--- Log opened Fri Feb 22 00:00:07 2013 | ||
-!- indigenous [~indigenou@85.210.227.43] has joined ##hplusroadmap | 00:05 | |
-!- indigenous [~indigenou@85.210.227.43] has quit [Changing host] | 00:05 | |
-!- indigenous [~indigenou@pdpc/supporter/student/indigenous] has joined ##hplusroadmap | 00:05 | |
-!- gnusha [~gnusha@131.252.130.248] has joined ##hplusroadmap | 00:19 | |
-!- Topic for ##hplusroadmap: biohacking, nootropics, transhumanism, open hardware | sponsored by george church and the NRA | http://gnusha.org/logs http://diyhpl.us/wiki http://groups.google.com/group/diybio | banned by the Federal Death Administration | 3.5 kidneys for sale | no questions asked | 00:19 | |
-!- Topic set by kanzure [~kanzure@131.252.130.248] [Sat Feb 2 18:24:43 2013] | 00:19 | |
[Users ##hplusroadmap] | 00:19 | |
[@fenn ] [ bkero ] [ heath ] [ NeuroWinter] [ sheena1 ] [ ThomasEgi ] | 00:19 | |
[@kanzure ] [ brownies ] [ HEx1 ] [ nmz787 ] [ Shehrazad ] [ Thorbinator] | 00:19 | |
[ _sol_ ] [ Charlie_ ] [ He||eshin ] [ nuba ] [ sivoais ] [ underscor ] | 00:19 | |
[ abetusk ] [ chido ] [ indigenous] [ OldCoder_ ] [ smeaaagle ] [ upgrayeddd ] | 00:19 | |
[ AlonzoTG ] [ Coornail ] [ ivan` ] [ paperbot ] [ sseehh_ ] [ Urchin ] | 00:19 | |
[ archbox_ ] [ curtiss ] [ JayDugger ] [ ParahSailin] [ strages_home] [ Vicarious ] | 00:19 | |
[ archels ] [ devrandom ] [ joehot ] [ pasky ] [ strangewarp ] [ Viper168 ] | 00:19 | |
[ ArmilusDajjal] [ etheros ] [ jrayhawk ] [ phryk ] [ streety ] [ wizrobe ] | 00:19 | |
[ augur_ ] [ gedankenstuecke] [ klafka ] [ rigel ] [ superkuh ] [ yoleaux ] | 00:19 | |
[ balrog ] [ gillies ] [ lichen ] [ Sanky ] [ sylph_mako ] | 00:19 | |
[ barriers ] [ gnusha ] [ lupfantomo] [ saurik ] [ Thomas42 ] | 00:19 | |
-!- Irssi: ##hplusroadmap: Total of 64 nicks [2 ops, 0 halfops, 0 voices, 62 normal] | 00:19 | |
-!- Channel ##hplusroadmap created Thu Feb 25 23:40:30 2010 | 00:19 | |
-!- Irssi: Join to ##hplusroadmap was synced in 6 secs | 00:19 | |
-!- klafka [~klafka@c-67-174-253-229.hsd1.ca.comcast.net] has quit [Remote host closed the connection] | 00:58 | |
-!- klafka [~klafka@c-67-174-253-229.hsd1.ca.comcast.net] has joined ##hplusroadmap | 00:59 | |
superkuh | paperbot: http://www.sciencemag.org/content/early/2013/02/20/science.1230883.full.pdf | 01:00 |
---|---|---|
paperbot | no translator available, raw dump: http://diyhpl.us/~bryan/papers2/paperbot/74213320c17fd8e69ae059968badcfa4.txt | 01:00 |
-!- klafka [~klafka@c-67-174-253-229.hsd1.ca.comcast.net] has quit [Ping timeout: 264 seconds] | 01:03 | |
@kanzure | http://www.alibaba.com/product-gs/710041985/2013_Version_21000_KVA_Industrial_Submerged.html | 01:33 |
@kanzure | "Shaanxi Zhonglong Metallurgical Equipment Co. Ltd. was established in May 2009, with registered capital of RMB 5,000,000 (USD$ 94,156) and fixed assets 10,762,300 (USD$1,709,390). " | 01:33 |
@kanzure | huh, that's not a lot of capital. | 01:33 |
@kanzure | i mean, for an electric arc furnace manufacturer. | 01:33 |
-!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 244 seconds] | 01:44 | |
-!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap | 01:48 | |
@kanzure | "Dear Mr. Bishop, Your email below has been forwarded to me, as we cannot identify you as an athlete or official in the database. As you are not included in our list this year, could you perhaps clarify the reason for your request so that we can assist you efficiently?" | 01:54 |
@kanzure | perhaps my typing fame is not great enough. | 01:54 |
@kanzure | hmmm. | 01:54 |
@kanzure | maybe i need to make up an athlete and do some fake publicity | 01:55 |
-!- klafka [~klafka@c-24-6-18-31.hsd1.ca.comcast.net] has joined ##hplusroadmap | 02:09 | |
-!- gedankenstuecke [~bastian@phylomemetic-tree.de] has quit [Ping timeout: 252 seconds] | 02:12 | |
-!- klafka [~klafka@c-24-6-18-31.hsd1.ca.comcast.net] has quit [Ping timeout: 246 seconds] | 02:14 | |
-!- gedankenstuecke [~bastian@phylomemetic-tree.de] has joined ##hplusroadmap | 02:21 | |
-!- indigenous [~indigenou@pdpc/supporter/student/indigenous] has quit [Read error: Operation timed out] | 03:18 | |
-!- sylph_mako [~mako@103-9-42-1.flip.co.nz] has quit [Ping timeout: 276 seconds] | 04:48 | |
-!- EnLilaSko [~Nattzor@unaffiliated/enlilasko] has joined ##hplusroadmap | 05:14 | |
-!- Urchin[emacs] [~user@unaffiliated/urchin] has joined ##hplusroadmap | 05:17 | |
-!- wizrobe is now known as wizroafk | 05:41 | |
-!- Mariu [Jimmy98@89.41.57.33] has joined ##hplusroadmap | 05:48 | |
-!- augur_ [~augur@208.58.5.87] has quit [Remote host closed the connection] | 05:49 | |
-!- augur [~augur@129-2-129-34.wireless.umd.edu] has joined ##hplusroadmap | 06:21 | |
-!- sseehh_ [~me@24.131.65.218] has quit [Read error: Connection reset by peer] | 07:14 | |
-!- klafka [~klafka@c-67-174-253-229.hsd1.ca.comcast.net] has joined ##hplusroadmap | 07:17 | |
-!- yashgaroth [~ffffff@cpe-66-27-118-94.san.res.rr.com] has joined ##hplusroadmap | 07:19 | |
-!- sseehh_ [~me@c-24-131-65-218.hsd1.pa.comcast.net] has joined ##hplusroadmap | 07:39 | |
-!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 256 seconds] | 07:52 | |
-!- yashgaroth [~ffffff@cpe-66-27-118-94.san.res.rr.com] has quit [Quit: Leaving] | 07:53 | |
-!- ArmilusDajjal is now known as mutagen | 07:56 | |
nmz787 | whoo got this broken projecor working with an MSP430 launchpad! | 08:00 |
-!- delinquentme [~asdfasdf@coffeebar.static.monkeybrains.net] has joined ##hplusroadmap | 08:02 | |
-!- nsh [~nsh@wikipedia/nsh] has joined ##hplusroadmap | 08:09 | |
delinquentme | banned by the Federal Death Administration | 08:09 |
delinquentme | http://transhumanity.net/articles/entry/bill-gates-really-wants-to-becomeimmortal | 08:13 |
Mariu | I don't think he was clear enough on that issue, or was he ? | 08:14 |
-!- kanzure changed the topic of ##hplusroadmap to: biohacking, nootropics, transhumanism, open hardware | sponsored by george church and the NRA | http://gnusha.org/logs http://diyhpl.us/wiki http://groups.google.com/group/diybio | banned by the Federal Death Administration | home of the free-ranging paperbots | 08:30 | |
Mariu | :p | 08:30 |
sseehh_ | http://mashable.com/2013/02/20/mark-zuckerberg-sergey-brin-breakthrough-prize/ opensource democratic immortality http://news.yahoo.com/iceland-facebook-constitution-closer-reality-101210718.html | 08:35 |
Mariu | neat | 08:36 |
@kanzure | man i hate the news | 08:36 |
@kanzure | Mariu: do you say anything other than lol | 08:36 |
Mariu | kanzure: not really | 08:37 |
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap | 08:38 | |
ParahSailin | kanzure: interesting article that you quoted | 08:42 |
-!- klafka [~klafka@c-67-174-253-229.hsd1.ca.comcast.net] has quit [Remote host closed the connection] | 08:42 | |
-!- klafka [~klafka@c-67-174-253-229.hsd1.ca.comcast.net] has joined ##hplusroadmap | 08:42 | |
-!- klafka [~klafka@c-67-174-253-229.hsd1.ca.comcast.net] has quit [Ping timeout: 244 seconds] | 08:47 | |
ParahSailin | kanzure: i wonder how many farmers were also smart enough to buy soybeans from a distributor (not under contract with monsanto), select out the roundup ready ones, and unlike bowman, not be a retard and tell monsanto that they were doing it | 08:50 |
chris_99 | http://www.bristol.ac.uk/news/2013/9163.html | 08:53 |
@kanzure | .title | 08:58 |
yoleaux | Bristol University | News from the University | Floral signs go electric | 08:58 |
chris_99 | i found that pretty amazing | 08:58 |
-!- undefined_ [188341da@gateway/web/freenode/ip.24.131.65.218] has joined ##hplusroadmap | 08:59 | |
-!- _sol_ [~Sol@c-174-57-58-11.hsd1.pa.comcast.net] has quit [] | 09:00 | |
-!- joehot [~not@64.231.129.30] has quit [Ping timeout: 256 seconds] | 09:01 | |
-!- undefined_ [188341da@gateway/web/freenode/ip.24.131.65.218] has quit [Quit: Page closed] | 09:06 | |
sseehh_ | http://en.wikipedia.org/wiki/Crystal_healing http://www.bristol.ac.uk/news/2013/9163.html learning the difference between two colours when electric wiki.opencog.org/w/DeSTIN | 09:09 |
Mariu | .title | 09:10 |
yoleaux | Crystal healing - Wikipedia, the free encyclopedia | 09:10 |
@kanzure | sseehh_: linking to crystal healing and opencog sets off my troll alarms. please get the fuck out. | 09:11 |
@kanzure | Mariu: that title was obvious, why did you ask for it | 09:11 |
@kanzure | wikipedia hasn't changed their <title>s in years | 09:11 |
@kanzure | it's a function of the url | 09:12 |
Mariu | kanzure: got it. I was also curious about yoleaux .. didn't knew about it | 09:12 |
delinquentme | Mariu, thats the french yolo? | 09:22 |
Mariu | delinquentme: no idea | 09:22 |
-!- eudoxia [~eudoxia@r186-53-163-202.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap | 09:25 | |
-!- mutagen [~safitan@75-105-12-23.cust.wildblue.net] has left ##hplusroadmap [] | 09:27 | |
eudoxia | .wa mass of the Voyager 2 * distance between Adelaide and Montevideo | 09:28 |
yoleaux | Voyager 2: mass distance: from: Adelaide, South Australia, Australia; to: Montevideo, Uruguay: 5.825 million kg mi (kilogram miles); Unit conversions: 3.027×10⁷ lg tn ft (long ton-feet); 9226 lg tn km (long ton-kilometers); 5733 lg tn mi (long ton-miles); 9374 t km (metric ton-kilometers); 3.39×10⁷ sh tn ft (short ton-feet) | 09:28 |
eudoxia | gud, gud | 09:28 |
eudoxia | .botsnack | 09:28 |
yoleaux | :D | 09:28 |
ParahSailin | .title | 09:35 |
yoleaux | Crystal healing - Wikipedia, the free encyclopedia | 09:35 |
ParahSailin | i guess i should have noticed that was the bot | 09:36 |
-!- padz [~not@bas5-kingston08-3096457487.dsl.bell.ca] has joined ##hplusroadmap | 09:40 | |
-!- augur [~augur@129-2-129-34.wireless.umd.edu] has quit [Remote host closed the connection] | 09:45 | |
-!- augur [~augur@129-2-129-34.wireless.umd.edu] has joined ##hplusroadmap | 09:45 | |
-!- klafka [~klafka@204-16-157-18-static.ipnetworksinc.net] has joined ##hplusroadmap | 09:46 | |
-!- delinquentme [~asdfasdf@coffeebar.static.monkeybrains.net] has quit [Ping timeout: 244 seconds] | 09:47 | |
-!- augur [~augur@129-2-129-34.wireless.umd.edu] has quit [Ping timeout: 248 seconds] | 09:50 | |
-!- _Sol_ [~Sol@c-174-57-58-11.hsd1.pa.comcast.net] has joined ##hplusroadmap | 10:02 | |
-!- eudoxia [~eudoxia@r186-53-163-202.dialup.adsl.anteldata.net.uy] has quit [Quit: leaving] | 10:02 | |
klafka | wow that article on bill gates is awful | 10:03 |
klafka | just terrible | 10:03 |
klafka | well 'article' | 10:04 |
klafka | blog post | 10:04 |
Mariu | klafka: why ? | 10:04 |
klafka | 'bill gates said dont die but it sounds like he has no idea how' | 10:04 |
klafka | really? | 10:04 |
