--- Log opened Wed Sep 06 00:00:10 2017
01:21 -!- Urchin [~urchin@unaffiliated/urchin] has joined ##hplusroadmap
01:34 -!- Cory [~Cory@unaffiliated/cory] has joined ##hplusroadmap
01:39 -!- sachy [~sachy@nat.brmlab.cz] has joined ##hplusroadmap
01:53 -!- Gurkenglas [~Gurkengla@dslb-178-000-222-214.178.000.pools.vodafone-ip.de] has joined ##hplusroadmap
01:53 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has joined ##hplusroadmap
02:54 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has quit [Ping timeout: 240 seconds]
03:03 -!- strages [uid11297@gateway/web/irccloud.com/x-igxtjnlrlgjtkvnc] has quit [Quit: Connection closed for inactivity]
03:12 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has joined ##hplusroadmap
03:45 -!- darsie [~username@84-113-55-42.cable.dynamic.surfer.at] has joined ##hplusroadmap
04:58 -!- g0d355__ [~lmao@104.131.75.159] has joined ##hplusroadmap
05:04 -!- cluckj [~cluckj@static-173-59-27-112.phlapa.ftas.verizon.net] has quit [Quit: Leaving]
05:13 -!- Gurkenglas [~Gurkengla@dslb-178-000-222-214.178.000.pools.vodafone-ip.de] has quit [Ping timeout: 264 seconds]
05:55 -!- jtimon [~quassel@199.31.134.37.dynamic.jazztel.es] has joined ##hplusroadmap
06:06 -!- cluckj [~cluckj@static-173-59-27-112.phlapa.ftas.verizon.net] has joined ##hplusroadmap
06:18 -!- aeiousomething [~aeiousome@183.82.170.54] has quit [Quit: leaving]
06:19 -!- aeiousomething [~aeiousome@gateway/vpn/privateinternetaccess/aeiousomething] has joined ##hplusroadmap
06:19 -!- jtimon [~quassel@199.31.134.37.dynamic.jazztel.es] has quit [Ping timeout: 252 seconds]
06:30 -!- strages [uid11297@gateway/web/irccloud.com/x-poykyuxcvwcyxqpo] has joined ##hplusroadmap
06:38 -!- jaboja [~jaboja@jaboja.pl] has joined ##hplusroadmap
06:45 < kanzure> "Hardware architectures for deep neural learning" http://www.rle.mit.edu/eems/wp-content/uploads/2017/06/ISCA-2017-Hardware-Architectures-for-DNN-Tutorial.pdf
07:00 -!- darsie [~username@84-113-55-42.cable.dynamic.surfer.at] has quit [Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/]
07:02 -!- darsie [~username@84-113-55-42.cable.dynamic.surfer.at] has joined ##hplusroadmap
07:09 < JayDugger> .tell Asterion O'Neill's 2081 now available free for Kindle. (cf. PM lecture on non-fiction)
07:09 < yoleaux> JayDugger: I'll pass your message to Asterion.
07:10 < JayDugger> .botsnack
07:10 < yoleaux> :D
07:15 -!- ensign [~ensign@2001:41d0:8:d711::1] has quit [Ping timeout: 255 seconds]
07:29 -!- ensign [~ensign@2001:41d0:8:d711::1] has joined ##hplusroadmap
07:48 -!- aeiousomething [~aeiousome@gateway/vpn/privateinternetaccess/aeiousomething] has quit [Ping timeout: 248 seconds]
08:19 < kanzure> fltrz: http://www.terapaper.com/
08:20 < kanzure> i forget what this is but it's a tab i had opened in my browser
08:21 -!- jaboja [~jaboja@jaboja.pl] has quit [Ping timeout: 260 seconds]
08:29 < nmz787> kanzure: did you transcribe today's call? the link you sent wasn't for the most recent if by most recent you meant today's
08:34 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has quit [Remote host closed the connection]
08:36 -!- mrdata [~mrdata@unaffiliated/mrdata] has joined ##hplusroadmap
08:51 -!- preview [~quassel@2407:7000:842d:4000::3] has quit [Ping timeout: 255 seconds]
08:52 < kanzure> what
08:53 < kanzure> nmz787: fixed. thank you.
08:58 < brujo_biologica> FNFN
08:58 < brujo_biologica>   kanzure │ that should not require ethereum. are you sure that's all your doing?
08:58 < brujo_biologica>   kanzure │ that should not require ethereum. are you sure that's all your doing?
