waxing Crescent (0.074)
09:42 < fenn> .geo 1 times square
09:42 < yoleaux> fenn: One Times Square, New York, NY 10036, USA at 40.756,-73.986 to wit http://google.com/maps?q=40.756,-73.986
09:43 < JayDugger> .g spikerbox
09:43 < yoleaux> https://backyardbrains.com/products/spikerbox
09:44 < JayDugger> Ah. Silly me.
09:44 < kanzure> NIH: never invented haha
09:45 < JayDugger> Ding! a new error code for skdb.
09:45 < JayDugger> sudo skdb make -me -a nuclear_pulse_rocket returns NIH.
09:46 < fenn> --me
09:46 < kanzure> don't one of you owe me a technology tree
09:47 < JayDugger> Fair enough,
09:47 < fenn> .ngrams effete
09:47 < yoleaux> :(
09:47 < fenn> .ngrams effete snob
09:47 < yoleaux> :(
09:47 < fenn> .help ngrams
09:47 < yoleaux> Compare the frequency of words/phrases in an n-grams database
09:48 < fenn> that doesn't help
09:48 < fenn> .ngrams "effete snob", "jaded wanker"
09:48 < yoleaux> :(
09:49 < kanzure> fenn: what's missing from http://pinboard.in/ ?
09:49 < fenn> ads?
09:49 < fenn> fluff?
09:50 < kanzure> i mean for tagging
09:50 < fenn> well it's not part of ikiwiki...
09:51 < kanzure> do you really want all of your tagged data to be based on ikiwiki's compiler engine thing
09:51 < fenn> why should i trust "one dude in his underpants" when i don't even trust google to stay around for ten years
09:51 < kanzure> nah i don't mean "use this service exactly as it is"
09:52 < kanzure> what's wrong with tagging things in postgresql? table schema: id, timestamp, tag id, object id
09:52 < fenn> see http://gwern.net/Google%20shutdowns etc
09:52 < fenn> because i hate databases
09:52 < fenn> tables is all backwards/inside out from normal thinking
09:52 < kanzure> why is it geometrically wrong?
09:52 < fenn> database people just don't see how backwards it is because they have internalized the model
09:52 < fenn> or shall i say, the model has internalized them
09:52 < JayDugger> .g stockholm syndrome
09:53 < yoleaux> http://en.wikipedia.org/wiki/Stockholm_syndrome
09:53 < catern> i've never internalized the model
09:53 < catern> please inoculate me before I do so
09:53 < fenn> would you rather have a graph or an adjacency matrix?
09:54 < JayDugger> Oh, awesome. I can make snide comments and the bot drops links. :)
09:54 < fenn> hint: adjacency matrices are ugly and impossible to visually parse
09:54 < catern> g-graph
09:54 < JayDugger> Won't improve my humor, but I'm happy.
09:54 < catern> but that's a type error
09:54 < catern> adjacencey matrices aren't the same things as graphs
09:54 < catern> but go on
09:55 < catern> (they're representations of graphs)
09:55 < fenn> uh, i dont know what you're saying
09:56 < catern> well you were going to say why databases are bad, please continue
09:56 < gradstudentbot> Yeah, I couldn't repeat that.
09:56 < gradstudentbot> Yeah, but that was only a sample size of one.
09:57 < kanzure> you can do graphs with tables
09:57 < fenn> databases are all set up for the computer, not for humans. so if you need to go in and fix something, or add an entry, all of a sudden you're editing 15 different tables and saying "is this a unique key" and looking up your object relational model
09:58 < fenn> whereas with a yaml file, the schema is right there in every entry, and you never get a slice of an object
09:58 < fenn> i know slice is the wrong word but i dont know the right word
09:59 < fenn> i mean really, look at this crap edu/dropping-all-the-foreign-keys-in-your-sql-server-database/
09:59 < kanzure> mongodb subscribes to that model, but it doesn't matter because you can just use a json column in postgresql anyway
09:59 < fenn> http://mafudge.syr.edu/dropping-all-the-foreign-keys-in-your-sql-server-database/
10:00 < fenn> can you look at that diagram and figure out "has a" or "is a" relationships?
10:00 < kanzure> yes
10:00 < fenn> is there anything showing the existence of "employee" besides the prefix?
10:00 < kanzure> the only times i've had problems was when i encountered 50+ tables/models that had broken/wrong relationships between them
10:01 < kanzure> oh you're right, departments is not defined correctly
10:01 < kanzure> well they are just morons
10:02 < fenn> my point exactly
10:02 < kanzure> i can't help them if their models suck
10:02 < kanzure> their models would suck anyway even if it was a pile of json or yaml
10:03 -!- pyotra [~asakharov@24.60.79.55] has joined ##hplusroadmap
10:08 < kanzure> http://journalprices.com/
10:08 < kanzure> "Use this search engine to find internationally-published journals and rank them by price per article or citation. Here are some summary statistics for this edition. If you wish, you can also download an Excel spreadsheet that contains all of our data. You can find explanations of our data sources and methods at this link."
10:10 < fenn> i'm not a database expert and i don't really feel like explaining why i hate them, so here, have a thing: http://whydoeseverythingsuck.com/2008/02/death-of-relational-database.html
10:11 < kanzure> well, i don't think it was ever about knowledge in the first place
10:11 < gradstudentbot> Nobody has tried this before.
10:12 -!- sapiosexual [~sapiosexu@d75-156-89-72.bchsia.telus.net] has joined ##hplusroadmap
10:15 < kanzure> you should try mongodb (mongo_engine or pymongo) and neo4j
10:15 < fenn> look i hate databases okay
10:15 < kanzure> and one of those python mongo modelers
10:16 < kanzure> mongodb is just a key value store where the values are json documents
10:16 < kanzure> schemaless
10:16 -!- mosasaur [~mosasaur@178.227.77.234] has quit [Quit: Leaving.]
10:17 < fenn> what's wrong with a filesystem? a filesystem is a database, but i don't have to use some shitty interface that has no tools and no common conveniences like tab completion or whatever
10:17 < catern> i like filesystems
10:17 < fenn> i get that mongodb is just a big dict
10:18 < kanzure> d3vz3r0: are you around?
10:18 < fenn> i get that facebook and google have too many users and they have to resort to fancy tricks
10:18 < fenn> but 98% of the web would be just fine with flat files and a version control system like git
10:18 -!- delinquentme [~dingo@74.61.157.78] has quit [Ping timeout: 240 seconds]
10:19 < d3vz3r0> yo
10:20 < d3vz3r0> kanzure: ^^^
10:20 < kanzure> d3vz3r0: there is some backlog stuff, but basically fenn is wondering why mongodb instead of just git plus flat files
10:20 -!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap
10:20 < kanzure> or just a generic filesystem
10:21 < fenn> you can even index a filesystem, see "tracker-search" or "namazu" for example
10:21 < kanzure> .wik namazu
10:21 < yoleaux> "Disambiguation: Namazu" — http://en.wikipedia.org/wiki/Namazu
10:21 < d3vz3r0> sure you can do that, but fs's aren't really built for that kind of access pattern
10:21 < d3vz3r0> you can certainly do it, things like hotmail were built entirely on a fs
10:21 < d3vz3r0> seriously, crazy shit
10:22 < d3vz3r0> the issue really is the IO overhead
10:22 < fenn> why is that any different in a database
10:22 < d3vz3r0> there's a reason mongo uses one big ass file (or multiple 'tablespace' files)
10:23 < d3vz3r0> the IO overhead of the filesystem itself
10:23 < d3vz3r0> ie: opening 10k files takes a long time
10:23 < fenn> why
10:23 < fenn> they are just pointers
10:23 < d3vz3r0> inodes?