Mariu | yeah, that's the thing that bug?ed me as well | 10:04 |
@kanzure | that was just a reddit thing, it's bullshit | 10:04 |
klafka | his terse 2 word response makes you think that | 10:05 |
@kanzure | this is why i hate the news. stop posting it. | 10:05 |
klafka | i didn't post it | 10:05 |
klafka | i don't post shit | 10:05 |
@kanzure | someone else did | 10:05 |
@kanzure | not you | 10:05 |
klafka | right | 10:05 |
nmz787 | kanzure: i have this project asus_x1220h_fan_override_msp430 | 10:11 |
nmz787 | i want to add it to git | 10:11 |
nmz787 | what's a better name? | 10:11 |
@kanzure | what's wrong with that name? | 10:11 |
nmz787 | can you change a github repo later? | 10:11 |
@kanzure | what do you mean change? | 10:11 |
-!- sheena1 [~home@174.4.134.159] has quit [Ping timeout: 246 seconds] | 10:11 | |
@kanzure | rename? | 10:11 |
@kanzure | there is a github rename feature, but i don't understand your question about git | 10:12 |
nmz787 | i just played my first movie on this hacked projector :P | 10:20 |
nmz787 | focused at about 8cm wide | 10:21 |
-!- delinquentme [~asdfasdf@50-193-54-102-static.hfc.comcastbusiness.net] has joined ##hplusroadmap | 10:23 | |
-!- delinquentme [~asdfasdf@50-193-54-102-static.hfc.comcastbusiness.net] has quit [Ping timeout: 255 seconds] | 10:35 | |
-!- nsh [~nsh@wikipedia/nsh] has quit [Remote host closed the connection] | 10:44 | |
-!- sheena1 [~home@174.4.134.159] has joined ##hplusroadmap | 10:58 | |
-!- delinquentme [~asdfasdf@64-13-13-62.eug.clearwire-dns.net] has joined ##hplusroadmap | 11:08 | |
nmz787 | spam https://github.com/nmz787/DLP-Projector-fan-and-lightbulb-override-msp430 | 11:26 |
@kanzure | what is .ino? | 11:26 |
-!- delinquentme [~asdfasdf@64-13-13-62.eug.clearwire-dns.net] has quit [Quit: Leaving] | 11:28 | |
klafka | i like when books give 'chapter dependence trees' | 11:28 |
klafka | it's sooooo nerdy but also useful | 11:28 |
-!- sseehh_ [~me@c-24-131-65-218.hsd1.pa.comcast.net] has quit [Read error: Connection reset by peer] | 11:31 | |
-!- sseehh_ [~me@c-24-131-65-218.hsd1.pa.comcast.net] has joined ##hplusroadmap | 11:31 | |
-!- augur [~augur@208.58.5.87] has joined ##hplusroadmap | 11:38 | |
-!- juri_ [~juri@funkykitty.net] has joined ##hplusroadmap | 11:42 | |
bkero | http://rationalwiki.org/wiki/Roko%27s_basilisk lul | 11:44 |
bkero | "The notion is taken sufficiently seriously by some LessWrong posters that they try to work out how to erase evidence of themselves so a future unfriendly AI can't reconstruct a copy of them to torture." | 11:44 |
Mariu | that happens when they get burried | 11:46 |
@kanzure | i'm not even sure where to begin with the list of reasons why they caring about that is useless. | 11:50 |
@kanzure | bkero: i recommend not bothering with lesswrong | 11:50 |
bkero | kanzure: I was just amused by that last part. | 11:51 |
@kanzure | huh, the petition about open access policies got a response | 11:55 |
@kanzure | https://petitions.whitehouse.gov/response/increasing-public-access-results-scientific-research | 11:55 |
@kanzure | "To that end, I have issued a memorandum today (.pdf) to Federal agencies that directs those with more than $100 million in research and development expenditures to develop plans to make the results of federally-funded research publically available free of charge within 12 months after original publication." | 11:56 |
@kanzure | haha "and the need to ensure that the valuable contributions that the scientific publishing industry provides are not lost." | 11:56 |
@kanzure | michael nielsen replied, http://news.ycombinator.com/item?id=5266065 | 11:57 |
@kanzure | peter suber, https://plus.google.com/109377556796183035206/posts/8hzviMJeVHJ | 11:57 |
-!- sheena1 [~home@174.4.134.159] has quit [Quit: Leaving.] | 11:58 | |
-!- sheena1 [~home@174.4.134.159] has joined ##hplusroadmap | 11:58 | |
@kanzure | directive: http://www.whitehouse.gov/sites/default/files/microsites/ostp/ostp_public_access_memo_2013.pdf | 11:58 |
-!- sheena1 [~home@174.4.134.159] has quit [Client Quit] | 12:01 | |
-!- sheena1 [~home@174.4.134.159] has joined ##hplusroadmap | 12:02 | |
@fenn | roko's basilisk is like a fucked up version of pascal's wager | 12:16 |
@kanzure | obviously the correct strategy would be index poisoning, not data erasing | 12:18 |
@kanzure | but the whole point is messed up anyway | 12:18 |
-!- sheena1 [~home@174.4.134.159] has quit [Ping timeout: 244 seconds] | 12:18 | |
@fenn | it reminds me of the baptist hellfire preachers on TV | 12:19 |
ParahSailin | wait holy crap, did they actually do something with a petition? | 12:19 |
Vicarious | ohai | 12:19 |
ParahSailin | learning about the whole roko's basilisk thing is when i first realized there was something horribly wrong with LW and EY | 12:20 |
@fenn | was it really? i would have thought it was the "try to take over the world before anyone else does so we can institute eternal totalitarian friendliness" | 12:20 |
@fenn | anyway, the OSTP petition response is the most useful thing i've seen in a while | 12:21 |
ParahSailin | well that was the first thing that i personally heard of | 12:21 |
@kanzure | fenn: no, the problem is that that concept looks like a good idea if its your first-ever exposure to the concepts. | 12:21 |
@fenn | it just doesn't seem like a very good plan in the first place | 12:24 |
@kanzure | ParahSailin: yeah, this seems like positive response to a petition. i am particularly surprised because people were saying that there were some publisher shills on OSTP or something, and that they would block it. | 12:24 |
ParahSailin | is that the first ever actual response to a petition? | 12:25 |
@fenn | what does "undermine any right under the provisions of title 17 or 35 | 12:26 |
@fenn | USC" mean? | 12:26 |
@kanzure | USC means US Code.. look it up i guess. | 12:27 |
@kanzure | http://www.copyright.gov/title17/ | 12:27 |
@fenn | title 17 is the law concerning copyright, and 35 is patents | 12:27 |
@fenn | so does it just mean "this doesn't actually nullify copyright law or anything crazy" | 12:27 |
@kanzure | it's hard to tell. it might mean that the publishers get to keep a proprietary copyright to the article, separate to the public domain version? | 12:28 |
ParahSailin | i guess this is maybe a pressure relief valve after their doodz murdered one of our doodz | 12:28 |
@fenn | or mayb the president just doesn't give a fuck about publishers | 12:29 |
@kanzure | i doubt this was the president's decision to come up with this response | 12:30 |
@fenn | i have no idea how any of this actually works | 12:30 |
@kanzure | http://www.whitehouse.gov/sites/default/files/microsites/ostp/ostp_public_access_memo_2013.pdf | 12:31 |
@fenn | ugh reading government speak makes my brain hurt | 12:33 |
ParahSailin | what do you think the chances of a random bag of soybeans from chinatown's gonna be monsanto roundup ready seeds | 12:35 |
sseehh_ | http://www.foxnews.com/politics/2013/02/22/nasa-missile-defense-tech-leaked-to-china-sources-say/#ixzz2Lex4wLcN | 12:36 |
@fenn | nasa doesn't do missile defense, wtf are they on about | 12:39 |
@kanzure | sseehh_: i would appreciate it if you stop linking to these terrible things | 12:39 |
sseehh_ | kanzure: im sorry maybe someone else can | 12:40 |
strangewarp | oh Roko's Basilisk. I think you'd have to have an incredibly cynical vision of the set of possible superhuman minds arising under possible topologies to believe that exists at any worrisome magnitude ... | 12:42 |
@kanzure | but even if it did exist, so what? | 12:45 |
@fenn | oh interesting, i know will marshall, he runs the rainbow mansion talks | 12:46 |
strangewarp | Well, if you're a cognitive materialist who is extremely cynical about the mind-space of possible AGI, then it's worrying. But I'm only the former, not the latter. | 12:46 |
@kanzure | god i hate what this channel has become | 12:47 |
-!- sseehh_ was kicked from ##hplusroadmap by kanzure [sseehh_] | 12:47 | |
-!- Mariu was kicked from ##hplusroadmap by kanzure [Mariu] | 12:47 | |
-!- Mariu [Jimmy98@89.41.57.33] has joined ##hplusroadmap | 12:47 | |
-!- mode/##hplusroadmap [+q Mariu!*@*] by kanzure | 12:48 | |
strangewarp | hahaha Yudkowsky is really terrifiedof Roko's Basilisk, wow | 13:04 |
@fenn | careful, admitting knowledge of the basilisk entitles it to punish you | 13:06 |
@fenn | oh noes i did it | 13:07 |
strangewarp | oh nooooo | 13:07 |
@fenn | now i have to delete all my profilez | 13:07 |
@fenn | goodbye cruel internet | 13:08 |
strangewarp | Resimulation based on derivations from online profiles / logs would be causal trade, not acausal trade, anyway. And.. I seem to be being a pedant, because that doesn't change the point much. Whoever wrote this rationalwiki page should be tapped across the knuckles though... | 13:20 |
strangewarp | Oh, well, if they meant it would resimulate every possibe mind that knows about the Basilisk, then that would be acausal trade, but ... I don't care, I'm so over this | 13:21 |
@fenn | the basilisk appreciates your comments | 13:21 |
strangewarp | heh | 13:22 |
@fenn | http://xkcd.com/380/ | 13:24 |
-!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Quit: Leaving] | 13:32 | |
@kanzure | strange, i have to "save" in bios before linux will boot up past "Waiting for /dev to be fully populated" | 13:34 |
@kanzure | "What's needed is a single day of coordinated action. Where all scientists and researchers move to free publications in one big move." | 13:36 |
@kanzure | i somehow doubt that could be coordinated. | 13:36 |
jrayhawk | It might just be a factor of boot time; the idea is to wait for controllers to boot and settle. | 13:37 |
@kanzure | no, it never boots unless i go through bios and "save changes" (there were no changes) | 13:37 |
@kanzure | saurik: do you have a google glass device? i'm interested in getting my hands on one to start poking around at the internals. | 13:39 |
@kanzure | i mean, erm, the google scouter. | 13:40 |
@fenn | it could be a bad battery? | 13:43 |
-!- sylph_mako [~mako@103-9-42-1.flip.co.nz] has joined ##hplusroadmap | 13:43 | |
@kanzure | battery lasts forever (4 hours; could be better, but whatever) | 13:44 |
@kanzure | i'm not sure what google scouter is actually powered by, i thought it was probably bluetooth to an android phone that does most of the work | 13:45 |
@kanzure | "“When we ship this, we will have a cloud-based API that will allow developers to integrate with Glass, which enables a wide variety of Glass services while keeping a consistent user experience” Parviz confirms. “It’s the same API that we used to build the e-mail and calendar services that we test on Glass."" | 13:47 |