09:01 -!- jtimon [~quassel@199.31.134.37.dynamic.jazztel.es] has joined ##hplusroadmap
09:04 < kanzure> who?
09:07 -!- Malvolio [~Malvolio@unaffiliated/malvolio] has quit [Ping timeout: 240 seconds]
09:08 < fltrz> kanzure: I can't read arabic?
09:08 < fltrz> is this a sci-hub mirro?
09:12 < fltrz> kanzure: brujo_biologica is quoting you, it's what you said when I was sketching the crowdsourced through ethereum idea
09:12 < fltrz> not sure what FNFN means?\
09:13 < kanzure> alright well first step is to learn arabic
09:13 < fltrz> kanzure: you know arabic?
09:14 < fltrz> it's not even clear the site is free?
09:15 -!- gene-hacker [~tetrapod@CMU-943103.ANDREW.CMU.EDU] has joined ##hplusroadmap
09:15 < gene-hacker> hey kanzure, is there anywhere one can download all of scihub?
09:16 < fltrz> nice, this article https://www.researchgate.net/publication/263702575_Topological_Wiener_Indices_and_Polynomials_of_C84_Fullerene_Nanocage has a big phat "TeraPaper.com" watermark on it
09:16 < gene-hacker> or is that in the libgen repo torrent
09:17 < nmz787> gene-hacker: are you in PGH now?
09:17 < gene-hacker> maybe
09:17 < kanzure> gene-hacker: i'm working on it... it's a slow download.
09:17 < kanzure> it's 50 TB
09:17 < gene-hacker> HOLY SHIT
09:17 < nmz787> gene-hacker: that's where I grew up ;)
09:17 < kanzure> i tried bribing them but they have $400k and they don't care
09:17 < gene-hacker> nice
09:18 < kanzure> gene-hacker: http://diyhpl.us/~bryan/scimag/torrents/
09:18 < gene-hacker> you are downloading it?
09:18 < gene-hacker> the libgen torrent?
09:18 < kanzure> ~supposedly~ this is libgen's mirror of scihub's content but nobody knows for sure
09:19 < fltrz> brujo_biologica: lol at FNFN
09:19 < nmz787> gene-hacker: are you in school there now?
09:19 < kanzure> school is for plebs. gene hacker has ascended to postdoc status.
09:19 < gene-hacker> fool
09:19 < fltrz> kanzure: how slow is the download on average?
09:20 < kanzure> i was getting 10 kb/sec but i was talking with bram yesterday and i think i have a reasonable idea for hacking libtorrent to figure that out without actually allocating 1000 GB of disk space
09:20 < gene-hacker> I have gone post-post-doc
09:20 < kanzure> never go SSJ4
09:20 < nmz787> gene-hacker: nice, what lab?
09:21 < gene-hacker> I'd rather not say in a public channel
09:22 < kanzure> hype lab
09:22 < nmz787> uptight lab?
09:22 -!- Storyteller [~Storytell@unaffiliated/storyteller] has joined ##hplusroadmap
09:22 < gene-hacker> I wonder if it would be better to put it on IPFS
09:22 < kanzure> siacoin would be better, plus we have the siacoin person in here -)
09:22 < kanzure> :-)
09:22 < gene-hacker> yeah basically
09:23 < kanzure> the problem is getting the data
09:23 < kanzure> they have to want to distribute the data
09:23 < kanzure> i'm thinking about setting up a server in seychelles next to them
09:23 < kanzure> it's off the coast of fucking mali or something? i don't know
09:24 < fltrz> kanzure: 10kbps?? then scraping the userfacing sci-hub portal would be much faster even if only a single user, let alone many in parallel
09:24 < gene-hacker> wait, you're scraping it?
09:24 < kanzure> well maybe some new seeders have showed up
09:24 < kanzure> so we should check
09:24 < fltrz> kanzure: another advantage is that  since the index is public you can order them from small to big
09:24 < kanzure> it might have been 2 years ago when i last checked.
09:25 < kanzure> well the torrents are 100 GB each and each one has like 100 zips each 1 GB in size
09:25 < nmz787> kanzure: I think seychelles is a tourist spot for many Indians... but also one of the places on the top list of endangered-by-sea-level-rise
09:25 < kanzure> so you would have to pick the zips that have the larger number of files based on like... the range in the filename.
09:25 < kanzure> nmz787: scihub is hosted there.