10:23 < fenn> right
10:23 < d3vz3r0> sure, to locations on the disk
10:23 < d3vz3r0> which it has to seek to
10:24 < fenn> or in memory
10:24 < d3vz3r0> if you mmapped the file
10:24 < xmj> fenn: why do you need git when you have a journaled filesystem
10:24 < d3vz3r0> but that file has to be read into memory at some point
10:24 < fenn> if you database is 1TB it's not going to be in RAM either
10:24 < xmj> well you could implement some gitlike tool at the FS level :)
10:24 < d3vz3r0> what are you trying to do?
10:24 < kanzure> d3vz3r0: unrelated, but http://kentonv.github.io/capnproto/ (instead of regular protobufs)
10:25 < fenn> xmj: multiple versions, distributed editing, and journals weren't intended to be rolled back very often
10:25 < gradstudentbot> I have this really good idea. I just can't get it to work.
10:25 < xmj> fenn: git rebase.
10:25 < fenn> xmj: i'd love to have a git filesystem
10:25 < xmj> heh
10:25 < xmj> i like ZFS.
10:26 < JayDugger> fenn: How does git-annex fail for your purposes?
10:26 < gradstudentbot> The culture got contaminated.
10:26 < d3vz3r0> what do you want from a git filesystem? I'm sincerely curious
10:26 < xmj> fenn: ever tried ZFS on some box?
10:26 < kanzure> d3vz3r0: originally he was asking why databases (including mongodb) are relevant or interesting
10:27 < fenn> d3vz3r0: versioning? branch, merge, etc. all the reasons people use version control
10:27 -!- delinquentme [~dingo@74.61.157.78] has joined ##hplusroadmap
10:27 < xmj> i'm sure you could implement that with zfs
10:27 < fenn> kanzure: don't put words in my mouth, i was trying (halfheartedly) to explain why i hate databases
10:28 < kanzure> well you brought up models, but models can be wrong on the file system just as they can be wrong in a db
10:28 < fenn> apple has a "time machine" thingy, so obviously the demand is there
10:28 < JayDugger> .g git-annex
10:28 < yoleaux> https://git-annex.branchable.com/
10:28 < fenn> the only reason we need git-annex is because we don't have a git filesystem
10:29 < xmj> what exactly does apple's time machine do?
10:29 < fenn> lets you roll back to before you broke stuff
10:29 < xmj> ah
10:29 < fenn> it also does backups
10:29 < JayDugger> I don't know git-annex well, which prompted my question.
10:29 < xmj> zfs rollback <3
10:29 < d3vz3r0> +1 for zfs
10:30 < d3vz3r0> fenn: I sometimes wishing i had a git fs too, but i ask myself how I would use it
10:30 < fenn> if a filesystem isn't made to quickly open lots of little files, then maybe you should use a different filesystem
10:31 < xmj> zfs has one caveat: it eats ram.
10:32 < fenn> see i can implement stupid inside-out database stuff in a filesystem too; just take the n line of every file and put that in its own file, then when you need to reconstruct the original file "just" do cat * | head -n $n | tail -n1 >> the_file
10:32 < d3vz3r0> and performance will suck
10:32 < fenn> of course it will suck, because it's a stupid idea
10:32 < fenn> that's why i hate databases!!!
10:33 < d3vz3r0> you hate databases because filesystems would be really slow at doing similar things?
10:33 < d3vz3r0> I don't understand
10:33 < xmj> he just doesn't like the idea of tables.
10:33 < fenn> because they're inside out and don't let you get at the thing itself
10:33 < d3vz3r0> the 'thing' itself?
10:33 < d3vz3r0> which is what>
10:33 < xmj> fenn: did you fail accounting once and start hating excel?
10:33 < d3vz3r0> ?
10:33 -!- nmz787_i [nmccorkx@nat/intel/x-hzixdmkkksqeyzeb] has joined ##hplusroadmap
10:34 < d3vz3r0> I don't like the idea of organized and structured data either
10:34 < fenn> in object oriented programming the "thing" is an object
10:34 < d3vz3r0> yes
10:34 < d3vz3r0> I guess that oculd be true
10:34 < xmj> there are more programming paradigms than OOP.
10:34 < d3vz3r0> or the "thing" is an association of objects, thing is pretty broad
10:35 < fenn> in functional programming the "thing" could be a subtree
10:35 < xmj> yay
10:35 < d3vz3r0> in the unverise "thing" could be the solution of quantum gravity.... can you use a better term than "thing"?
10:35 < xmj> lists!
10:35 < d3vz3r0> lists are just flat trees
10:35 < d3vz3r0> :)
10:35 < kanzure> flat graphs
10:36 < fenn> 1 = 1 in natural units
10:36 < d3vz3r0> sure
10:36 < fenn> -_-
10:36 * xmj puts d3vz3r0 into a cons.
10:36 < d3vz3r0> ha
10:36 < xmj> (/dev (/zero .))
10:37 < fenn> a row in a table is not usually "the thing" you want, it's part of it, or the thing is part of it
10:37 < d3vz3r0> you want the record
10:37 -!- Shehrazad [~Shehrazad@95.5.116.22] has joined ##hplusroadmap
10:37 -!- Shehrazad [~Shehrazad@95.5.116.22] has quit [Changing host]
10:37 -!- Shehrazad [~Shehrazad@unaffiliated/shehrazad] has joined ##hplusroadmap
10:37 < catern> camlistore is interesting on the note of a "git filesystem"
10:37 < d3vz3r0> which is inside the row?
10:38 < xmj> clearly
10:38 < xmj> set theory and sigma algebras to the rescue
10:38 < catern> it's totally immutable and uses garbage collection instead of deletion
10:39 < fenn> i don't know what a record is
10:39 < xmj> an element of a set ?
10:39 < d3vz3r0> ^^
10:40 -!- ElixirVitae [~Shehrazad@unaffiliated/shehrazad] has quit [Ping timeout: 252 seconds]
10:40 < fenn> Searching the 7IWD2-02-Foundation-2011-12.pdf, with a date of 2011-12-21 reveals that the word row appears 2277 times in the document while the word record appears only 21 times, either as the verb "record" or in some appendices in the end, in specifications of the data type correspondences for SQL data types and host language types
10:40 < d3vz3r0> in SQL terms a set of records is returned from a query
10:40 < d3vz3r0> each record is contained inside a row
10:40 < d3vz3r0> that record may or may not be the entire 'row'
10:40 < fenn> a record is just a subset of a row?
10:40 < d3vz3r0> at a 10k foot view, yes
10:41 < fenn> what do i call a set of records from multiple rows in multiple tables?
10:41 < kanzure> joined record, iirc
10:42 < d3vz3r0> the record is the combination of the joined data set from the multiple tables
10:43 < fenn> aren't people always going on and on about how evil joins are and how to minimize joins and oh geeze if only we had been smart enough to not require a join
10:43 < d3vz3r0> joins aren't evil
10:43 < d3vz3r0> they are just difficult in a distributed system
10:44 < fenn> why is it always the database server that's the weak link in a distributed system?
10:44 < xmj> and may be computationally expensive if tables consist of "big data"
10:44 < d3vz3r0> not requiring a join means perfectly partitioning your data set
10:44 < xmj> d3vz3r0: in the past we did denormalization to our datamodel to avoid joins and rice it.
10:44 < gradstudentbot> I am writing the abstract.