@kanzure | that sounds terrible. | 13:47 |
@kanzure | "With those APIs, developers will be able to deliver select data to a Glass user, rather than overwhelming them with all the information that might fit onto a typical smartphone screen. " | 13:47 |
@kanzure | overwhelming them? you mean make them die from dehydration, surely. | 13:48 |
@kanzure | i was thinking that if there's no api yet, it would be better to just give developers something to work with in the mean time, and then work on actual compatibility when the sdk emerges | 14:00 |
@fenn | i mean a bad bios battery, so your bios is uninitialized or somehow set wrong when you boot up, instead of whatever state the rom saves when you do save and exit | 14:03 |
@kanzure | this also happens when it sleeps, if that's helpful information. | 14:04 |
@fenn | personally i avoid google's interface whenever possible, i don't see that changing just because it's on a HUD. but hey i guess someone likes pointless whitespace | 14:05 |
-!- etheros [~etheros@2601:0:8980:7d:92e6:baff:fe83:8bff] has quit [Read error: No route to host] | 14:06 | |
@kanzure | i'm trying to figure out who will be the first to get facial recognition working | 14:06 |
@fenn | facebook, obviously | 14:08 |
@kanzure | i see a lot of stuff about iphone apps that are able to recognize your friends... but who cares? you probably have <500 friends and i bet they look different. not useful. | 14:09 |
@fenn | what i'm getting from this video is that google glass will make me a skydiving trapeeze artist fashion model and parent who likes to do ice sculpture as a hobby | 14:10 |
@kanzure | your mistake is assuming that you should bother to watch a marketing video | 14:10 |
@fenn | i'm wondering if it does voice recognition onboard, is it totally useless without an internet connection? | 14:11 |
@kanzure | it seems that nobody has details out yet | 14:12 |
@kanzure | so, assume the worst | 14:12 |
@fenn | worst in this case requires a powerful imagination | 14:13 |
@fenn | Explorers will each need to pre-order a Glass Explorer Edition for $1500 plus tax and attend a special pick-up experience, in person, in New York, San Francisco or Los Angeles. | 14:14 |
@fenn | there's no way it costs them that much to make the beta models | 14:14 |
@kanzure | how about.. on-board gsm, no wifi, only wireless charging, no sdk, all interactions filtered through google.com, software is installed remotely (if at all) | 14:14 |
@kanzure | nah google is just bad at pricing | 14:14 |
@fenn | they didnt charge anything for the chrome books | 14:15 |
@fenn | it's essentially the same thing with a smaller battery and no keyboard :P | 14:15 |
@kanzure | they are trying to use price as a marketing signal or something | 14:15 |
@kanzure | "it's a premium product" etc. | 14:15 |
@fenn | i guess if you can't afford that you can't afford to go skydiving or whatever | 14:16 |
@kanzure | i'm pretty sure skydiving is $50ish per jump? | 14:16 |
@kanzure | "If you are looking to just make one tandem jump then you are looking at about $200 for the jump and $50 to $100 for video." | 14:16 |
@fenn | whatever happened to companies selling products to customers | 14:18 |
@fenn | why can't i just buy a fucking LCOS display | 14:18 |
@kanzure | it would be really easy to sell a person recognition system to anyone in sales or politics | 14:19 |
@fenn | not if it makes you look like a dork | 14:20 |
@kanzure | the problem can be simplified by only recognizing individuals that are using the same device/service (based on gps/their location data, instead of the weird way that 'bump' requires you to meet with people) | 14:20 |
@kanzure | people would pay more for recognizing people "out of network" | 14:21 |
@kanzure | well, if it makes them look like a dork, maybe the pirate patch will catch on again. who knows. | 14:22 |
@fenn | if people are using the device you can just use bluetooth or wifi or whatever | 14:22 |
@fenn | no need for facial recognition in that case | 14:23 |
@kanzure | oh, wifi fingerprinting to find someone in a crowd. that would be neat. | 14:24 |
@kanzure | or i mean, for two individuals to find each other in a crowd. | 14:24 |
@fenn | dear google i plan to mount my glass to my AR-15 babykiller and ask for directions to the nearest elementary school | 14:25 |
@kanzure | still, the usefulness of that system is really only when it can 1) identify people you've never met, 2) identify people who aren't signed up with whatever awful thing you're using, and 3) work offline at least a little bit. | 14:25 |
@fenn | i dont think recognizing out of network faces offline is feasible or to be expected | 14:26 |
@kanzure | i think it is feasible (in a "it would take a tremendous amount of work, but it would be possible" way) for when you know where you are going in advance (like a conference) that has a public listing of who the hell is attending. | 14:27 |
@fenn | nah, maybe in five years we'll have the flops/joule but as it is, just detecting the faces takes all the CPU you can throw at it | 14:27 |
@fenn | in a crowd at least | 14:27 |
@fenn | also consider that nobody tags faces when they're pointing away from the camera | 14:28 |
@fenn | so you'd essentially be forced to use data gathered from on-device head tracking | 14:29 |
@fenn | otherwise how would you ever know what so and so looks like from the side | 14:29 |
@kanzure | even with a fast internet connection, person recognition is going to take at least a few seconds.. it would be awkward to just be sitting there making small talk until you remember who the hell a person is. | 14:29 |
@fenn | i dont see why it should take a few seconds | 14:29 |
@kanzure | yeah i haven't thought about data gathering. it would be nice to have facebook/linkedin's data, but i don't have that. | 14:29 |
@kanzure | because the minimum ping to google is 50ms or something stupid (9ms if you're on a good network). add up the roundtrip cost.. | 14:30 |
@fenn | that's still less than a tenth of a second in network overhead | 14:30 |
@kanzure | well, let's add it up? | 14:30 |
@fenn | this is sort of like asking "how long is a string" | 14:32 |
@kanzure | i think an estimate is a reasonable thing to do | 14:32 |
@kanzure | let's say that you turn your head in a direction and see a person 3 feet from you, with a beautiful smile. | 14:32 |
@fenn | it depends on how many faces you're searching through, how they're sorted/hashed, how many nodes you have to compute with, what the data rate is between user, server, server, and god knows what API calls you have to make | 14:32 |
@kanzure | first frame in the mpeg stream is done being collected in, uh, some amount of time (i don't even know what's reasonable here.. 10 ms?) | 14:33 |
@fenn | opencv can do real-time face detection, so 1/30 s = 33ms | 14:33 |
@kanzure | no that's detecting whether or not a face exists in the photo | 14:33 |
@fenn | correct | 14:33 |
@fenn | you have to detect a face before uploading it | 14:34 |
@fenn | otherwise you're just constantly uploading crap | 14:34 |
@kanzure | what about just uploading everything? i dunno if that would be better. | 14:34 |
@fenn | hm. seems like you'd lose somehow on resolution/compression | 14:35 |
@fenn | also consider the power requirement of constantly sending data over wifi | 14:35 |
@kanzure | phones are already doing that and i think battery technology is going to improve | 14:35 |
@fenn | phones are much better at downloading than uploading because you can have gigantic grid powered antenna arrays at the telco station | 14:36 |
@kanzure | i mean over wifi | 14:36 |
@fenn | same principle applies | 14:36 |
@kanzure | also, even just 1 hour of this system at full throttle would be very neat. | 14:36 |
@fenn | sure, but what's the point of trying to recognize the backs of peoples' heads? | 14:37 |
@kanzure | that's where the barcode is, you just have to look harder.. | 14:37 |
-!- klafka [~klafka@204-16-157-18-static.ipnetworksinc.net] has quit [Remote host closed the connection] | 14:37 | |
saurik | kanzure: no | 14:37 |
@fenn | too bad the paranoid schizos aren't right more often, this would be so much easier if we all had the mark of the beast | 14:38 |
-!- klafka [~klafka@204-16-157-18-static.ipnetworksinc.net] has joined ##hplusroadmap | 14:38 | |
@kanzure | saurik: ok, thanks. | 14:38 |
-!- klafka [~klafka@204-16-157-18-static.ipnetworksinc.net] has quit [Read error: Connection reset by peer] | 14:38 | |
-!- klafka [~klafka@204-16-157-18-static.ipnetworksinc.net] has joined ##hplusroadmap | 14:39 | |
@kanzure | fenn: i guess even if the full round trip was 1 or even 2 seconds, that would still be better. the majority of the time, i don't know a person. i know only a fraction of the 7 billion people, so this system would be a few billion times better than the current situation. :P | 14:39 |
@fenn | maybe of interest, i have this file i'm not sure where it came from, possibly steve mann? shows a highly edited stream of faces from some computer vision conference: http://fennetic.net/irc/face001.mp4 | 14:40 |
@fenn | turn down the sound before watching | 14:40 |
@fenn | hm probably not steve mann because the camera isn't at eye level | 14:43 |
@fenn | probably from this thing http://www.robots.ox.ac.uk/~bob/research/research_wearables.html | 14:55 |
@kanzure | network connections are usually congested during conferences anyway, hrm | 15:00 |
@fenn | assuming you could compute the face matrix (set of eigenfaces) on-board, it wouldn't take that much storage space for all the people in the world | 15:03 |
@fenn | say each face has 100 bytes of uniquely identifying features, that's "only" 800GB | 15:04 |
archels | http://www.kurzweilai.net/2013-world-congress-in-computer-science-computer-engineering-applied-computing | 15:05 |
archels | wait, I thought WORLDCOMP had been called out as bullshit? | 15:05 |
@fenn | looks like real life systems require about 50 bytes per face | 15:06 |
@kanzure | archels: yes, it's bullshit. so is kurzweilai.net.. | 15:07 |
@fenn | eigenfaces are the stuff of nightmares | 15:09 |
archels | haha | 15:10 |
* archels sleeps | 15:10 | |
rigel | i saw a great twitter thing today | 15:13 |
rigel | "transhumanism wants to be scientology when it grows up" | 15:13 |
@kanzure | rigel: that's not very productive | 15:15 |
@fenn | are you a troll | 15:15 |
ThomasEgi | scientrollogy ? | 15:15 |
@fenn | operation clambait | 15:16 |
ThomasEgi | is trollscience involved? | 15:16 |