09:25 < kanzure> according to whois
09:25 < gene-hacker> you know I wonder what it would take to put it on tape
09:25 < gene-hacker> and mail the damn thing
09:25 < kanzure> yes mailing would be a good idea
09:25 < nmz787> time for a vacation
09:26 -!- Malvolio [~Malvolio@unaffiliated/malvolio] has joined ##hplusroadmap
09:26 < kanzure> well she's not in seychelles
09:26 < nmz787> the tape drive is a she?
09:26 < kanzure> the hosting is located there
09:26 < gene-hacker> the god of libgen?
09:26 < fltrz> nmz787: yes, a twin even
09:27 < fltrz> they roll on top of each other
09:27 < fltrz> kanzure: what you mentioned yesterday about the papers being cached on cloudflare.. that is interesting
09:28 < kanzure> well only some.... e.g. the ones that users have requested, during the period for which the site has been hosted on cloudflare.
09:28 < kanzure> unfortunately youcan't do url enumeration because the url includes the hash of the file plus author surname plus year
09:28 < kanzure> if we have the hash from libgen database dump then maybe........
09:29 < fltrz> md5 or other?
09:29 < gene-hacker> I wonder what it would take to pull off a data-heist, talk with the person that runs libgen, get it on tape, get it out of the country
09:29 < fltrz> perhaps sci-hub wouldnt mind adding a column in their sql database?
09:30 < gene-hacker> man that fullerene paper is weird
09:30 < kanzure> fltrz: you need to go ask mindwrapper yourself because she is not replying to me lately
09:31 < fltrz> hmm, I abstain social media, never had titterz
09:31 < fltrz> *twiddles
09:31 < kanzure> gchat
09:31 < kanzure> vk.com too
09:31 < kanzure> https://vk.com/sci_hub
09:33 < kanzure> gene-hacker: for large-scale star shade deployment, instead of bubble expansion w/ atmospheric injection,  jeffrey epstein goon suggests using nuke to expand a stretchable material to 10 km width/diameter in orbit.
09:34 < gene-hacker> that seems like whatever material you're stretching would get destroyed by the nuke
09:35 < gene-hacker> also that probably violates a couple international treaties
09:35 < kanzure> fltrz: http://dabamirror.sci-hub.bz/e1186183c368da740fcaa8cdeb413137/solomatov2007.pdf  <-- url structure
09:36 < fltrz> kanzure: I had said I was going to make some lorenz curve plots (I still didnt make them), but I did crawl through the index with some statistics, and the one with the highest bytes/page turns out to be a book
09:36 < fltrz> books can also have DOI's...
09:36 < fltrz> it was a history/art book, so I can imagine theres quite a lot of scanned books, so I believe that purely articles would be much less than 50TB
09:37 < fltrz> thats another reason for scraping IMHO
09:37 < fltrz> with the sql index, you can more intelligently select which to download
09:38 < fltrz> I wonder if there is an objective way to map DOI -> boolean book vs article
09:38 < gene-hacker> there are probably a bunch of scanned articles
09:38 < kanzure> my arabic friends tell me "ah it's iran's arxiv" (re: terapaper.com)
09:38 < kanzure> gene-hacker: start downloading those torrents :-)
09:39 -!- Malvolio [~Malvolio@unaffiliated/malvolio] has quit [Ping timeout: 260 seconds]
09:39 < gene-hacker> well right now I need a more stable method to download huge torrents
09:40 -!- Malvolio [~Malvolio@unaffiliated/malvolio] has joined ##hplusroadmap
09:40 < gene-hacker> using my router with 250 MB of memory and a 1 GHz processor as a seedbox
09:40 < kanzure> seychelles datacenter-- they sell hosting for like $3/mo
09:41 < gene-hacker> what do you use to torrent big stuff?
09:44 < kanzure> yeaaah that's a reasonable question
09:44 < kanzure> i have no idea
09:44 < kanzure> i was using aws for a while with a cluster. but that's somewhat expensive, especially for 50 TB.
09:44 < fltrz> kanzure: regarding url structure, yes observing the database its not entirely clear how they arrive at filename (solomatov2007.pdf), is this consistently surname & year ?
09:44 < kanzure> well does the database have that hash? i thikn name+year is worth trying.
09:46 < brujo_biologica> http://www.biorxiv.org/content/early/2017/03/07/114553//misl.cs.washington.edu/
09:48 < nmz787> gene-hacker: so I can't tell from googlings if you're still in OR or now in PA... I never did make it down to see you here in OR
09:48 < nmz787> (still haven't been to Corvallis at all)
09:50 < nmz787> would be a bit sad if I missed you around these parts...