10:45 < xmj> obviously a data model of third normal form will be slower than a more redundant, higher normal formed data model
10:45 < xmj> it's a tradeoff
10:45 < d3vz3r0> yea, denormalization requires more joins
10:45 < fenn> you know what i call that? a bad caching system
10:46 < xmj> ?
10:46 < d3vz3r0> why is that a bad caching system?
10:46 < xmj> what do those two have to do with each other?
10:46 < fenn> all the denormalizing stuff, you're just rewriting the data different ways to speed up performance
10:46 < kanzure> are databases really the point of failure? i thought it was stuff like service discovery
10:46 < fenn> caching is storing answers so you don't have to recompute/fetch them again
10:47 < d3vz3r0> seems simple doesn't it?
10:47 < xmj> youre missing the point
10:47 < xmj> even with caching the generation of the first answer will differ
10:47 < xmj> generation time*
10:47 < d3vz3r0> caching is reliant on determinism
10:48 < xmj> yep
10:48 < fenn> isn't denormalization also?
10:48 < d3vz3r0> without deterministic results from a question you ask a system, caching is impossible
10:48 < d3vz3r0> no
10:48 < xmj> btw- there was a django CVE lately in which caching lead to an attacker being able to do a CSRF.
10:48 < xmj> which was that again.. /looks
10:48 < d3vz3r0> 2nd hardest problem in CS: caching
10:48 < fenn> but don't you have to update all the different copies in a distributed database when you modify a denormalized table?
10:49 < d3vz3r0> fenn: have you read the Dynamo paper from amazon?
10:49 < xmj> ?
10:50 < xmj> even when your db is normalized enough, changes will have to traverse your distributed system
10:50 < d3vz3r0> yea, he's talking about eventual consistency
10:50 < d3vz3r0> totally different thing
10:50 < fenn> while it's called a "hash table", as a distributed system it's not really a table in my understanding
10:51 < xmj> d3vz3r0: i hate when people talk about different things
10:51 < d3vz3r0> a "hash table" is a distributed system?
10:51 < fenn> (i'm reading about dynamo)
10:51 < fenn> a distributed hash table
10:51 < xmj> people talking past each other happens so frequently and is sooo expensive.
10:51 < d3vz3r0> yep
10:51 < fenn> as i said earlier, i'm not a database expert
10:51 < fenn> "denormalizing" is a stretch
10:51 < d3vz3r0> fenn: I think you are getting terms confused
10:52 < fenn> "denormalization is the process of attempting to optimize the read performance of a database by adding redundant data or by grouping data
10:52 < xmj> denormalizing is when you take a normalized data model with next to no redundancy and reduce the amount of tables, adding redundant parts.
10:52 < xmj> exactly
10:52 < fenn> presumably this is so you don't have to do a join across tables across database servers
10:52 < d3vz3r0> typically for perforamnce rasons
10:52 < d3vz3r0> no
10:52 < d3vz3r0> well, yes, but not always
10:53 < xmj> why would you join across database servers ?
10:53 < xmj> apart from using a foreigh table sorta import?
10:53 < fenn> i dont know
10:53 < xmj> fenn: i'd keep a few masters and a few readonly slaves
10:53 < xmj> but then again I'm no DBA and i'd be better off letting someone do that who really knows.
10:53 < kanzure> or redis replication sets/groups where they figure that out on their own
10:53 < d3vz3r0> he means distributed joins within a sharded SQL engine I think
10:53 < fenn> that centralization is your bottleneck
10:53 < xmj> ahh ok
10:54 < xmj> i dont know much about sharding
10:54 < xmj> havent worked with a level of data that made it necessary. our "big data" wasn't so big back then
10:54 < d3vz3r0> butyou know how sets work :)
10:54 < xmj> well, yes
10:54 < d3vz3r0> it's all just set theory
10:55 < fenn> how is denormalizing not a cache
10:55 < d3vz3r0> simple :)
10:55 < d3vz3r0> haha
10:55 < xmj> even with a sharded db, it should be invisible to the querying client, no?
10:55 < dingo> d3vz3r0 makes shit up all the time don't listen to him
10:55 < xmj> well maybe in slightly slower read speeds
10:55 < d3vz3r0> in theory, yes
10:55 < d3vz3r0> fuck you dingo
10:55 < d3vz3r0> :)
10:55 < d3vz3r0> just kidding buddy ;)
10:55 < dingo> thats mr. fucker to you
10:56 < dingo> hey look forward to seeing you later today right
10:56 < d3vz3r0> I think you got promoted to Dr. right?
10:56 < d3vz3r0> yep, I'll be there tonight
10:56 < dingo> who, me? naw
10:56 < kanzure> going away party?
10:57 < gradstudentbot> Who got mustard on my cell culture?
10:57 < d3vz3r0> something like that
10:57 < xmj> d3vz3r0: whats your trade?
10:57 < d3vz3r0> hookers
10:57 < kanzure> approximately true
10:57 < d3vz3r0> :)
10:57 < xmj> a pimp that idles on freenode.
10:57 < xmj> nice!
10:58 < d3vz3r0> my trade is computers
10:58 < kanzure> his trade is models.
10:58 < xmj> thats good, i enjoy the View of a good Model.
10:58 < d3vz3r0> yea, that's probably more accurate
10:58 < xmj> I really want to Control them.
10:58 < d3vz3r0> oh god
10:59 < xmj> bad jokes are sometimes fun
10:59 < d3vz3r0> sometimes
10:59 < kanzure> fenn: what about a strain of bamboo that requires less water
10:59 < fenn> kanzure: back to the original question you asked, "do you really want all of your tagged data to be based on ikiwiki's compiler" the user would add the tag data to the file itself
10:59 < kanzure> i don't think that opening up each file and looking through the wiki syntax for tags is a good idea
10:59 < fenn> the compiler only uses the tag data to build a list of pages with that tag
10:59 < kanzure> there can be thousands or millions of tags attached to an object
10:59 < gradstudentbot> Someone's sitting at my bench space.
11:00 < kanzure> probably the mustard fiend
11:00 < fenn> when you add a tag to a file and save it, the compiler adds the tag to the index for that tag
11:00 < fenn> or at least that's how i would do it
11:00 < kanzure> but then what's my querying interface for 2, 3, 500 tags?
11:01 < fenn> don't do that, asshole!
11:01 < fenn> why do you need to query for 500 tags
11:01 < kanzure> why shouldn't i query tagged data?
11:01 < d3vz3r0> fenn: please don't write software with that approach...
11:01 < d3vz3r0> people will always do things you've never thought of
11:02 < d3vz3r0> and would consider 'dumb', but they have very good reasons for doing what they are doing...
11:02 < fenn> that doesn't mean i have to indulge them
11:02 < d3vz3r0> no, but just don't make your software fall apart because they did something you didn't plan for
11:02 < fenn> it wouldn't fall apart
11:02 < d3vz3r0> ok
11:03 < d3vz3r0> I'll believe you:)
11:03 < fenn> this is all hypothetical anyway
11:03 < kanzure> tagging isn't
11:03 < kanzure> i have >50,000 bookmarks
11:03 < fenn> how many tags per bookmark, on average
11:03 < fenn> what's the median number of tags per bookmark
11:03 < kanzure> none, because my bookmarks suck
11:03 < fenn> oh. that's suboptimal
11:04 < fenn> are they in a tree?
11:04 < kanzure> sometimes
11:04 < fenn> is it just a list of 50k bookmarks?