rigel | i know, i just thought it was funny | 15:23 |
strangewarp | rigel: Found the tweet in question. Not surprised that the guy behind it namechecks Dale Carrico in a positive way in the same line. | 15:35 |
strangewarp | Dale Carrico epitomizes everything that's wrong with the reform-before-technology crowd. | 15:37 |
strangewarp | blah. | 15:37 |
@fenn | this is the "feed the starving children in africa before going to the moon" argument? | 15:40 |
strangewarp | basically yeah. Also he's decided the transhumanist project, in every form, is a cult, because .. something something serious grown-up intellectual with realistic concerns something something. | 15:41 |
@fenn | he likes to use big words | 15:42 |
strangewarp | I'd say he's a comedic study in the power of signalling, except so many people take him seriously, ugh. | 15:43 |
Urchin[emacs] | those are the problems of treating X as a literary genre | 15:44 |
-!- EnLilaSko [~Nattzor@unaffiliated/enlilasko] has quit [Remote host closed the connection] | 15:44 | |
Urchin[emacs] | whatever X is (unless it's an acutal literary genre, of course) | 15:44 |
Urchin[emacs] | though I have encountered cultic transhumanists | 15:45 |
@fenn | every group of humans larger than three is a cult, get over it | 15:46 |
@kanzure | cult of one | 15:46 |
Urchin[emacs] | no, it's not | 15:46 |
Urchin[emacs] | cults have a fairly consistent set of traits that mark them as such | 15:48 |
Urchin[emacs] | so not every group fits, far from it | 15:50 |
@fenn | wikipedia defines it as a "group whose beliefs or practices are considered abnormal or bizarre by the larger society" | 15:50 |
Urchin[emacs] | though you will find a crazy worshiper in just about any group | 15:50 |
strangewarp | ...Hm. Maybe this is why Yudkowsky so vehemently opposes Roko's basilisk, even though he says it's unlikely. He doesn't want people panicking about sin and giving him every last cent of their money. | 15:50 |
@fenn | i'm sure yudkowsky wouldn't mind if people gave him every last cent. the problem with the basilisk is it's too similar to russel's teapot; it exposes the infinite AI boogeyman for being the bad argument that it is | 15:51 |
strangewarp | Oh indeed, I'm not saying he would be doing it out of altruism, I'm saying it would look terrible to an outside observer if people were dumping all their cash on him out of fear. | 15:52 |
@fenn | that's already how it looks to an outsider | 15:52 |
strangewarp | Well, explicit fear, anyway. Low-level fear is already the case.. yeah | 15:52 |
Urchin[emacs] | giving money is one thing, but a guy who castrates himself because singularity is near is a completely different thing (and it has happened) | 15:54 |
@fenn | people don't donate to SIAI (or whatever it's called now) because they think it's the best way to make magical AI overlords appear, they do it out of fear that evil AI overlords will appear if they don't | 15:54 |
* strangewarp nods | 15:55 | |
Urchin[emacs] | I'm not sure that that's the case | 15:55 |
@fenn | Urchin[emacs]: link? | 15:55 |
Urchin[emacs] | I do know some people in a neighboring country that donated to siai but they were singularitarians interested in benefit to themselves | 15:56 |
Urchin[emacs] | I have not donated | 15:56 |
Urchin[emacs] | fenn: do you have a study on the reasons for donating to SIAI yourself? | 15:57 |
@fenn | please provide further references to above anecdote regarding castration and the singularity | 15:57 |
Urchin[emacs] | the stuff about castaration was on the old betterhumans site (might still exist in some kind of archive somewhere, but the site is defunct for years) | 15:59 |
juri_ | i'm a singularist (in a loose sense, also a taoist). and i just don't want kids. | 15:59 |
klafka | https://petitions.whitehouse.gov/response/increasing-public-access-results-scientific-research | 16:00 |
juri_ | nasty little buggers would always be getting in the way of my crazy projects. | 16:00 |
Urchin[emacs] | http://posthumanblues.blogspot.com/2006/03/how-my-castration-relates-to.html | 16:01 |
juri_ | 'MOM! little jimmy's got himself stuck in the vacuum chamber again!' | 16:01 |
juri_ | "good. maybe now he'll understand what i feel like when my blood is boiling." | 16:01 |
Urchin[emacs] | wait, again? | 16:01 |
juri_ | I've owned two chambers in my life, and am building a third. by the time my kids could talk, i'm imagining my house would be filled with the things. | 16:03 |
@fenn | hey we know how to use the internet here http://web.archive.org/web/20100524154905/http://www.betterhumans.com/blogs/cybert/archive/2006/03/20/5117.aspx | 16:03 |
juri_ | but, it would be filled with less of them, if i had kids. and that's a problem. | 16:03 |
@fenn | i love that "marduk" is posting on that thread | 16:05 |
-!- nsh [~nsh@wikipedia/nsh] has joined ##hplusroadmap | 16:06 | |
@fenn | oh wait, i'm thinking "zardoz" | 16:07 |
-!- strophariad [~str@unaffiliated/strophariad] has joined ##hplusroadmap | 16:10 | |
strangewarp | Well, I might pursue castration eventually, but that would be for gender-identity reasons seperate from my transhumanist concerns. | 16:10 |
@fenn | well anyway i would say castration is about as transhumanist as lepht_anon's hand-stabbing (which is to say, not at all) | 16:10 |
Urchin[emacs] | I sort of like my genitals | 16:14 |
Urchin[emacs] | it's the whole being human that I have a problem with | 16:14 |
strophariad | Gigolo Joe? | 16:21 |
@fenn | bleh i give up. if anyone finds a survey explaining why people have donated to SIAI let me know | 16:27 |
* strophariad would rather donate to Roger Williams: http://localroger.com/prime-intellect/ | 16:45 | |
@fenn | there's always the math mines, for when you run out of war, famine, etc | 16:47 |
@fenn | oh noes the internet doesn't know my meme | 16:47 |
@fenn | mathematics research, called "truth mining" in the future. Especially in this "new world" where everyone is literally a computer, one might expect that math research would have been completely automated to the point that it does not involve consciousness or creativity. However, at least according to the wise old virtual being in the book, this will never be the case. Math research requires some | 16:48 |
@fenn | intuition, and discovery requires having a different outlook than those who came before you. | 16:49 |
@fenn | strophariad: is this prime intellect really worth reading? | 16:51 |
strophariad | that reminds me of hearing a few years ago how some group mined the news-tickers at the bottom of CNN, FOX, etc., claiming to have uncovered classified military operations or somesuch. tho I imagine language-processing will be just as important as raw number-crunching. | 16:51 |
@fenn | data mining is not the same thing at all | 16:52 |
strophariad | I'm very glad I read it, tho I don't know the singularity-ficiton genre as well as I imagine many here do. it's very dark, requiring at least as much suspension of moral distaste as it does disbelief. blurs the line between utopia and dystopia in some thoughtful ways. | 16:53 |
strophariad | just saying X reminds me of Y, how my mind tends to work, or not :-) | 16:54 |
@fenn | you should read some greg egan if you liked it | 16:54 |
strophariad | sounds interesting, based on his wikipedia description | 16:55 |
@fenn | there's the general problem of trying to communicate the day to day existence and thoughts of acorporeal beings | 16:56 |
strophariad | or even to recognize them, I suspect | 16:56 |
@fenn | so the genre will necessarily return to the crutch of the humanoid narrator | 16:56 |
strophariad | ah | 16:57 |
@fenn | but so far this "prime intellect" is not much better than Siri or the talking paperclip | 16:57 |
strophariad | well, tMMoPI has a third-person omniscient narrator, and the main character's foremost motivation is her resentment of the supreme being | 16:58 |
rigel | oh, just so we're clear, i share some goals with transhumanists but an quite opposed to others. I also think that the argument you've characterized carrico as making does carry some weight. certainly it's not the whole ballgame. | 16:58 |
rigel | also, everything ray kurzweil says outside of a narrow engineering framework is garbage | 16:59 |
rigel | and in 30 years people will point and laugh at his naivete | 16:59 |
rigel | (they already do) | 16:59 |
rigel | but hes also an easy target | 16:59 |
@fenn | but it's such great television | 16:59 |
@fenn | why be right when you can be entertaining | 17:00 |
strophariad | Terence McKenna knew that well, and bowed out gracefully before he could be proven wrong :-) | 17:00 |
rigel | terence was more of a bard than a scientist | 17:00 |
rigel | like a 90/10 kind of "more" | 17:01 |
strophariad | yes | 17:01 |
rigel | i have certainly tripped the fuck out on some shit he was saying because holy crap that shit is heavy man | 17:01 |
strophariad | He was an ethnobotanist, a psychedelics connoisseur, a technoshaman futurist, a lecturer and raconteur. | 17:01 |
rigel | but it's also garbage | 17:01 |
strophariad | a myth-maker :-) | 17:02 |
rigel | its feel-good shamanism | 17:02 |
rigel | pop shamanism | 17:02 |
rigel | which is fine i guess, until you have people like fucking sting going on about ayahuasca and tantra | 17:02 |
@fenn | sting! | 17:02 |
strangewarp | yeah, it would be comfortable if Kurzweil were right, but he only seems to be about 35% right on a regular basis, judging from an analysis LessWrong did of his various predictions... (I don't have much love for LessWrong either, mind, but that post wasuseful) | 17:02 |
strangewarp | *was useful | 17:03 |
strophariad | well, sort of feel-good, if telling people they can expect the species to become another smear in the shale without something like what he hoped would happen is popular or uplifting | 17:03 |
-!- strangewarp [~C@c-67-176-51-26.hsd1.co.comcast.net] has quit [Remote host closed the connection] | 17:03 | |
-!- strangewarp [~C@c-67-176-51-26.hsd1.co.comcast.net] has joined ##hplusroadmap | 17:03 | |
-!- yashgaroth [~ffffff@cpe-66-27-118-94.san.res.rr.com] has joined ##hplusroadmap | 17:04 | |
rigel | i mean, you take a look at his shit about the king wen series of hexagrams and tell me that's not either a gigantic elaborate put-on or the ramblings of someone fairly well brain damaged | 17:05 |
rigel | probably both | 17:05 |
rigel | (terence, i mean) | 17:05 |
strophariad | probably both. he spoke a lot of bullshit, probably without even fully recognizing it as such, but it was some of the most erudite silver-tongued imaginative bullshit I've encountered :-) | 17:06 |