09:51 < fltrz> kanzure: I am searching the hash, it will take half an hour or so (its over USB2)
09:51 < kanzure> you need a bloom filter for fast lookup :-)
09:51 < fltrz> yes
09:51 < kanzure> (unfortunately ocnstructing the bloom filter will also take a while....)
09:52 < kanzure> or an ordered hash dictionary i guess. that owuld work.
09:53 < fltrz> I was intending to implement those lookups, but rotated projects back to kinect & time machine
09:53 < kanzure> forward time machine is a reasoable goal
09:53 < kanzure> gosh why so many typos.   i've typed like 30k words in the past day (transcripts).
09:53 < kanzure> (between the bitcoin developer stuff and hgp-write)
09:53 < fltrz> because I encountered the book, so I continue other projects while I let the "DOI->is it book/article?" linger in back of my mind
09:54 < gene-hacker> did you catch the eclipse?
09:54 < nmz787> yeah, just from my driveway
09:54 < kanzure> btw there's a public submit api thing to get DOI number
09:54 < nmz787> 94%
09:54 < fltrz> kanzure: its not to enter the time machine, its only information, and it'd be backwards (otherwise its just memory)
09:54 < kanzure> backwards time travel? only forward time travel works.
09:56 < fltrz> I guess the jury is still out (I could see how it could work, but I am very skeptical myself, I first need to reproduce a certain phenomenon that was reported, if I see the same [which I expect not to] I think it is possible to have backwards time travel)
09:56 < gene-hacker> yeah you aren't missing out on much in Corvallis
09:57 < brujo_biologica> forgive me for the babble. accidental right-clicks
09:57 -!- p0nziph0ne [~p0nziph0n@unaffiliated/p0nziph0ne] has joined ##hplusroadmap
09:57 < brujo_biologica> damn puTTY
09:57 < fltrz> " <kanzure> btw there's a public submit api thing to get DOI number" < not sure I can follow, f(x)=DOI(x), what is 'x'?
09:58 < kanzure> i think either the document or the url to the document
09:59 < kanzure> to get a new DOI number assigned
10:00 < fltrz> I thought it was a paying thing
10:01 < kanzure> i don't know their business model or who they are
10:01 < kanzure> it might be OCLC
10:01 < kanzure> dunno
10:03 < fltrz> kanzure: it looks like an MD5 hash (same size) but I didn't find it, was it recently retrieved into scihub? perhaps its reversed md5 or smth?
10:03 < kanzure> let's try something else
10:03 < kanzure> give me a file from your list and its md5 and its author and let's just check that way
10:04 < fltrz> ok
10:04 < kanzure> and if it's not on their server, then we go through their frontend UI and search for the doi for the object you picked, and then after it loads, we check their cloudflare cache thing
10:04 < fltrz> better, I give the DOI
10:05 < fltrz>  10.1002/(sici)(1997)5:1<1::aid-nt1>3.0.co;2-8 | 93b76bc6875ce7957eeec1247e7b83b9 | Yasuhiro Itagaki; Tsuyoshi Fujita; Hideo Naoki; Tadashi Yasuhara; Marta Andriantsiferana; Terumi Nakajima | 1997 |
10:05 < kanzure> one that you have not used with scihub
10:05 -!- Storyteller [~Storytell@unaffiliated/storyteller] has left ##hplusroadmap ["Leaving"]
10:05 < fltrz> thats DOI (weird DOI no?) then MD5, then author string, then year
10:05 < kanzure> ok i am just guessing on the url, i picked http://dabamirror.sci-hub.bz/93b76bc6875ce7957eeec1247e7b83b9/itagaki1997.pdf
10:05 < kanzure> and i get a captcha
10:05 < kanzure> and now a 404
10:05 < fltrz> just enter the doi I guess
10:06 < kanzure> what is the actual doi here? your paste is not good
10:06 < fltrz> the md5 is modified
10:06 < fltrz> the DOI is  " 10.1002/(sici)(1997)5:1<1::aid-nt1>3.0.co;2-8 "
10:06 < fltrz> but it seems awful
10:06 < fltrz> let me just take another paper same exercise
10:06 < kanzure> wow the DOI format is shit
10:06 < kanzure> no this is what the website says too http://onlinelibrary.wiley.com/doi/10.1002/(SICI)(1997)5:2%3C47::AID-NT1%3E3.0.CO;2-X/abstract