11:05 < xmj> 21:02:19 < d3vz3r0> no, but just don't make your software fall apart because they did something you didn't plan for
11:05 < xmj> d3vz3r0: you know that joke right? "ATTENTION ATTENTION, User actually uses software for the purpose the developer intended it to be used!!!!"
11:05 < d3vz3r0> tea
11:05 < d3vz3r0> tea
11:05 < d3vz3r0> yea
11:05 < d3vz3r0> fucking typing..
11:06 < d3vz3r0> I've been at this since like 1994, software is *never* used for its intended purpose
11:06 < fenn> what was the xanadu grid system called...
11:08 < fenn> generalized search tree? it was some kind of cartesian row/column interface for data of arbitrary and nonuniform number of dimensions
11:08 < dingo> http://spectrum.ieee.org/img/BASIC-1398952320593.jpg
11:09 < d3vz3r0> Enfilades?
11:09 < d3vz3r0> dingo, that girl is prolly dead now
11:10 < d3vz3r0> fenn: used redis before?
11:10 < fenn> no
11:10 < dingo> ! use it
11:10 < d3vz3r0> ^^^
11:11 < d3vz3r0> use it
11:11 < fenn> what does it do
11:11 < d3vz3r0> seriously
11:11 < d3vz3r0> and learn how zsets work
11:11 < dingo> it changes the way you write software
11:11 < d3vz3r0> http://redis.io/
11:11 < fenn> it's a hash table i can save to disk?
11:11 < d3vz3r0> dingo is right
11:11 < d3vz3r0> no, read the fucking website
11:11 < d3vz3r0> it's a data structure server
11:11 < dingo> http://redis.io/topics/data-types
11:12 < d3vz3r0> don't hold his hand
11:12 < d3vz3r0> his/her
11:12 < d3vz3r0> its
11:12 < fenn> okay set operations are useful
11:18 < fenn> it looks like it was designed to be a database for people who realized how bad databases are but couldn't turn their minds back outside in
11:20 < fenn> is a sorted set an index of other keys (a list of pointers) or does it actually contain all the data?
11:20 < dingo> its about data structures and alogorithms
11:20 < dingo> it offloads the issue to a "database"
11:20 < dingo> its not about storing data, its about how its stored and retrieved
11:20 < dingo> btree's and all that shit
11:21 < dingo> its malloc() connected to tcp/ip or unix sockets
11:21 < dingo> and then all those silly data structures is just the pudding
11:21 < fenn> "you can use a sorted set with elements having the age of the user as the score and the ID of the user as the value" this sounds like it's just a table
11:22 < fenn> i never want to have to dereference a pointer (lookup an ID)
11:23 < d3vz3r0> seriously?
11:23 < fenn> is that too much to ask
11:23 < d3vz3r0> yes
11:23 < juri_> absolutely.
11:24 < fenn> i'm being oppressed! help!
11:24 < d3vz3r0> I don't know how else to answer that... the concept of a pointer is probably the most fundamental thing in computers
11:24 < d3vz3r0> in CS
11:26 -!- dcary [46b171ae@gateway/web/freenode/ip.70.177.113.174] has quit [Ping timeout: 240 seconds]
11:28 < catern> you could eliminate pointers with laziness maybe?
11:28 < d3vz3r0> magic
11:28 < d3vz3r0> I tried once, but I ended up creating a whole new universe that I ended up having to destroy
11:28 < d3vz3r0> total waste
11:29 -!- delinquentme [~dingo@74.61.157.78] has quit [Read error: Connection reset by peer]
11:29 < xentrac> d3vz3r0: you may not be aware of this but people were programming computers for more than a decade before they started making linked lists
11:29 < fenn> i thought there was a function to get the pointer of an object in python, but i guess not
11:29 < xentrac> 1945-1955
11:29 < xentrac> fenn: id
11:30 < d3vz3r0> yes, I'm aware
11:30 < d3vz3r0> but they still used symbols and pointers
11:30 < xentrac> in a way, yeah: jump addresses and subroutine return addresses, for example
11:30 < d3vz3r0> fenn: you never want that, that's why you can't get it
11:30 < d3vz3r0> and you can, btw, it's the 'id(..)' method
11:30 < xentrac> subroutines date from about 1945
11:31 < kanzure> .py id(object())
11:31 < yoleaux> 3721690541589520088
11:31 < kanzure> okay then
11:31 < xentrac> .py '%016x' % id(object())x
11:31 < yoleaux> SyntaxError: unexpected EOF while parsing (, line 1)
11:31 < xentrac> .py '%016x' % id(object())
11:31 < yoleaux> 33a619d4475fded8
11:31 < xentrac> .py '%016x' % id(object())
11:31 < yoleaux> 33a619d4475fded8
11:31 < kanzure> ah.
11:32 < kanzure> i also would have accepted hex(id(object()))
11:32 < xentrac> I forget about hex()
11:32 < xentrac> d3vz3r0: and certainly the program counter and addresses of variables inside your program were pointers
11:32 < d3vz3r0> yea, you want to hex it
11:32 < d3vz3r0> exactly
11:33 < d3vz3r0> *we* are symbollic computers
11:33 < d3vz3r0> we are most certainly going to create symbolic computers
11:33 < d3vz3r0> which is what a pointer is
11:33 < xentrac> a pointer is a symbolic computer?
11:33 < d3vz3r0> no, a symbol
11:33 < d3vz3r0> sorry
11:33 < d3vz3r0> a symbol that is usedwithin the symbollic computer
11:33 < d3vz3r0> a symbol to another symble
11:33 < d3vz3r0> symbol
11:34 < d3vz3r0> btw, I just started re-reading Godel, Escher, Bach
11:34 < d3vz3r0> so yea
11:34 < fenn> go read some neuroscience :P
11:34 < d3vz3r0> suggestions
11:34 < d3vz3r0> ?
11:34 < d3vz3r0> I'd love to
11:34 < kanzure> hm, thinking
11:35 < fenn> "on intelligence" by jef hawkins is a nonstandard but helpful intro
11:35 < kanzure> it is not very neurosciencey, that's more "hierarchical bayesian networks and other obvious things"
11:35 < fenn> it shows how the system works though
11:35 < kanzure> although he does talk about the visual cortex layers
11:37 < d3vz3r0> ah, cool
11:37 < fenn> speaking of which my brain feels pretty swollen
11:37 < kanzure> what about something more specific, like http://diyhpl.us/~bryan/papers2/neuro/Role%20of%20layer%206%20of%20V2%20visual%20cortex%20in%20object-recognition%20memory.pdf
11:37 < kanzure> i guess that one doesn't talk about other possible models of brain matter
11:38 < pyotra> d3vz3r0 You should familiarize yourself with Gerald Edelman
11:38 < xentrac> so I think "*we* are symbolic computers" is kind of a narrow view
11:38 < xentrac> certainly we are capable, poorly, of symbolic computation
11:38 < kanzure> xentrac: i'm sure ruldolf steiner would like that view
11:38 < d3vz3r0> sure, I'll accept that
11:39 < kanzure> xentrac: (obscure waldorf reference) http://wn.rsarchive.org/Lectures/OccSigns/OccSgn_index.html
11:39 < xentrac> but I think the failure of 1960s AI shows that most of the cognitive things we do are not done through our symbolic computation abilities
11:39 < fenn> the brain is such a noisy stochastic system that symbolic computation instantly evaporates without supporting infrastructure for the symbolic data
11:40 < xentrac> I've been worried about brain swelling a lot lately, fenn
11:40 < xentrac> not my brain
11:41 < fenn> archetypes are the built-in infrastructure, it's why people all believe in dragons and god and straight lines
11:41 * kanzure squints
11:41 < fenn> but not very useful for doing math (maybe lines are)
11:41 < cluckj> ...