rigel | the used car salesman-cum-guru archetype, he certainly was that | 17:06 |
-!- klafka [~klafka@204-16-157-18-static.ipnetworksinc.net] has quit [Remote host closed the connection] | 17:07 | |
-!- klafka [~klafka@204-16-157-18-static.ipnetworksinc.net] has joined ##hplusroadmap | 17:08 | |
strophariad | even as he told people to think for themselves and despised the "white man with all the answers at the front of the room" image | 17:08 |
strangewarp | I think it's interesting that Google hired him to manage a development team for a long-shot project of some sort, though. While he's kooky, I'm interested to see what they accomplish. | 17:10 |
@fenn | are you talking about kurzweil or mckenna? | 17:11 |
strangewarp | Kurzweil | 17:11 |
* strophariad awaits more from project glass | 17:11 | |
* fenn awaits nothing surprising from project glass | 17:11 | |
-!- klafka [~klafka@204-16-157-18-static.ipnetworksinc.net] has quit [Ping timeout: 248 seconds] | 17:12 | |
@fenn | maybe they'll even come up with an augmented reality cookbook, for DIY cooking! | 17:12 |
strangewarp | At some point I need cheap HUD glasses of some sort with a user-friendly graphics API, just so I don't have to be looking down-and-diagonally at my custom music software on a computer screen while singing.. blah :P | 17:13 |
strophariad | they got my hopes up somewhat, having been imagining the transformative possibilities of AR since first hearing about Steve Mann ard 2003. as limited as the vision in their promotional videos are, they're still the most inspiring I've seen. | 17:13 |
strophariad | in contrast, a couple years ago when Intel made a press-release about their new AR lab, it was even more hum-drum. "here's a gadget that will monitor your fridge and tell you what recipes you can make with the food you have!" | 17:14 |
@fenn | strangewarp: have you considered a projector system like sixth sense? | 17:15 |
* strangewarp googles that... | 17:15 | |
@fenn | you'll probably want to use a camera to do hand tracking anyway | 17:15 |
Urchin[emacs] | I want hud glasses that correct my eyesight and conceal my eyes | 17:15 |
strangewarp | Oh yeah, I already control it with a custom instrument, it just has an additional GUI on the computer screen that I'd much prefer to pipe through to a HUD of some kind | 17:16 |
Urchin[emacs] | I have looked into doing something like Steve Mann's stuff, but I don't have much skill with electronic circuts and stuff like that | 17:18 |
@fenn | there aren't very many wide angle displays available, and those that exist are pretty low resolution | 17:19 |
@fenn | even the "oculus rift" will be quite pixellated | 17:19 |
strangewarp | The Oculus Rift would be more than enough for my purposes, heh | 17:20 |
@fenn | i'm pretty pissed that there's still no consumer virtual retinal display | 17:20 |
@fenn | also the low light performance of small cameras leaves a lot to be desired | 17:20 |
strangewarp | Still, I've got to find something that doesn't totally obscure vision, since I'd of course be using it on stage, and looking at the lights and dials on my instruments as well.. | 17:21 |
@fenn | larger cameras tend to be expensive and hard to interface with | 17:21 |
@fenn | strangewarp: why not just get an old myvu | 17:21 |
* strangewarp googles that too... | 17:22 | |
Urchin[emacs] | I found the exeamples on the wiki page interesting: http://en.wikipedia.org/wiki/Eyetap | 17:25 |
@fenn | i want to know where to get a miniature "heat sensor" | 17:26 |
sylph_mako | fenn, fuck yes, VRDs are what I'm hanging out for. | 17:30 |
sylph_mako | Any idea what's going on with that these days? | 17:30 |
@fenn | microvision licensed the tech to brother (of inkjet printer fame) who have been farting around doing nothing significant as far as i can tell | 17:31 |
@fenn | i'm considering building my own from scratch | 17:31 |
@fenn | it's really rather simple technology as long as you dont have to build your own RGB laser from scratch | 17:32 |
@fenn | fortunately for us the TI pico projector has one premade | 17:32 |
@fenn | monochrome will do fine though | 17:33 |
@fenn | here's one way to do it: www-personal.umich.edu/~awtar/PHD/TLO_scanner1.pdf | 17:34 |
@fenn | here's a much more low tech way to do it http://laserpointerforums.com/f47/homemade-scanning-laser-color-projector-tv-32628.html | 17:35 |
sylph_mako | Hahahah. are you sure this is safe. | 17:36 |
@fenn | the part i dont know how to deal with yet is that the eye doesn't orbit around the pupil, so the pupil actually moves off axis | 17:36 |
@fenn | it might be as simple as projecting onto a half silvered concave mirror | 17:36 |
sylph_mako | One of the things I find exciting about this technology is that you end up packaging on eye-tracking along with it- well, if that didn't work. | 17:37 |
@fenn | you definitely want a safety interlock in case the mirror stops moving, but that's fairly simple to implement. | 17:37 |
Urchin[emacs] | fenn: the diagram on wiki is actually for night vision goggles | 17:38 |
@fenn | alright i can't read this crap anymore, they're giving serious consideration to asimov's laws of robotics | 17:40 |
@kanzure | for whatever it's worth i have a full history of most donations to siai from 200x-2009ish | 17:40 |
@kanzure | if anyone wants that data (off the record) | 17:40 |
@fenn | does it have a comments field attached? | 17:42 |
@kanzure | fenn: the reason you should be interested in glass is because you will have more friends writing software for you. | 17:44 |
@kanzure | rigel: you are grossly misinformed if you think that kurzweil can actually do any of those things. | 17:46 |
@kanzure | http://www.breakthroughprizeinlifesciences.org/ | 17:48 |
@kanzure | "The Breakthrough Prize in Life Sciences is founded by Art Levinson, Sergey Brin, Anne Wojcicki, Mark Zuckerberg and Priscilla Chan, and Yuri Milner to recognize excellence in research aimed at curing intractable diseases and extending human life" | 17:48 |
strangewarp | I'll probably eventually end up using Glass (or a Glass knockoff, or possibly a secondhand MyVu) and a RasPi + USB hub on stage, instead of a laptop... hmm | 17:49 |
@kanzure | neat they game money to yamanaka | 17:49 |
@kanzure | *gave | 17:49 |
@kanzure | i don't recognize all of the names on http://www.breakthroughprizeinlifesciences.org/laureates | 17:50 |
@kanzure | fenn: no it does not have a comments field | 17:51 |
@fenn | "Lawrence realized that he had not really created Prime Intellect to make the world a better place. He had created it to prove he could do it, to bask in the glory, and to prove himself the equal of God. He had created for the momentary pleasure of personal success, and he had not cared about the distant outcome." | 17:54 |
@fenn | le sigh | 17:54 |
-!- aristarchus [~aristarch@unaffiliated/aristarchus] has joined ##hplusroadmap | 17:54 | |
@fenn | kanzure you are assuming they will let people write native apps | 17:54 |
strophariad | fenn: not that motivation you would have given him? | 17:55 |
strophariad | -that +the | 17:55 |
@fenn | strophariad: it's just bad writing, ignores "show dont tell" | 17:55 |
strophariad | yes, it's very first-drafty, but still thoughtful, imo, if not as well-crafted as it could be | 17:55 |
@fenn | and the whole faster than light speed chip thing is just unnecessary | 17:56 |
strophariad | and iirc, it does do some showing later, a chapter devoted to the design and emergence of Prime Intellect | 17:59 |
@kanzure | fenn: yes, and even if they don't, there should be a free software effort. | 17:59 |
@fenn | i haven't seen shit for free software on android | 18:01 |
@fenn | there's more free software for the amiga | 18:01 |
@fenn | maybe i'm spoiled | 18:02 |
@kanzure | linux installs just fine | 18:08 |
strophariad | I'd be happy with BASIC on my dumbphone. | 18:08 |
@kanzure | why would you buy a crippled phone? that seems silly. | 18:08 |
@kanzure | me ranting about sourceforge http://news.ycombinator.com/item?id=5268220 | 18:14 |
strophariad | dumbphone = cripped phone? | 18:14 |
strophariad | *crippled | 18:14 |
-!- SphericalMouse [62cfd610@gateway/web/freenode/ip.98.207.214.16] has joined ##hplusroadmap | 18:15 | |
SphericalMouse | sup? | 18:16 |
@kanzure | strophariad: yes | 18:17 |
@kanzure | strophariad: geek things about myvu, i guess. | 18:17 |
@kanzure | oops, wrong person | 18:18 |
strophariad | kanzure: oh. well, my facetious remark abt Basic aside, it does what I need it to do, and I don't see a point in spending upwards of three figures a month for more bells and whistles | 18:18 |
-!- klafka [~klafka@c-24-6-18-31.hsd1.ca.comcast.net] has joined ##hplusroadmap | 18:18 | |
strophariad | well, there you are, anyway :-) | 18:18 |
@kanzure | strophariad: smartphones cost <$50 these days. | 18:19 |
@kanzure | i think you're bad at math | 18:19 |
strophariad | and the plan? the internet-access that would be the only reason to have one? | 18:19 |
@kanzure | these phones tend to be wifi capable | 18:19 |
strophariad | and? | 18:20 |
@kanzure | those 3 figure plans are just for people who want to pay that, the phones still work without those plans -_- | 18:20 |
@kanzure | you can use totally different plans if you want.. why would the phone care? | 18:20 |
strophariad | k, so my phone is crippled for not having wifi access? a keyboard? | 18:20 |
@kanzure | these phones are bluetooth capable too, so you just use your regular keyboard | 18:21 |
@kanzure | i think you're making this harder than it is | 18:21 |
strophariad | so are many dumbphones | 18:21 |
strophariad | incl mine | 18:21 |
strophariad | not that I need it | 18:21 |
@kanzure | but why would you want a phone that doesn't run linux? | 18:21 |
SphericalMouse | ~lolz~ | 18:22 |
strophariad | heh, why would I want one that does? or a laptop, for that matter? what has having linux on your phone gained you? | 18:22 |
@kanzure | millions of man hours of industrial civilization at my fingertips | 18:22 |
@kanzure | and package management | 18:22 |
SphericalMouse | frankly, if a phone can't shitpost for you automatically, what is the point? | 18:23 |
@kanzure | shitwhat? | 18:23 |
-!- klafka [~klafka@c-24-6-18-31.hsd1.ca.comcast.net] has quit [Ping timeout: 256 seconds] | 18:23 | |
SphericalMouse | Are you claiming to be unaware of shitposting? | 18:24 |
@kanzure | trolling? | 18:24 |
@kanzure | you can do that from a dumbphone too, i don't see the point | 18:24 |
SphericalMouse | trolling is different than shitposting | 18:24 |