10:07 < kanzure> can you sort by DOI length and let's see the worst offenders
10:07 < fltrz>  10.1002/chin.200606115 | 83b5def982a99660ad64c7a9590f8079 | Dong-Mei Shen; Chao Liu; Qing-Yun Chen | 2006
10:09 < fltrz> I'm not going to sort by DOI for worst offenders now (the sql indices were disabled, building them took ages when importing so I just disabled the keys)
10:09 < kanzure> 404 on http://dabamirror.sci-hub.bz/83b5def982a99660ad64c7a9590f8079/shen2006.pdf
10:09 < kanzure> alright i'm getting endless captchas on scihub now, i'm blocked
10:10 -!- Gurkenglas [~Gurkengla@dslb-178-000-222-214.178.000.pools.vodafone-ip.de] has joined ##hplusroadmap
10:10 < fltrz> kanzure: where did you get the url structure from? I don't see it on scihub
10:10 < kanzure> http://dacemirror.sci-hub.io/journal-article/7a5eb50b00c44dec840b5eee5667d571/shen2006.pdf?download=true
10:10 < kanzure> it's an iframe
10:10 < kanzure> what is dacemirror
10:11 < fltrz> kanzure: on my first try Im getting infinite captcha as well, I guess they use some DDoS detection or whatever, and it noticed our 2 ip's asking for same file or smth
10:11 < kanzure> here is a real one: http://dacemirror.sci-hub.io/journal-article/e1a64158d10e2af6a9dab0f425eb4db0/mangels2001.pdf
10:12 < kanzure> yea it's probably really really aggressive blocking
10:12 < kanzure> or they broke something recently :-)
10:13 < fltrz> yes either one of those 2
10:13 < fltrz> or she's in this channel :)
10:23 -!- gene-hacker [~tetrapod@CMU-943103.ANDREW.CMU.EDU] has quit [Ping timeout: 246 seconds]
10:27 -!- atrus6 [~atrus6@72.241.82.247] has joined ##hplusroadmap
10:32 -!- jaboja [~jaboja@jaboja.pl] has joined ##hplusroadmap
10:43 < kanzure> "Its based on solid phase extraction (SPE) and thin layer chromatography (TLC), and our goal is make it accessible enough so that anyone can test using ChemPrint and we are running a larger user trials at University of Georgia, Athens, GA."   https://experiment.com/projects/fojmgoyinptvpdgiylye
10:48 < kanzure> .title https://news.ycombinator.com/item?id=15184809
10:48 < yoleaux> 23andMe is raising about $200M, led by Sequoia | Hacker News
10:52 -!- gene-hacker [~tetrapod@ok-u2.wv.cc.cmu.edu] has joined ##hplusroadmap
10:54 -!- mrdata_ [~mrdata@unaffiliated/mrdata] has joined ##hplusroadmap
10:57 -!- mrdata [~mrdata@unaffiliated/mrdata] has quit [Ping timeout: 248 seconds]
11:02 < fltrz> kanzure: I think somethings wrong with sci-hub in general?
11:02 < nmz787> hmm, so in locomotion, it seems the most efficient you could get would be where all the heat produced is at the moving-volume's perimeter, i.e. the friction from air rubbing, and tires/wheels rubbing
11:03 < kanzure> fltrz: yea maybe they updated their captcha stuff within the past ~24 hours. it's definitely strange behavior.   or you and i are blocked :-) congratulations.
11:03 < nmz787> or even with protein locomotion, the protein's friction with water/solubles
11:04 < fltrz> nmz787: thats why proposals like hyperloop propose partially evacuating the transport tunnel, and in using magnetism can be very low friction
11:04 < nmz787> sure
11:05  * nmz787 was just reading a random page about steam trains smoke being black under load
11:05 < fltrz> nmz787: an alternative would be wormholes/portals...
11:05 < gene-hacker> well if you really want to get into transport efficiency check out the von Karmen-Gabrielli diagram
11:06 < gene-hacker> https://en.wikipedia.org/wiki/Von_K%C3%A1rm%C3%A1n%E2%80%93Gabrielli_diagram
11:07 < gene-hacker> it's a plot of tractive force vs speed
11:08 < gene-hacker> https://trainsnboatsnplanes.wordpress.com/2010/01/07/the-price-of-speed/
11:09 < fltrz> wow horses are more efficient than jet planes according to the diagram
11:14 < fltrz> they're also more efficient than cars, but not sure how old this diagram is
11:14 < gene-hacker> which diagram?