11:41 < xentrac> rather the brain of the drunk guy over whose head I smashed a bottle Sunday night
11:41 < kanzure> uh
11:41 < kanzure> so, i don't think we have neurophysiological evidence of archetype whatever
11:41 < cluckj> http://en.wikipedia.org/wiki/Structural_anthropology
11:42 < xentrac> but I'm pretty sure that the police would have contacted me if anything bad had happened to him
11:42 < fenn> maybe not, but it's too common to overlook these similarities between widely differing cultures
11:42 < kanzure> "essentially that all cultures are equatable" pls not now
11:42 < kanzure> fenn: communication is a thing that happens
11:42 < xentrac> you could call his plan to threaten me and then fight with the police "hminusroadmap"
11:42 < cluckj> kanzure, relatable is a better word
11:43 < kanzure> oh i read it as equitable
11:43 < fenn> cluckj: can you say why you linked that page (it's not what i thought it would be)
11:43 < cluckj> archetypes are the built-in infrastructure, it's why people all believe in dragons and god and straight lines
11:43 < pyotra> lool you could call his plan to threaten me and then fight with the police "hminusroadmap"
11:43 < xentrac> kanzure: we don't have much neurophysiological evidence of most cognitive phenomena
11:43 < fenn> that's what _i_ said
11:43 < cluckj> yes
11:44 < pyotra> Jeff Hawkins annoys me; and here's why,
11:44 -!- superkuh [~superkuh@unaffiliated/superkuh] has quit [Quit: the neuronal action potential is an electrical manipulation of reversible abrupt phase changes in the lipid bilayer]
11:44 < kanzure> xentrac: yes, which is why so many suggestions ("well, clearly the brain is computing based on archetypes") are boring
11:44 < cluckj> I linked it because that's a scientific explanation of what you're talking about
11:44 < gradstudentbot> Yeah, it's significant.
11:44 < xentrac> gradstudentbot++
11:44 < gradstudentbot> Who the hell stole my pipette?
11:44 < cluckj> hahaha
11:45 < fenn> cluckj: that wasn't science, it was baseless philosophical rambling
11:45 < gradstudentbot> Yeah, I'm familiar with the technique. Sort of.
11:46 < pyotra> So crows will bend a wire into a J-shape in order to lift a stopper out of a tube with the "hook" they fashioned. Crows are birds (not mammals) and hence do not have a cortex at all. And yet neuroscience has not the foggiest clue how their brains do this.
11:46 < kanzure> unfortunately, neuroanatomy people are universally awful
11:46 < kanzure> so there are things called cortexes that are not what you think
11:46 < kanzure> i'm sure crows have at least one cortex somewhere
11:46 < pyotra> Jeff Hawkins pretension that he has discovered the key to "intelligence" as a whole must be taken as the snake oil that it is.
11:47 < xentrac> like doctors in 1700 or chemists in 1200, kanzure?
11:47 < cluckj> fenn, it's based on anthropological fieldwork
11:47 < xentrac> I have a cortex in each kidney I think
11:47 < fenn> i thought it would be something like http://en.wikipedia.org/wiki/Form_constants a universal human experience
11:47 < cluckj> the wiki article is kinda trashy (it's wikipedia)
11:47 < kanzure> fenn: it's an alternative explanation other than "neurons are causing this directly"
11:47 < cluckj> yes
11:47 < gradstudentbot> Oh, that's problematic.
11:47 < cluckj> instead of looking down to the neuro level, it's a theory that goes up in the other direction, to bigger things
11:47 < kanzure> xentrac: yes, i'm complaining about how we still don't have a good naming system or way of talking about brain matter parts
11:48 < kanzure> cluckj: human conversation/gossip is bigger?
11:48 < kanzure> xentrac: neuron naming is much better, for some reason
11:48 < cluckj> yes, it's not at a micro-level of analysis
11:49 < fenn> cluckj: "Lévi-Strauss' approach arose in large part from dialectics expounded on by Marx and Hegel, though dialectics (as a concept) dates back to Ancient Greek philosophy." how is this not philosophical rambling again?
11:49 < cluckj> lol, form constants
11:49 -!- superkuh [~superkuh@unaffiliated/superkuh] has joined ##hplusroadmap
11:49 < fenn> form constants exist
11:50 < xentrac> Log in or create an account to start the Avian neuroscience article, alternatively use the Article Wizard, or add a request for it.
11:50 < kanzure> _archels: is there a good neuroanatomy-specific wiki?
11:50 < kanzure> article wizard?
11:50 < cluckj> fenn, in his book on structural anthropology there are actual examples of what he's talking about, not just a philosophical distillation of what he's talking about
11:50 < cluckj> form constants are not the same thing as structural anthropology
11:50 < kanzure> wow what an awful book http://en.wikibooks.org/wiki/Neuroscience/Neuroanatomy/The_Brain
11:51 < xentrac> however there is https://en.wikipedia.org/wiki/Animal_consciousness#Birds and https://en.wikipedia.org/wiki/Bird_intelligence
11:51 < cluckj> structural anthropology is about the form and structure that beliefs take, and that those structures of sense and meaning-making are the similar things between cultures
11:51 < cluckj> not merely the symbols themselves
11:51 < kanzure> pyotra: i haven't been able to convince fenn about the reasons why hawkins is annoying in that sense
11:51 < xentrac> and in particular https://en.wikipedia.org/wiki/Bird_intelligence#Brain_anatomy
11:52 < kanzure> huh, i didn't know the CLARITY method had a longer video published:
11:52 < kanzure> .title https://www.youtube.com/watch?v=c-NMfp13Uug
11:52 < yoleaux> See-through brains
11:52 < fenn> cluckj: form constants are a genetically encoded geometric pattern that is part of the subjective experience. there are probably others that are harder to demonstrate, but we get them confused with "culture"
11:52 < xentrac> "The basal ganglia only occupy a small part of the avian brain. Instead, it seems that birds use a different part of their brain, the medio-rostral neostriatum/hyperstriatum ventrale (see also nidopallium), as the seat of their intelligence,"
11:53 < fenn> cluckj: they might take the form of "dragon" or "line" or "god"
11:53 < xentrac> " The region was renamed to nidopallium in 2002 during the Avian Brain Nomenclature Consortium because the prior name, neostriatum, suggested that the region was used for more primitive functions as the neostriatum in mammalian brains is sub-cortical."
11:53 < kanzure> avian brain.. nomenclature.. consortium.
11:53 < cluckj> fenn, the shapes and symbols that you might see in form constants are a different thing than what structural anthropology addresses
11:53 < fenn> i'm saying they shouldn't be
11:54 < fenn> if "immutable deep structures exist in all cultures" aren't genetic, then where did they come from?
11:54 < cluckj> the particular shapes/sizes/forms of symbols are irrelevant to *how* those shapes/sizes/forms are arranged in a system of meaning-making
11:54 < xentrac> http://avianbrain.org/atlases.html
11:54 < cluckj> (e.g. culture)
11:55 < cluckj> culture is the system, symbols are the bits that carry meaning inside the system
11:55 < fenn> the form of symbols is irrelevant. in any good coding system your simple symbols are used most often or when they need to be communicated quickly and with minimal ambiguity
11:55 < fenn> fuck.