@fenn | please enlighten us savages | 18:26 |
SphericalMouse | http://forums.somethingawful.com/dictionary.php?act=3&topicid=2362 | 18:26 |
@fenn | wants me to register | 18:27 |
SphericalMouse | hmm | 18:27 |
@fenn | urban dictionary defines it as, "In general, being a bad person." | 18:27 |
@fenn | how is this different from trolling again? | 18:28 |
SphericalMouse | shitposting is low content but offensive posting. trolling is more designed to provoke someone into enraged response. | 18:28 |
@fenn | sounds like it's just a lame excuse for not being good at trolling | 18:28 |
@kanzure | i demand an honorary somethingawful membership | 18:29 |
SphericalMouse | they serve different purposes | 18:29 |
bkero | I have actual somethingawful membership. | 18:29 |
SphericalMouse | pay your 10bux | 18:29 |
yashgaroth | bkero can I touch it | 18:29 |
bkero | no | 18:29 |
-!- wizroafk [~userdi@c-76-23-254-105.hsd1.ct.comcast.net] has quit [Ping timeout: 256 seconds] | 18:30 | |
abetusk | http://www.whitehouse.gov/sites/default/files/microsites/ostp/ostp_public_access_memo_2013.pdf | 18:30 |
* strophariad awaits the hplustomtom from his higher-tech acqauintances here | 18:34 | |
@kanzure | awaits what ? | 18:35 |
@kanzure | abetusk: check the logs from earlier | 18:35 |
strophariad | http://en.wikipedia.org/wiki/TomTom | 18:36 |
@kanzure | SphericalMouse: i have resisted paying for somethingawful for >10 years. i am not about to break now. | 18:36 |
SphericalMouse | but think of the shitposting | 18:36 |
SphericalMouse | what advantages do is there with an ethernet over a USB data tranfer system? | 18:37 |
SphericalMouse | i fear the wine has a hold on my brain again | 18:37 |
@kanzure | well, most usb ports are probably usb2 at this point, and usb3 isn't widely deployed yet.. | 18:38 |
SphericalMouse | so? | 18:38 |
-!- klafka [~klafka@c-24-6-18-31.hsd1.ca.comcast.net] has joined ##hplusroadmap | 18:38 | |
@kanzure | well you're not going to get 20 MB/sec on usb2, but you could conceivably get that on ethernet. | 18:38 |
SphericalMouse | we were thinking of using a usb system for the fluid robot to control the various sub units. partly because arduino has that standard. its not like you need extreme bandwidth there | 18:39 |
@kanzure | oh wait. usb2 was supposed to work up to 60 MB/sec. | 18:39 |
@kanzure | SphericalMouse: what about linuxcnc? | 18:39 |
SphericalMouse | the cubespawn guy claims to aim for an ehternet data bus | 18:39 |
@kanzure | i think ethernet is fine, but it shouldn't really matter.. just put a computer in front of the machine, and just interface with the machine, right? | 18:40 |
SphericalMouse | I know nothing about that, but it looks like software, not hardware | 18:40 |
SphericalMouse | basically | 18:41 |
@kanzure | what is the specific problem you are solving ? | 18:41 |
SphericalMouse | we want to run various pieces of equipment from a central machine. mostly those will be simple commands or low amounts of data | 18:42 |
SphericalMouse | cubespawn seems to want to do continuous data and image analysis | 18:43 |
SphericalMouse | mostly, we don't care about images, except for microscopy purposes | 18:43 |
SphericalMouse | probably never real time image based control systems | 18:43 |
@kanzure | if your only options are ethernet and usb, then i would recommend ethernet. | 18:44 |
SphericalMouse | well, most arduinos don't have ethernet built in | 18:45 |
@kanzure | then don't use arduino | 18:45 |
SphericalMouse | so, should we find other solutions that do? find compatables that do? | 18:45 |
@kanzure | use a computer. like a pc. | 18:46 |
@kanzure | fenn: can you make me sound less crazy here? | 18:46 |
SphericalMouse | we will have a pc as the main controller, fo rsure | 18:46 |
SphericalMouse | but if we have some thermal cycler, for example, we probably want that to locally have some sort of onboard controller | 18:47 |
Urchin[emacs] | raspberry pi? | 18:50 |
SphericalMouse | a pi is complete overkill for something that is supposed to drive something up and down in temperature | 18:51 |
-!- wizrobe [~userdi@c-76-23-254-105.hsd1.ct.comcast.net] has joined ##hplusroadmap | 18:58 | |
strophariad | anyone here experiment with tDCS? | 18:58 |
@kanzure | a few people, yes. i forget who exactly. | 18:59 |
@kanzure | i suggest ultrasound instead of tdcs if you're considering your options. | 18:59 |
strophariad | would you recommend some resources on that? | 19:00 |
@kanzure | strophariad: http://diyhpl.us/~bryan/papers2/neuro/ultrasound/ | 19:02 |
@kanzure | and tdcs things: http://diyhpl.us/~bryan/papers2/neuro/tdcs/ | 19:02 |
strophariad | thanks much | 19:02 |
@fenn | i dont see any point in using USB for robot control if you have GPIO at your disposal | 19:05 |
@fenn | ethernet makes a better realtime network bus because it's simpler | 19:05 |
@fenn | also it's designed for multipoint and has better electrical characteristics | 19:06 |
@fenn | usb has a nominal max of what, three meters? | 19:06 |
@kanzure | what about just using full motherboards instead of just arduinos? am i making things up here? | 19:07 |
@fenn | anyway if all you want to do is turn a relay on and off, use the "digital IO" pins on the arduino | 19:07 |
@kanzure | the problem with arduino is that there's no package infrastructure | 19:08 |
@kanzure | and people end up rewriting a lot of firmware | 19:08 |
@fenn | hm. okay raspberry pi then | 19:08 |
@kanzure | i didn't even look at what ports are available on raspberry pi. | 19:09 |
@fenn | arduino does have various hardware advantages, like ADCs, timers, and hardware support for various serial protocols | 19:09 |
@fenn | and TTL | 19:09 |
@fenn | pi is 3.3V so you need level shifting for a lot of things | 19:10 |
nmz787 | if you want long range and low bandwidth you should consider RS485 | 19:10 |
nmz787 | level shifting is easy though, why do so many people complain about it? | 19:10 |
nmz787 | not saying you're complaining fenn | 19:11 |
@fenn | because it's one more chip to deal with | 19:11 |
nmz787 | but 1 more | 19:11 |
@kanzure | if there was some package manager for arduino things, i would feel better about it | 19:11 |
@fenn | i'm not sure it makes sense to have a package manager for a device which only does one thing at a time | 19:13 |
nmz787 | the arduino website seems pretty searchable to me | 19:13 |
nmz787 | i guess they might not have a central code repo | 19:13 |
nmz787 | parallax does though for the propeller | 19:13 |
@kanzure | it's not about searchable | 19:13 |
@fenn | a code repo isnt quite the same thing | 19:13 |
@kanzure | searchability isn't the point at all | 19:13 |
nmz787 | but there generally aren't dependencies for arduino stuff that isn't in the same file | 19:14 |
@kanzure | libraries have dependencies all the time | 19:14 |
nmz787 | yeah but they're a few files, not 10s or 100s | 19:15 |
nmz787 | for the arduino* | 19:15 |
@fenn | this boils down to static vs dynamic libraries | 19:15 |
@fenn | the real magic of a package manager is that things can be upgraded and still work | 19:16 |
@kanzure | so what.. you guys just copy and paste code from some library into your project, and pray that it doesn't break or just pretend upstream never updates? | 19:16 |
@fenn | i think most arduino code never gets upgraded | 19:16 |
@kanzure | or distributed :P | 19:17 |
nmz787 | generally upstream isn't changing within a processor type | 19:19 |
nmz787 | so if a new chip comes out in the processor family, you generally just switch the pinout and muxing... which all easily fits in one file | 19:19 |
nmz787 | otherwise most operations are standard C | 19:20 |
@kanzure | i mean, there should still be libawful-dev packages to put files in the right place | 19:21 |
@kanzure | or how npm works (local repo management, unless you force global system-wide with -G) | 19:21 |
@kanzure | upstream should change though- there's no reason to assume that the first version is the best thing ever | 19:21 |
@kanzure | fenn: why is there all this dumb culture around microcontroller programming anyway? | 19:22 |
@kanzure | fenn: i mean, there's a linux/embedded community that seems to be sane. but why is all this other crap broken? | 19:22 |
nmz787 | there are lots of arduino git projects | 19:22 |
@kanzure | so you just include everything as a submodule ? | 19:23 |
@kanzure | *as separate git submodules ? | 19:24 |
@kanzure | so far npm has been rather nice to work with. bundler is okay but it gets very slow. | 19:28 |
@fenn | kanzure: arduino is popular because it's an offshoot of processing, which was invented for dumb people | 19:28 |
@fenn | or perhaps i should say stupid people | 19:29 |
@fenn | anyway, there's plenty of C and C++ out there that's not "arduino" but runs fine on the arduino hardware | 19:29 |
@kanzure | oh, good point | 19:30 |
@fenn | but it's not even in the slackware stage yet, there's no need for an operating system and it would just waste space and cpu in most applications | 19:31 |
@kanzure | what would waste space? | 19:32 |
@fenn | the ram and flash required to run and store the OS | 19:32 |
@kanzure | for C? | 19:32 |
@fenn | we're talking about tens of bytes of ram here | 19:32 |
@kanzure | but why? | 19:33 |
@fenn | i dont know | 19:33 |
@kanzure | oh no your $200 liquid handler needs a $5 computer... big deal ? | 19:33 |
@fenn | it would seem that ram should be nearly unlimited, and yet it isn't. i don't know the rationale behind this | 19:33 |
@fenn | maybe ThomasEgi knows | 19:35 |
@kanzure | what's the argument against wiring up a thermocycler to something that runs linux? | 19:35 |
@fenn | lack of ADC, less ruggedness and need for level shifting. i think that's about it | 19:36 |
ThomasEgi | what what? | 19:36 |
@fenn | ThomasEgi: why do avr's have such pathetically small amounts of ram? | 19:36 |
ThomasEgi | because they are _micro_ controllers | 19:37 |
ThomasEgi | they controll stuff | 19:37 |
@fenn | but ram is so easy to make, and you can only chop chips up so small before you get diminishing returns | 19:38 |
ThomasEgi | they don't handle databases or flashy gui's | 19:38 |
ThomasEgi | yeah and.. how exactly are you going to address all that ram with an 8bit bus? | 19:38 |
@fenn | are microcontrollers cheap because they use outdated manufacturing processes? | 19:38 |
@kanzure | bank switching :P | 19:39 |
ThomasEgi | .... ahm. are you a bit.. dense or so? | 19:39 |
ThomasEgi | they are small. efficient, simple, modular circuits for controlling stuff. | 19:39 |