11:15 < fltrz> also not sure if the mass in the efficiency calculation is the freight mass or the total vehicle mass.
11:15 < gene-hacker> http://www.ingenia.org.uk/Content/ingenia/issues/issue22/Imperial.pdf
11:15 < fltrz> the von Karmen-Gabrielli diagram you linked
11:15 < gene-hacker> it's total vehicle mass
11:15 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has joined ##hplusroadmap
11:16 < fltrz> so the economical efficiency is still something different
11:16 < gene-hacker> no, it depends on the regime
11:16 < gene-hacker> the commercial airliner is more efficient than a horse at certain speeds
11:17 < gene-hacker> last link posted is the most recent one
11:17 < gene-hacker> closer to the line is more efficient
11:20 < fltrz> I don't see horses in the last link?
11:20 < gene-hacker> the last link has an update diagram
11:20 < gene-hacker> it doesn't have horses though
11:22 < fltrz> at some point ICBM may become most efficient?
11:22 < fltrz> priority mail by rockets
11:24 < fltrz> well, most efficient is still at quasistatic speeds of course
11:24 < gene-hacker> suborbital airliners have been proposed for new york to tokyo in 30 minutes
11:25 < gene-hacker> the biggest problem is that getting up to speed takes a rocket
11:26 < kanzure> well we can reuse rocket boosters these days so that's not impossible
11:42 -!- gene-hacker [~tetrapod@ok-u2.wv.cc.cmu.edu] has quit [Ping timeout: 248 seconds]
11:55 < fltrz> kanzure: sci-hub still works through i.e. https://de.hideproxy.me/
11:55 < fltrz> but! I dont see the original url structure...
11:56 < fltrz> http://cyber.sci-hub.cc/MTAuMTAwMi9hamhiLjEwMTg3/leonard2003.pdf
11:57 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has quit [Remote host closed the connection]
11:59 < fltrz> kanzure: I understand how the filename is generated: given the Author string, split by spaces, pass only valid ascii characters + year... i.e.
11:59 < fltrz> 10.1002/ajh.2830420317 | 857a57a58cf2c74f0588709f4fa9a6de | Slinasi Özsoylu  | 1993
11:59 < fltrz> becomes zsoylu1993.pdf
12:00 < fltrz> but I don't know how they would handle multiple articles, same author same year
12:00 < fltrz> oh it dont matter, the hash kind of string would presumably be different
12:01 < fltrz> also not sure what hash that is "MTAuMTAwMi9hamhiLjEwMTg3" thats 24 characters
12:01 < fltrz> base64?
12:02 < fltrz> yep base64(DOI) !!!
12:02 < fltrz> kanzure: ^
12:04 < fltrz> so "http://cyber.sci-hub.cc/"+base64(DOI)+"/"+dropnonasciiorsmth(split(Author," "))+Year+".pdf"
12:05 < kanzure> confirmed? or do you get perpetual captcha?
12:06 < fltrz> I did base64decode on the mysterious hashlike string
12:06 < fltrz> I instantly recognized the DOI string I pasted in scihub to test
12:06 < kanzure> you need to test the url before requesting it in scihub because requesting it might prime the cache and we need to know if the cache is already primed
12:06 < fltrz> I did not of course verify with 50Megapapers, to check which dropnonascii-ish function was used
12:07 < fltrz> ah
12:07 < fltrz> well, my IP might still be blocked, so I should wait a few days before testing?
12:08 < fltrz> I get "https://de.hideproxy.me/onclick=%22location.href=parseURL('http://cyber.sci-hub.cc/MTAuMTAwMi9hamguMjgzMDQyMDMxNw==/zsoylu1993.pdf?download=true%27)%22"
12:08 < kanzure> i thought you had your dedonkeyproxy thing
12:08 < fltrz> so not sure where the original scihub url ends
12:08 < kanzure> "download=true" is part of scihub url
12:08 < fltrz> %27 = ?
12:08 < kanzure> this is called url encoding
12:09 < fltrz> so when I generate a test, I just try my generated string up till "...download=true" in hideproxy? or also the %27)%22?
12:12 < kanzure> the %27%27 is required for hidewhoever reasons
12:12 < kanzure> onclick= stuff is javascript embedded in your url
12:13 < kanzure> it's url-encoded javascript stuff
12:13 < kanzure> btw it's not "MTAuMTAwMi9hamhiLjEwMTg3" it's "MTAuMTAwMi9hamhiLjEwMTg3==" (base64 stuff)
12:13 < fltrz> kanzure: confirmed! generated a whole url from sql database row, presented with a captcha, pdf!