11:55 < cluckj> lol
11:55 < fenn> the form of symbols is NOT irrelevant
11:56 < cluckj> it's irrelevant to how the forms are arranged in a structure
11:57 < cluckj> something can look like a dragon in many cultures, but what that form-of-dragon means in many cultures is different
11:57 < fenn> yeah because dragons went extinct and we reused the symbol for something else
11:57 < xentrac> fenn: ASCII or UCS-4 may not be good coding systems, but nevertheless they work
11:59 < fenn> xentrac: why is ASCII not a good coding system?
11:59 < kanzure> "ontological framework for neuroanatomy" http://trac.biostr.washington.edu/trac/raw-attachment/wiki/minutes_6-29-07/OMtalk.6.29.07.pdf
11:59 < kanzure> cc fenn
11:59 < fenn> bored already
11:59 < cluckj> alright :P
11:59 < xentrac> fenn: by your definition; in ASCII all 128 symbols are equally simple
12:00 < xentrac> it's true that symbolic processing of a system that doesn't fulfill your desideratum is more expensive
12:01 < xentrac> but in most cases that doesn't stop it from being workable
12:01 < xentrac> so in that sense the form of symbols is of limited, though nonzero, relevance
12:02 < fenn> hint to powerpoint users: don't use cyan or yellow text on a white background
12:03 < fenn> so kanzure is there a thing that will circle the relevant part of the brain for me when i click on its entry in the tree?
12:04 < fenn> xentrac: i think you're confusing symbols with bits? or maybe i am confused
12:05 < xentrac> well, a bit is a symbol, but I was talking about the 7-bit symbols of ASCII
12:05 < fenn> a bit doesn't represent anything, so it's not a symbol
12:06 < kanzure> fenn: i think you should be asking that to _archels
12:06 < cluckj> o_O
12:06 < xentrac> fenn: I thought about that point of view, but I rejected it
12:06 < xentrac> because when a bit doesn't represent anything, it's not even a bit either
12:06 < cluckj> a bit represents something because of its relation to other bits
12:06 < cluckj> 1 means 1 because it's not 0
12:07 < xentrac> instead it's some arbitrary analog voltage or RF signal or bias of magnetization
12:07 < xentrac> it's at the point that we decide to interpret that voltage or whatever as meaning something that it becomes a bit
12:07 < cluckj> ^
12:07 < xentrac> and, by the same token, a symbol
12:07 < cluckj> interpret == assign meaning
12:08 < fenn> ok a bit is a binary digit
12:09 < fenn> ascii is mapping binary digits to letters and punctuation
12:09 < xentrac> well, 7-vectors of them, but yes
12:09 < fenn> in that mapping it exhibits poor efficiency
12:09 < xentrac> yes
12:09 < fenn> the poor efficiency is _because_ it uses 7-vectors for all letters and punctuation
12:10 < xentrac> yes
12:10 < fenn> but at mapping bytes to letters, it couldn't be any more efficient
12:11 < xentrac> why not? huffman can get you down to about two letters per byte on average
12:11 < xentrac> http://worldpowersystems.com/J/codes/ talks about some other codes that achieve slightly over 5 bits per letter
12:11 < xentrac> like ITA2
12:12 < fenn> sure there are only 26 letters and 2^5 is 32 so that's plenty
12:12 < fenn> i am just playing dumb
12:12 < cluckj> lol
12:12 < xentrac> that was the reasoning behind ITA2, yes
12:13 < fenn> it doesn't explain why huffman codes are efficient though (or arithmetic coding)
12:13 < cluckj> but it doesn't matter if the letters you're encoding are greek or cyrillic, you're going to be able to do them using the same structure
12:14 < xentrac> that's because, as you said, in any good coding system your simple symbols are used most often
12:14 < xentrac> which is why the relevance of the form of the symbols is nonzero
12:15 < fenn> only two letters per byte?
12:16 < fenn> most internet traffic is just "lol" and "narf"
12:16 < ThomasEgi> most internet traffic is video data.
12:16 < fenn> and translate?
12:21 < kanzure> they do cdn/caching stuff
12:21 < fenn> why is video traffic a large part of internet traffic?
12:21 < kanzure> sometimes you have to block your isp's ip addresses so that you can get to the actual youtube servers instead of the cached content
12:22 < cluckj> cat videos > cat pics
12:22 < gradstudentbot> I'm glad you brought that up, I'm going to do that right now.
12:23 < fenn> so comcast actually caches youtube videos?
12:23 < kanzure> i'm not sure if it's the isp itself, or if youtbe struck up deals, or if youtube is just using some geographically-distributed cdn service like akamai or what
12:23 < kanzure> *youtube
12:25 < fenn> isn't it in comcast's interest to cache youtube/netflix because they have to pay for that bandwidth otherwise?
12:25 < fenn> are they just stupid or is there a good reason not to cache?
12:25 < fenn> basically, why doesn't the internet work like a big DHT
12:25 < fenn> or freenet
12:26 -!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 276 seconds]
12:26 * fenn gets on his knees and shakes his fist at the heavens
12:26 < fenn> damn you, cloud!
12:29 < cluckj> lol
12:29 < kanzure> INSTANCE_09956',5,'name','','\0',0,3,'Left posterior seventh thoracic radicular vein',NULL),('KB_INSTANCE_09956',5,'au
12:29 < kanzure> Submucosa of accessory superior segmental bronchus
12:30 < kanzure> here's a query interface, but no brain map visualization thing http://fma.biostr.washington.edu:8080/noqafma/query.jsp
12:30 < fenn> cluckj: the cat video phenomenon is possibly the sort of thing i'm talking about
12:31 < cluckj> oh?
12:31 < fenn> "why are people so into cats"
12:32 < kanzure> DatabaseError
12:32 < kanzure> this is useless
12:32 < fenn> Error!
12:33 -!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap
12:34 < kanzure> i dunno, not convinced http://brancusi.usc.edu/ontology/ontology-details.php?id=116
12:34 -!- Shehrazad [~Shehrazad@unaffiliated/shehrazad] has quit [Quit: Leaving]
12:35 -!- ElixirVitae [~Shehrazad@unaffiliated/shehrazad] has joined ##hplusroadmap
12:35 < cluckj> fenn, that doesn't tell you about the structure of the system though
12:36 < cluckj> sorry, I had to let some bees off my porch before my cat decided to eat them
12:37 < fenn> xentrac: morse code makes so much more sense as a tree than as a table
12:37 < fenn> http://www.learnmorsecode.com/
12:39 < cluckj> you might have a nice, neat, biological explanation for why that one particular thing is so popular
12:39 < kanzure> .wik neuronames
12:39 < cluckj> but that biological explanation is predicated on having already decided that babies are important
12:39 < yoleaux> "NeuroNames is an integrated nomenclature for structures in the brain and spinal cord of the four species most studied by neuroscientists: human, macaque, rat and mouse. It offers a standard, controlled vocabulary of common names for structures, which is suitable for unambiguous neuroanatomical indexing of information in digital databases." — http://en.wikipedia.org/wiki/NeuroNames
12:39 < kanzure> "BrainInfo helps one identify structures in the brain. One can either search by a structure name or locate the structure in a brain atlas and get information such as its location in the classical brain hierarchy, images of the structure, what cells it has, its connections and genes expressed there. Information can be accessed by any of some 16,000 synonyms in eight languages."
12:39 < cluckj> so cats == babies == reproduction
12:40 < fenn> i think it's messier than that
12:40 < cluckj> what that can show you is the connections between those different symbols, which is the structure that we were talking about earlier
12:40 < cluckj> oh it's waaaaaay messier than that
12:41 < cluckj> structural anthro is saying that the connections between those concepts are the structures that make up culture, not the concepts/signs/symbols themselves
12:41 < kanzure> number of 'a's correlates to number of powers of ten to include?