@kanzure | fenn: i suppose another reason to not just use a pc is if you need microsecond timing resolution. besides just adc and level shifting. | 19:39 |
ThomasEgi | indeed | 19:40 |
ThomasEgi | not only micro, but even nanoseconds | 19:40 |
@kanzure | but i don't think you need microsecond timing for most lab automation. | 19:40 |
@fenn | apparently the atmega328 has 1824 bytes of ram, which implies something like bank switching is going on | 19:40 |
ThomasEgi | yeah have fun witching those by slamming in 4GB of ram. | 19:41 |
@fenn | anyway it would be faster to copy all the flash memory into ram and talk to it over serial than the current practice of running code straight from flash | 19:41 |
ThomasEgi | no really. they are that small because there is simply no need for them to be any bigger | 19:41 |
ThomasEgi | what was the inital problem about anyway? | 19:42 |
@fenn | there's not any cheap rugged hardware interface that runs linux | 19:43 |
@fenn | and all the associated gnu tools that we've been spoiled with | 19:43 |
ThomasEgi | why do you need to run linux on an interface? | 19:43 |
@fenn | to simplify hardware dependencies | 19:43 |
ThomasEgi | that makes no sense? | 19:43 |
@fenn | um, so i dont need two "things" | 19:43 |
ThomasEgi | a hardware interface without hardware dependencies? | 19:43 |
ThomasEgi | i mean.. this _is_ hardware. you are depending on it. there is no way not to. | 19:44 |
@fenn | like, i could run code on a raspberry pi, but it doesn't have good i/o the way an avr does | 19:44 |
ThomasEgi | then use the avr right away | 19:44 |
ThomasEgi | where's the problem? | 19:45 |
@fenn | but the avr has a crappy (nonexistent) OS | 19:45 |
ThomasEgi | it is a microcontroller | 19:45 |
ThomasEgi | there is no os at all. and why should it | 19:45 |
@fenn | it can't download code from the internet and self-update | 19:45 |
ThomasEgi | actually you can | 19:45 |
ThomasEgi | you'd have to add an apropriate interface and a bootloader for it . but it is overkill | 19:46 |
ThomasEgi | microcontrollers are made to do one job. | 19:46 |
@fenn | you have to add a bunch of stuff to it and at that point you might as well just use a raspberry pi, right | 19:46 |
ThomasEgi | how often wolud you update that firmware? | 19:46 |
@fenn | (actually you can run ethernet straight off the avr's gpio but that's hacky) | 19:47 |
SphericalMouse | seriously, a machine made to run a pcr program doesn't _need_ to run linux | 19:47 |
ThomasEgi | very true | 19:47 |
@fenn | i think this is overly limiting in scope | 19:47 |
-!- klafka [~klafka@c-24-6-18-31.hsd1.ca.comcast.net] has quit [Remote host closed the connection] | 19:48 | |
@fenn | you see "a pcr machine" whereas that same assemblage of stuff could be used for many other processes | 19:48 |
@fenn | maybe it needs to be synchronized with other assemblages of stuff | 19:48 |
SphericalMouse | that it does | 19:48 |
ThomasEgi | i see no problem | 19:49 |
SphericalMouse | but you can send commands via the serial port | 19:49 |
ThomasEgi | but you definetly woludnt syncronize microcontrollers with ethernet :D | 19:49 |
ThomasEgi | maybe i2c, spi , or serial | 19:49 |
ThomasEgi | most likely spi | 19:49 |
@fenn | ThomasEgi: but then i'd have to physically wire them together in some hacky way | 19:49 |
@fenn | vs just plugging in an rj45 cable | 19:49 |
ThomasEgi | ... by using wires. right | 19:49 |
ThomasEgi | and a rj45 is .. | 19:50 |
ThomasEgi | oh yeah. guess what | 19:50 |
@fenn | you missed the point | 19:50 |
ThomasEgi | a wire! | 19:50 |
ThomasEgi | just add a DB9 connector for your connection, and you'r set | 19:50 |
@fenn | ok fine, let it use wifi | 19:50 |
@kanzure | package management is useful for many things, not just firmware updates and re-using a pile of parts for other things | 19:50 |
@fenn | since we're being pedantic | 19:50 |
ThomasEgi | fenn, wifi is even worse | 19:50 |
@kanzure | it's also useful for things like easier development and projects with 100s of developers | 19:50 |
ThomasEgi | tons of hardware, transport layers, encryption.. | 19:50 |
ThomasEgi | if you want to get bits from a to b. you use a wire. | 19:51 |
ThomasEgi | as simple as that. | 19:51 |
@fenn | i refuse to believe i can't have my cake and eat it too | 19:51 |
ThomasEgi | no hardware developer would want to pack his bits up in 7 layers of networking and protocolls, and THEN send it over a wire. and THEN unpack it again | 19:51 |
ThomasEgi | when he can simply transmit it straigth from a to b | 19:51 |
ThomasEgi | it's like ways ways ways too much overhead | 19:52 |
@fenn | yes i know, but you're being dumb and can't see that an ethernet cable is different from soldering two wires together | 19:52 |
ThomasEgi | noone tells you to solder wires | 19:52 |
ThomasEgi | you can have whatever connector you want | 19:52 |
ThomasEgi | what you dont want however.. is the entire network layering crap arount ethernet | 19:52 |
@fenn | it's not that bad | 19:53 |
ThomasEgi | trust me.. IT IS! | 19:53 |
ThomasEgi | if you don'- have dedicated hardware to handle it.. | 19:53 |
ThomasEgi | it really is nuts | 19:53 |
ThomasEgi | and if you are low on processing power and ram. it is even more so | 19:53 |
@fenn | and thus we're back to the problem of why we can't have it both ways | 19:53 |
@kanzure | processing power and ram is really cheap | 19:53 |
@kanzure | this isn't 1981 | 19:54 |
ThomasEgi | kanzure, but flat out not available on microcontrollers | 19:54 |
@kanzure | then don't use a microcontroller? | 19:54 |
-!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap | 19:54 | |
ThomasEgi | then you are back on a SoC | 19:54 |
ThomasEgi | like the raspberry | 19:54 |
nsh | guys guys guys, this is a completely unreasonable discussion | 19:55 |
nsh | the solution is obviously XML | 19:55 |
@kanzure | ... | 19:55 |
* ThomasEgi files a kickbanignore request | 19:55 | |
-!- sheena1 [~home@174.4.134.159] has joined ##hplusroadmap | 19:55 | |
* fenn mutters something about carrier pigeons | 19:55 | |
@kanzure | ThomasEgi: motion denied, but subject will be sternly warned? | 19:56 |
ThomasEgi | i don't really care as i'll go to bed. | 19:56 |
ThomasEgi | conclusion: either go with the cost and high complexity of a SoC. | 19:56 |
@fenn | (cost is negligible these days) | 19:57 |
ThomasEgi | or get a microcontroller and a hand full of lines of c code and simply run it | 19:57 |
@kanzure | "hand full of lines of c".. some of these libraries are huuge. | 19:57 |
ThomasEgi | spoiler: the last option is a lot less prone to errors, eats tons less energy and is cheaper on top of it, and gives you predictable timing | 19:57 |
@kanzure | yeah, i'm about ready to go for a completely opposite argument | 19:58 |
@kanzure | linuxes.. everywhere. | 19:58 |
ThomasEgi | go for what you want. i build enough circuits. i told you what i learned. | 19:59 |
@fenn | i'm afraid the only reasonable solution is the silliest.. use both | 19:59 |
@kanzure | i don't see how we jumped from package management for arduino things to operating systems | 19:59 |
@fenn | too bad the pi doesn't have a microcontroller built in | 19:59 |
@kanzure | but i still think there's room for arduino packages | 19:59 |
@fenn | please elaborate how such a package manager would work | 20:00 |
@kanzure | have you used npm? | 20:00 |
@fenn | no | 20:00 |
ThomasEgi | if you are going to build a pcr. just greb an arduino. if that's too expensive. get the bare atmega only, and programm it via ISP | 20:00 |
@kanzure | i'm sure npm is bloated under the hood, but it works off of a json file describing the contents of the package (the folder it's in) | 20:01 |
@kanzure | by default, dependencies are installed in yourproject/npm_modules/ | 20:01 |
@fenn | okay, that's assuming you have a filesystem | 20:01 |
@kanzure | and then that's added to the PATH for require() statements in js to find those things | 20:01 |
@kanzure | i think it's reasonable to assume you have a file system when compiling arduino code things | 20:02 |
@fenn | so this package manager is for the host computer, not the microcontroller itself | 20:02 |
@kanzure | yes, for compile-time | 20:02 |
ThomasEgi | hint hint hint. you can get the compiled arduino code. and load it directly onto the microcontroller. works, even with no bootloader | 20:02 |
@kanzure | but why would you do that! | 20:02 |
@fenn | then why not just use deb packages or npm or whatever | 20:02 |
@kanzure | ThomasEgi: that's completely fucking crazy | 20:02 |
@fenn | no it's not, that's how debian works | 20:02 |
ThomasEgi | about as crazy as connecting a programming adapter and executing a signle line of code | 20:03 |
@kanzure | fenn: no, debian provides the binaries in packages.. wtf dude | 20:03 |
@kanzure | ThomasEgi: nope | 20:03 |
@fenn | what do you think "binaries" are? | 20:03 |
ThomasEgi | kanzure, i did that. it works. what's so crazy bout it? | 20:03 |
@kanzure | i thought the .deb format had a provision for pre-built binaries to be stored inside | 20:03 |
@kanzure | fenn: maybe i'm wrong? | 20:04 |
@fenn | that's correct | 20:04 |
@kanzure | you don't just drop in compiled code and leave it to rot.. that's.. no. don't do that. :( | 20:04 |
@fenn | there are also packages for various architectures, even ARM and weird stuff | 20:04 |
@kanzure | yeah, i use debian/arm packages on my android phone all the itme. | 20:04 |
@kanzure | *time | 20:04 |
@fenn | but the debian way is to have a source package and some sort of procedure for compiling the source into the binary | 20:04 |
@fenn | i think this compilation step is even supposed to be architecture independent | 20:05 |
@fenn | who knows how it works in practice | 20:05 |
@kanzure | my point is that if you are adding a binary in, it should be package managed. | 20:06 |
@fenn | anyway, make a new debian architecture with a limited set of packages for it | 20:06 |
ThomasEgi | are you still talking about microcontrollers and package management? | 20:06 |
@fenn | yeah | 20:06 |
@kanzure | ThomasEgi: i think you are seriously misunderstanding me | 20:06 |
ThomasEgi | you guys do realize that microcontrollers only run one signle piece of code | 20:07 |
ThomasEgi | like a firmware | 20:07 |
@kanzure | sigh. firmware images can be compiled dude. | 20:07 |
@kanzure | they don't just magically exist.. | 20:07 |
-!- klafka [~klafka@c-24-6-18-31.hsd1.ca.comcast.net] has joined ##hplusroadmap | 20:07 | |
ThomasEgi | yeah but there is no point in having a package management system. if the target system doesn't even support that | 20:07 |
@kanzure | the target system supports a ROM image, i never stated otherwise. | 20:08 |