12:14 < kanzure> hmmm.
12:14 -!- gene-hacker [~tetrapod@CMU-943103.ANDREW.CMU.EDU] has joined ##hplusroadmap
12:14 < kanzure> was random sample from database?
12:14 < fltrz> kanzure: true, I didn't notice a few characters were chopped off the DOI, it still looked like DOI...
12:14 < fltrz> kanzure: yes random
12:15 < kanzure> hmmmmm.
12:15 < fltrz> but could accidentally be in a cache?
12:16 < fltrz> kanzure: if you want I can write it up as a simple java application connecting to scimag sql db, user input sql ID => output url string
12:16 < fltrz> but what I wrote above is complete
12:16 < kanzure> cloudflare behavior might be check cache else request from real thing
12:17 < fltrz> [21:09] <fltrz> so "http://cyber.sci-hub.cc/"+base64(DOI)+"/"+dropnonasciiorsmth(split(Author," "))+Year+".pdf"
12:17 < fltrz> kanzure: true
12:17 < fltrz> but would cloudflare present me a new captcha?
12:17 < kanzure> the captcha stuff is 100% scihub
12:17 < kanzure> so this is probably just cloudflare in front of their server, captcha page is definitely their server
12:18 < kanzure> give me  full url for latest one please?
12:18 < fltrz> latest what?
12:18 < kanzure> http://cyber.sci-hub.cc/MTAuMTAwMi9hamhiLjEwMTg3/leonard2003.pdf ?
12:18 < fltrz> the one I generated myself?
12:18 < fltrz> http://cyber.sci-hub.cc/MTAuMTAwMi9hamhiLjIwMDIz/plochocki2004.pdf
12:18 < kanzure> hmm it did not ask me for captcha on this link i pasted above
12:18 < fltrz> this one I generated from sql row
12:19 < kanzure> no captcha on that one either for me
12:19 < kanzure> Server: cloudflare-nginx
12:19 < kanzure> CF-Cache-Status:MISS
12:19 < kanzure> CF-RAY:39a3c7a463ac6e2c-SJC
12:19 < fltrz> weird, perhaps because I  recently used them
12:20 < kanzure> CF-Cache-Status:HIT
12:20 < kanzure> CF-RAY:39a3c871f4fb6e2c-SJC
12:20 < fltrz> kanzure: let me generate a new url string which I don't try, so you try as first person
12:20 < kanzure> got "HIT" in reply to second request
12:21 < kanzure> url is failing for http://dabamirror.sci-hub.cc/MTAuMTAwMi9hamhiLjIwMDIz/plochocki2004.pdf and dacemirror... so it seems to only be cyber?
12:22 < fltrz> kanzure: freshly made up URL http://cyber.sci-hub.cc/MTAuMTAwMi9hamltLjEwMDQ1/sauter2002.pdf (haven't tried so perhaps I made some small mistake)
12:23 < fltrz> kanzure: I never saw the dabamirror, I think I asked before where its from
12:23 < kanzure> can you give me hash for this one
12:23 < fltrz> which hash?
12:24 < kanzure> for your url, got captcha but then it worked
12:24 < fltrz> the MD5?
12:24 < kanzure> md5 for this latest row you are working with
12:24 < fltrz> md5 of sauter: e70e60e11f110912dc6883d008256c1f
12:25 < kanzure> hmm.
12:25 < fltrz> in theory you could just md5sum the pdf?
12:25 < kanzure> check out the error message here http://twin.sci-hub.cc/MTAuMTAwMi9hamltLjEwMDQ1/sauter2002.pdf
12:25 < kanzure> [2017-09-06 22:25:22] некорректный запрос MTAuMTAwMi9hamltLjEwMDQ1/sauter2002.pdf
12:26 < kanzure> but this url works: http://twin.sci-hub.cc/e0f2608efb9e6923c2cc054655bf2477/10.1111@2041-210x.12735.pdf
12:26 < fltrz> kanzure: did the "cyber" url work for you?
12:26 < kanzure> yes i just said so
12:26 < kanzure> 12:24 < kanzure> for your url, got captcha but then it worked
12:26 < fltrz> oh ok, so we are checking out all the different mirrors?
12:26 < kanzure> we're hax0ring
12:27 < fltrz> not really, we are just interested in sci-hub nomenclature whicch they openly share with your browser
12:27 < kanzure> that was my answer to your question
12:27 < fltrz> kanzure: perhaps its md5(md5()) or smth?