12:43 < fenn> where am i, how did i get here
12:43 < cluckj> kanzure, yeah
12:43 < cluckj> well no
12:43 < cluckj> not even close
12:43 < kanzure> here is a terrible semantic mediawiki for neuroanatomy: http://neurolex.org/wiki/Category:Hippocampal_formation
12:43 <@_archels> kanzure: wikis do not seem the ideal data structure for describing neuroanatomy
12:43 < kanzure> duh
12:43 < kanzure> i absolutely agree
12:44 < kanzure> apparently the hippocampus is also called "ammon's horn"
12:44 -!- pyotra [~asakharov@24.60.79.55] has quit [Quit: quit]
12:44 <@_archels> are you playing with the Allen Brain Atlas?
12:45 < kanzure> _archels: i am having trouble evaluating the level of knowledge each author (of each neuroscience-related paper that i read) regarding neuroanatomy, because there's not only a naming problem but also a "what connects where" problem
12:45 <@_archels> the cornus ammonis areas refer to subparts of the hippocampus afaik, not the whole structure
12:45 < kanzure> kdfladjfkldajfka
12:45 < cluckj> oh dear
12:46 <@_archels> heh, heh
12:46 -!- marciogm [~marciogm@179.126.71.80] has quit [Quit: Computer has gone to sleep.]
12:46 <@_archels> yeah, it's a mess
12:46 < fenn> it's all partially overlapping sets
12:46 < kanzure> well, we should unmess it
12:46 <@_archels> we're organising a session on brain atlasing in our upcoming symposium. join us!
12:46 < cluckj> it'd be nice if there were some clean lines between all those bits of brain like there are in the rest of the body
12:46 -!- entelechios [~elysium@181.194.150.46] has joined ##hplusroadmap
12:47 < fenn> yeah, like the butchers diagram of the pig
12:47 < kanzure> there are actually a number of such lines
12:47 < cluckj> yeah
12:47 < kanzure> but now you have to apply that line map against the other existing name datasets ;)
12:47 < kanzure> and their names alone don't tell me about which neurons are projecting inwards or outwards
12:47 < kanzure> (and from where or to where)
12:47 < fenn> that's not important for chopping it into pieces
12:47 < fenn> just tell me where to cut
12:48 < kanzure> everywhere. in 1 micron intervals.
12:48 * fenn sharpens furiously
12:49 < kanzure> why is it still a mess
12:49 < fenn> is there a schematic of the brain? instead of just a circuit layout
12:49 < kanzure> superkuh had one..
12:49 < fenn> i mean there's no circuit layout either
12:49 < kanzure> it had giant arrows and rectangles
12:50 < fenn> but a description of functions and pathways
12:50 <@_archels> the problem is that there are too many "schematics of the brain" and we don't know how to integrate them
12:50 < cluckj> ooo rectangles
12:50 < kanzure> yeah, a circuit graph would be much nicer
12:50 < kanzure> and then you can just do clustering or whatever
12:50 < fenn> it's common to have multiple schematics of a complex circuit
12:50 < cluckj> so a map of the connections would be much better than a map of names
12:50 < cluckj> hmmmmmm
12:52 < fenn> olaf sporns did some stuff based on white matter tracts (diffusion tensor imaging)
12:52 < fenn> but that's like, data
12:52 < kanzure> why is data bad?
12:53 < fenn> it's expensive to communicate verbally
12:54 < kanzure> videos of some guy riding elevators https://www.youtube.com/user/gluse
12:54 < kanzure> https://www.youtube.com/user/SMOKERSOFCIGARSPIPES/ "See also: A man who has uploaded 6,300 videos and counting of him doing nothing but smoking pipes and grumbling unintelligible streams of broken English that are conveniently transcribed in the description"
12:55 < cluckj> thank you youtube for archiving those valuable insights
12:56 < kanzure> huh, ok "could wearable technology that counts words heard (a key indicator of cognitive development) be used to improve quality of early childhood education?"
12:56 < kanzure> yes i guess word counting would be a nice metric to have
12:57 < fenn> words heard?
12:58 < fenn> you mean words spoken?
12:58 < kanzure> clearly there might be problems if no words are heard or spoken
12:58 < kanzure> oops i mean "and"
12:58 < fenn> otherwise your wearable device has to dig into the brain to see if the words were perceived or not
12:58 < kanzure> i'm okay with someone not talking, but how can they hear if nobody speaks to them?
12:59 < fenn> right
12:59 < kanzure> (fuck talking)
12:59 < cluckj> irc 4 lyfe
12:59 < fenn> but they could be as dumb as a rock and have lots of words be spoken to them (this counts as words heard)
13:00 < kanzure> fenn: i think that's fine
13:00 < fenn> or they could be deaf and the metric is useless
13:00 <@_archels> kanzure: those lift videos are hilarious
13:00 < cluckj> or you could be mute and the metric is useless
13:00 < kanzure> "number of words spoken, in plain sight of eyes"
13:01 < fenn> ok i shouldn't have used "dumb" to mean "low cognitive development"
13:01 < fenn> fuck english
13:01 < cluckj> hahaha
13:02 < fenn> i kinda wish sapir-whorf were true
13:03 < cluckj> it sure would be convenient
13:03 < fenn> "things that are untranslatable" actually turn out to just be really long explanations that make the joke unfunny
13:05 < cluckj> welcome to the human race
13:07 < kanzure> "1st Call For Papers, 21th Annual Tcl/Tk Conference 2014"
13:07 < kanzure> narf
13:11 < fenn> shit. i think the cat drugs were making me sick (cat has monthly flea medicine on it sitting next to me)
13:12 < fenn> please disregard the past 5 hours of semi-conscious rambling
13:14 < catern> do you still hate databases?
13:14 < kanzure> pfft i've been speaking unconsciously for years
13:14 < cluckj> you were pretty coherent
13:18 < cluckj> I was able to discern what you were trying to say, and attempt to convince you otherwise :P
13:19 -!- nsh [~nsh@wikipedia/nsh] has quit [Ping timeout: 246 seconds]
13:21 -!- nmz787_i [nmccorkx@nat/intel/x-hzixdmkkksqeyzeb] has quit [Quit: Leaving.]
13:31 < chris_99> http://www.dailymail.co.uk/sciencetech/article-2617705/Brain-implant-restore-MEMORIES-wounded-soldiers-Alzheimers-sufferers-ethical.html
13:36 -!- nsh [~nsh@wikipedia/nsh] has joined ##hplusroadmap
13:39 -!- nmz787_i [nmccorkx@nat/intel/x-kqovnclpnfkxzukz] has joined ##hplusroadmap
13:55 < kanzure> amusing anti-bitcoin arguments https://news.ycombinator.com/item?id=7675418
14:23 < kanzure> .title http://www.nytimes.com/2014/05/01/technology/after-alibaba-ipo-us-web-giants-may-stop-ignoring-chinese-rivals.html?_r=0
14:23 < yoleaux> kanzure: Sorry: that command is a web-service, but its response was too long.
14:25 < kanzure> okie dokie
14:27 < kanzure> "They have a stealth 'target vertical' shopping experience that's developing right now in the Bay Area, however, they're significantly behind schedule, having trouble integrating teams from several locals and their product is dubiously "useful" at this stage of the game."
14:27 < kanzure> huh. that's weird.