@fenn | kanzure: in this case, how is a package manager better than just doing a git pull? | 20:08 |
ThomasEgi | anyway. introducing a package management for a microcontrollers seems just wrong in so many ways. | 20:09 |
ThomasEgi | for sharing it and the sourcecode, simply tarball it up with compile instructions. | 20:09 |
@kanzure | ThomasEgi: so if you have 100 different third party libraries in your arduino project, you just manually update the different binaries? | 20:09 |
@fenn | that assumes it's a monolithic package and has no dependencies | 20:09 |
@fenn | and maybe your dependencies have dependencies | 20:09 |
ThomasEgi | kanzure, an arduino code ends up as one binary. not different individual libraries. | 20:10 |
ThomasEgi | they are only existant on the development machine | 20:10 |
@fenn | the code still has to play nice | 20:10 |
@fenn | can't have variables go missing | 20:10 |
ThomasEgi | once you start compiling your code. it'll collect all those libraries and stuff it into a single rom image which you later upload onto your microcontroller | 20:11 |
@fenn | someone has to set all that up | 20:11 |
@kanzure | fenn: git submodules probably wouldn't work because if there's a common library shared between multiple 2 or more libraries, you would just be submodule cloning all over the place? | 20:11 |
@fenn | is there such a thing as a git submodule dependency loop? | 20:12 |
@fenn | i haven't really used that feature much | 20:12 |
-!- klafka [~klafka@c-24-6-18-31.hsd1.ca.comcast.net] has quit [Ping timeout: 255 seconds] | 20:13 | |
@kanzure | i wonder if github could be broken with a loop like that | 20:13 |
@fenn | since everything it referenced by hash it might just work out fine | 20:13 |
@fenn | have your ouroboros and eat it too | 20:14 |
ThomasEgi | i give up. it's not like any of that would exist on a microcontroller anyway. it's just c code. modules go in as includes like everything else. that's it. aaaand i'm off to bed. | 20:14 |
@fenn | thanks for playing ThomasEgi | 20:14 |
-!- ThomasEgi [~thomas@panda3d/ThomasEgi] has quit [Quit: ...unyaaa ~~~] | 20:14 | |
@kanzure | i know he's wrong. libwhatever-dev proves him wrong. | 20:14 |
@fenn | i dont see any arduino dev packages | 20:15 |
@kanzure | there are none | 20:15 |
@fenn | except for arduino itself that is | 20:15 |
-!- sheena1 [~home@174.4.134.159] has quit [Ping timeout: 276 seconds] | 20:16 | |
@fenn | seems like a poor way to run an open source project | 20:16 |
@kanzure | microcontroller programmer culture is weird | 20:16 |
@kanzure | "let's throw out everything we know about releasing software" | 20:16 |
@kanzure | s/releasing/developing | 20:16 |
@kanzure | i suppose i should go package some arduino things into a .deb or something | 20:17 |
@fenn | i think in large part it's because you want some sense of control over exactly what the device is doing | 20:17 |
@fenn | if you're calling some library it could mess up your timing, fiddle with IRQs, reset timers, who knows what | 20:18 |
@kanzure | me? or the owner of a chip in general? | 20:18 |
@kanzure | ah | 20:18 |
@fenn | you = a microcontroller programmer | 20:18 |
@kanzure | in the nodejs community in particular, it's considered very bad to release a package that blocks on io anywhere, unless you EXPLICITLY MAKE IT VERY WELL KNOWN | 20:19 |
@fenn | so the arduino code base is (was at least) not very efficiently written, and the libavr stuff was much better | 20:19 |
@kanzure | i never used libavr things. what was there? | 20:19 |
@fenn | or was it avrlib, i forget which was which | 20:19 |
@kanzure | compiler toolchain? | 20:20 |
@fenn | http://www.procyonengineering.com/embedded/avr/avrlib/ | 20:20 |
@kanzure | (and why not just gcc/llvm?) | 20:20 |
@kanzure | ah just a library of useful things. | 20:20 |
@fenn | sorry, libavr is the toolchain, avrlib is the standard library | 20:20 |
@fenn | arduino also uses libavr to compile | 20:21 |
@kanzure | ThomasEgi sounds like a gentoo user, except this port of gentoo doesn't exist yet or somthing :P | 20:21 |
-!- sheena1 [~home@174.4.134.159] has joined ##hplusroadmap | 20:21 | |
@fenn | gentoo would work, but things arent even at that point | 20:22 |
@fenn | it's tarballs and floppy disks | 20:22 |
@fenn | or zip files and .exe installers in this case | 20:23 |
@kanzure | so much for taking advantage of a third-party ecosystem efficiently | 20:23 |
@kanzure | client-side js is also at a similar stage. bower was supposed to be a thing to do better package management for client-side js but i haven't seen it going anywhere lately (i also haven't checked, so whatever). | 20:25 |
* kanzure drives home | 20:25 | |
@fenn | it seems like a lot of the work is already done, and there's nothing else to do, so why bother setting up a fancy development community when no development is going on. maybe my perception is wrong | 20:25 |
@fenn | at least raspberry pi uses debian | 20:26 |
@fenn | "we gave you VT100 and ethernet, what more do you want, now get off my lawn" | 20:27 |
-!- SphericalMouse [62cfd610@gateway/web/freenode/ip.98.207.214.16] has quit [Ping timeout: 245 seconds] | 20:49 | |
nmz787 | I agree with this 20:17 <@fenn> i think in large part it's because you want some sense of control | 20:51 |
nmz787 | over exactly what the device is doing | 20:51 |
nmz787 | 20:18 <@fenn> if you're calling some library it could mess up your timing, | 20:51 |
nmz787 | fiddle with IRQs, reset timers, who knows what | 20:51 |
-!- lichen [~lichen@c-24-21-206-64.hsd1.or.comcast.net] has quit [Quit: Lost terminal] | 20:56 | |
@kanzure | nmz787: that's not much of a problem though.. i mean how many libraries are going to reset timers really? | 21:23 |
@kanzure | which ones currently do that? | 21:24 |
-!- aristarchus [~aristarch@unaffiliated/aristarchus] has quit [Quit: Leaving] | 21:33 | |
-!- sheena2 [~home@174.4.134.159] has joined ##hplusroadmap | 21:50 | |
-!- sheena1 [~home@174.4.134.159] has quit [Read error: Connection reset by peer] | 21:50 | |
-!- pwang__ [~pwang@unaffiliated/parahsailin] has joined ##hplusroadmap | 21:56 | |
-!- ParahSail1n [~pwang@unaffiliated/parahsailin] has joined ##hplusroadmap | 21:56 | |
-!- pwang__ [~pwang@unaffiliated/parahsailin] has quit [Quit: Leaving] | 22:05 | |
-!- devrandom [~devrandom@gateway/tor-sasl/niftyzero1] has quit [Ping timeout: 276 seconds] | 22:12 | |
nmz787 | kanzure: lots of the time related functions | 22:17 |
nmz787 | delays | 22:17 |
nmz787 | serial comms | 22:17 |
nmz787 | tone production | 22:17 |
nmz787 | some PWM forms | 22:17 |
@kanzure | can't you just store the current clock value somewhere? | 22:17 |
nmz787 | so you can't really mash all of those functions/libs in one project | 22:17 |
nmz787 | it's not about the value, you can, it's about timing things to happen at specific times | 22:18 |
nmz787 | and if those things take a long time, they can interfere with other things using the timer that are on a tight schedule | 22:18 |
@kanzure | presumably each function will have documentation that tells you how many cycles they take | 22:18 |
nmz787 | no | 22:19 |
nmz787 | well | 22:19 |
nmz787 | you would have to dig all that up | 22:19 |
nmz787 | you just look with an oscilloscope | 22:19 |
@kanzure | what? why wouldn't a library tell you? | 22:19 |
nmz787 | or write it in assembly | 22:19 |
nmz787 | or get a scheduler | 22:20 |
nmz787 | i don't see them | 22:20 |
-!- Urchin[emacs] [~user@unaffiliated/urchin] has quit [Ping timeout: 264 seconds] | 22:20 | |
nmz787 | remember that often these programs are less than 64 or even 2 kilobytes | 22:24 |
nmz787 | not mega or gigabytes | 22:24 |
nmz787 | so it might just be a different realm of problem | 22:24 |
nmz787 | there might be a better way to handle things | 22:25 |
nmz787 | but I don't really have any ideas other than a central code repo | 22:25 |
@kanzure | just because it's 2 kb is no excuse to have shitty documentation | 22:25 |
rigel | can someone tell me how the shit this virtualenv crap works | 22:37 |
rigel | i have a brand new, fresh vm of arch | 22:37 |
@kanzure | sure. wazzup? | 22:37 |
rigel | and i havent the faintest idea how to set up the stuff that i apparently NEED TO DO THESE DAYS to work with python packages | 22:37 |
rigel | i.e. pip/virtualenv/virtualenvwrapper | 22:37 |
@kanzure | i like to use venv-burrito to setup virtualenv and virtualenvwrapper | 22:37 |
rigel | oh youve got to be shitting me | 22:38 |
@kanzure | curl -s https://raw.github.com/brainsik/virtualenv-burrito/master/virtualenv-burrito.sh | bash | 22:38 |
@kanzure | but first you will need pip: | 22:39 |
@kanzure | curl -O https://github.com/pypa/pip/raw/master/contrib/get-pip.py && sudo python get-pip.py | 22:39 |
rigel | if im getting pip first why dont i just get virtualenv first | 22:39 |
@kanzure | arch might have a python-setuptools package that might provide easy_install or pip, who knows. | 22:39 |
rigel | doesnt that include pip? | 22:39 |
@kanzure | virtualenv has a setup.py which requires the setuptools package, i think | 22:40 |
rigel | does it matter that i have both py2.7 and 3 installed? | 22:40 |
@kanzure | no | 22:40 |
@kanzure | just make sure you use the same version when installing all this shit | 22:40 |
@kanzure | so i suggest you be explicit with python about it | 22:40 |
@kanzure | python2.7 dowhatever | 22:40 |
rigel | this shit used to be easy | 22:40 |
rigel | what the goddamn fuck happened | 22:41 |
@kanzure | arch? not really. | 22:41 |
rigel | no, i mean python | 22:41 |
@kanzure | easy_install pip virtualenv virtualenvwrapper | 22:41 |
rigel | as user, right, not as root | 22:42 |
rigel | i hate installing shit as root | 22:42 |
rigel | gives me the willies | 22:42 |
rigel | meh. it worked as root, i guess | 22:45 |
-!- superkuh [~superkuh@unaffiliated/superkuh] has quit [Quit: the neuronal action potential is an electrical manipulation of reversible abrupt phase changes in the lipid bilayer] | 22:55 | |
@kanzure | so it turns out my aunt owns a home next to mission/24th in san francisco | 23:02 |
-!- yashgaroth [~ffffff@cpe-66-27-118-94.san.res.rr.com] has quit [Quit: Leaving] | 23:07 | |
nmz787 | kanzure: crazy | 23:28 |
nmz787 | kanzure: someone you're close with or grew up around? | 23:28 |
@kanzure | no | 23:29 |
-!- joshcryer [~g@unaffiliated/joshcryer] has joined ##hplusroadmap | 23:49 | |
joshcryer | http://www.altdevblogaday.com/2013/02/22/latency-mitigation-strategies/ | 23:49 |
--- Log closed Sat Feb 23 00:00:08 2013 |
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!