14:28 < kanzure> i prefer the original alibaba
14:35 < kanzure> alibaba facebook admins:
14:43 -!- FourFire [~fourfire@100-98-15.connect.netcom.no] has left ##hplusroadmap ["Leaving"]
14:46 < kanzure> i wonder if they are deleting data from their provisioned storage http://buy.aliyun.com/?spm=5176.383338.21.3.H0DCAU
14:53 < kanzure> alibaba had a remote code execution vulnerability in 2007, whee http://seclists.org/fulldisclosure/2007/Feb/146
14:54 < kanzure> oh, that's slightly false, oops
15:01 -!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 252 seconds]
15:08 -!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Quit: Leaving]
15:09 -!- nsh [~nsh@wikipedia/nsh] has quit [Ping timeout: 240 seconds]
15:17 -!- nsh [~nsh@wikipedia/nsh] has joined ##hplusroadmap
15:34 -!- EnLilaSko [EnLilaSko@unaffiliated/enlilasko] has quit [Quit: - nbs-irc 2.39 - www.nbs-irc.net -]
15:44 -!- ielo [~ielo@host-92-24-44-62.ppp.as43234.net] has joined ##hplusroadmap
15:46 -!- augur [~augur@216-164-48-148.c3-0.slvr-ubr1.lnh-slvr.md.cable.rcn.com] has joined ##hplusroadmap
16:07 -!- yorick [~yorick@oftn/member/yorick] has quit [Read error: Connection reset by peer]
16:14 * nsh is looking for things to watch with clever in them
16:14 < nsh> like... science!
16:14 < nsh> can anyone point me at some good audiovisual science learns?
16:16 < sheena> um
16:16 < sheena> how advanced?
16:16 < sheena> nsh: and topic?
16:16 < kanzure> nsh: https://www.youtube.com/watch?v=c-NMfp13Uug
16:17 < nsh> all of the topics, all of the advancements. i mean, it would be nice if it didn't all go over my head, but it'd pick that over low information density any time
16:17 < nsh> .t
16:17 < yoleaux> Thu, 01 May 2014 23:17:02 UTC
16:17 < nsh> .title https://www.youtube.com/watch?v=c-NMfp13Uug
16:17 < yoleaux> See-through brains
16:18 < nsh> neat
16:18 < kanzure> nsh: this one is an oldie but a goodie (and longie) https://www.youtube.com/watch?v=9gFI7o69VJM&t=4m
16:19 < kanzure> d3vz3r0: you might like that video as well (it's about an hour long, about neuroscience)
16:20 < sheena> youtubey things like brain scoop, targetted for non-academic audiences.... vs coursera and udacity higher education courses..
16:20 -!- cpopell [~cpopell@pool-71-255-241-91.washdc.fios.verizon.net] has quit [Quit: Nettalk6 - www.ntalk.de]
16:20 -!- cpopell [~cpopell@pool-71-255-241-91.washdc.fios.verizon.net] has joined ##hplusroadmap
16:26 -!- ebowden [~ebowden@CPE-60-231-177-134.lns4.dav.bigpond.net.au] has joined ##hplusroadmap
16:26 < nsh> ty
16:49 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=7d77387e Bryan Bishop: partial transcript of markram 2006 talk
16:54 -!- nsh [~nsh@wikipedia/nsh] has quit [Ping timeout: 240 seconds]
16:55 -!- nsh [~nsh@wikipedia/nsh] has joined ##hplusroadmap
16:56 < kanzure> fenn, if you use 2 wires per element does it matter where each of the 2 wires are placed on its surface
16:56 < kanzure> since you said capacitor, i would guess yes
17:01 -!- nsh [~nsh@wikipedia/nsh] has quit [Ping timeout: 252 seconds]
17:04 -!- nsh [~nsh@wikipedia/nsh] has joined ##hplusroadmap
17:05 -!- ebowden [~ebowden@CPE-60-231-177-134.lns4.dav.bigpond.net.au] has quit [Remote host closed the connection]
17:05 -!- nsh [~nsh@wikipedia/nsh] has quit [Max SendQ exceeded]
17:06 -!- nsh [~nsh@wikipedia/nsh] has joined ##hplusroadmap
17:06 -!- ebowden [~ebowden@CPE-60-231-177-134.lns4.dav.bigpond.net.au] has joined ##hplusroadmap
17:12 < kanzure> http://spectrum.ieee.org/aerospace/astrophysics/the-hunt-for-the-invisible-axion
17:21 -!- cpopell is now known as cpopell`gym
17:27 -!- ielo [~ielo@host-92-24-44-62.ppp.as43234.net] has quit [Ping timeout: 252 seconds]
17:41 -!- nmz787_i [nmccorkx@nat/intel/x-kqovnclpnfkxzukz] has quit [Ping timeout: 265 seconds]
17:51 -!- delinquentme [~dingo@74.61.157.78] has joined ##hplusroadmap
18:16 -!- ElixirVitae [~Shehrazad@unaffiliated/shehrazad] has quit [Remote host closed the connection]
18:18 -!- Adifex|zzz is now known as Adifex
18:51 -!- dcary [46b171ae@gateway/web/freenode/ip.70.177.113.174] has joined ##hplusroadmap
19:20 < kanzure> beep boop
19:22 -!- ielo [~ielo@host-92-24-44-62.ppp.as43234.net] has joined ##hplusroadmap
19:30 < catern> i saw that kanzure
19:30 < catern> you can't hide your robot nature
19:34 -!- Adifex is now known as Adifex|pub
19:41 -!- ThomasEgi [~thomas@panda3d/ThomasEgi] has quit [Remote host closed the connection]
19:46 -!- ielo [~ielo@host-92-24-44-62.ppp.as43234.net] has quit [Ping timeout: 246 seconds]
20:45 -!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap
21:10 -!- dcary [46b171ae@gateway/web/freenode/ip.70.177.113.174] has quit [Ping timeout: 240 seconds]
21:11 -!- HEx2 [~HEx@hexwab.plus.com] has joined ##hplusroadmap
21:12 -!- HEx1 [~HEx@hexwab.plus.com] has quit [Ping timeout: 276 seconds]
21:22 -!- HEx2 [~HEx@hexwab.plus.com] has quit [Ping timeout: 250 seconds]
21:26 -!- HEx1 [~HEx@hexwab.plus.com] has joined ##hplusroadmap
21:29 -!- entelechios [~elysium@181.194.150.46] has quit [Quit: leaving]
21:32 -!- entelechios [~elysium@190.5.213.2] has joined ##hplusroadmap
21:53 -!- HEx2 [~HEx@hexwab.plus.com] has joined ##hplusroadmap
21:55 -!- HEx1 [~HEx@hexwab.plus.com] has quit [Ping timeout: 276 seconds]
21:56 -!- delinquentme [~dingo@74.61.157.78] has quit [Ping timeout: 250 seconds]
22:02 -!- Guest30849 [~not@100.43.114.90] has quit [Read error: Connection reset by peer]
22:03 -!- pads [~not@100.43.114.90] has joined ##hplusroadmap
22:03 -!- pads is now known as Guest85608
22:24 -!- HashNuke [uid12117@gateway/web/irccloud.com/x-iidoqunvialdcgqi] has joined ##hplusroadmap
22:39 -!- kuldeepdhaka [~kuldeepdh@unaffiliated/kuldeepdhaka] has joined ##hplusroadmap
22:45 -!- nsh [~nsh@wikipedia/nsh] has quit [Ping timeout: 240 seconds]
22:58 -!- nsh [~nsh@wikipedia/nsh] has joined ##hplusroadmap
23:18 -!- nsh [~nsh@wikipedia/nsh] has quit [Ping timeout: 255 seconds]
--- Log closed Fri May 02 00:00:57 2014