12:28 < fltrz> or md5(DOI)?
12:28 < fltrz> because then they don't need to do a lookup in sql
12:28 < fltrz> !!
12:29 < fltrz> kanzure: why would we be interested in the other mirrors? to find the fastest one?
12:29 < kanzure> url enumeration is an important task when taking over remote servers
12:30 < fltrz> :-o ! I'm not taking over anything :-o
12:30 < fltrz> you sound like rick from rick and morty
12:30 < kanzure> i am the most optimal rick
12:31  * fltrz cigarette
12:32 < kanzure> i am trying e70e60e11f110912dc6883d008256c1f/DOI.pdf but not sure how to encode 10.1002/ajim.10045.pdf i guess it's 10.1002%2Fajim.10045.pdf
12:32 < kanzure> not getting good result
12:35 < fltrz> kanzure: where do you get all the alternative.sci-hub.cc mirrors from? are you hoping to find FTP jackpot or smth?
12:35 < kanzure> https://torrentfreak.com/elsevier-wants-cloudflare-to-expose-pirate-sites-160917/
12:35 < kanzure> fltrz: robtex
12:37 < kanzure> fltrz: http://sci-hub.cc/downloads/doi.7z
12:37 < fltrz> what is that?
12:38 < fltrz> list of all DOI's?
12:38 < kanzure> ttps://github.com/greenelab/scihub/tree/master/download/scihub-dois
12:39 < kanzure> i don't know man; i'm just making this up as i go
12:39 < fltrz> yeah, I got the sql database
12:39 < kanzure> "62,835,101 DOI in alphabetical order"
12:39 < kanzure> check out this magic url: http://sci-hub.cc/downloads/613f/10.1007@s00262-017-2043-6.pdf
12:39 < kanzure> this loads a pdf for me
12:39 < kanzure> 613f is probably first 4 characters of the md5 of the file?
12:39 < kanzure> let's try e0f2/10.1111@2041-210x.12735.pdf
12:40 < kanzure> 404: http://sci-hub.cc/downloads/e0f2/10.1111@2041-210x.12735.pdf
12:40 < fltrz> the sql table has 64195945 rows
12:40 < kanzure> where does 613f come from
12:41 < fltrz> kanzure: how did you generate the URL @ 21:44?
12:41 < kanzure> i don't seem to be synchronized with your clock heh
12:42 < kanzure> if you mean http://sci-hub.cc/downloads/613f/10.1007@s00262-017-2043-6.pdf then my answer is random googling
12:44 < fltrz> anyway, I think knowing one mirror's formatting is sufficient?
12:45 < kanzure> this one works: http://sci-hub.cc/downloads/9f30/10.1080@15213269.2017.1328312.pdf
12:45 < kanzure> fltrz: oh i suppose. but captchas.
12:45 < fltrz> kanzure: I think captcha's everywhere? even moscow.sci-hub.cc?
12:46 < kanzure> this works: http://sci-hub.io/downloads/681b/10.1515@zpch-2017-0952.pdf
12:46 < fltrz> the four characters are?
12:47 < fltrz> what encoding changes "/" into "@" ?
12:47 < kanzure> i dunno, i thought it was hash[0:4] or hash[-4:] but nope
12:47 < kanzure> yea i dunno, that's not a thing, it might be custom
12:47 < fltrz> but then how did you find <kanzure> this one works: http://sci-hub.cc/downloads/9f30/10.1080@15213269.2017.1328312.pdf
12:47 < kanzure> google
12:47 < fltrz> oh lol ok
12:48 < fltrz> so the advantage of the later series of links is it does not involve author and year from a table?
12:48 < kanzure> and so far no captcha
12:48 < fltrz> apart from mystery 4 characters its DOI
12:48 < fltrz> oh no captcha?
12:49 < fltrz> is there an older article (non 2017) in this style?
12:50 < fltrz> kanzure: try downloading the pdf, and md5sum to see if it is present in the hash somewhere
12:50 < fltrz> I verified that the md5 in the tables was always the md5 of the .pdf
12:50 < kanzure> f0f46f8fb5223d04181bf7a4179d98c2  10.1080@15213269.2017.1328312.pdf
12:51 < fltrz> hmm nope
12:51 < kanzure> so...... the developer of this site is a php programmer. what sort of php library splits a lot of files up into directories like this?
12:52 < fltrz> kanzure: found it https://md5.gromweb.com/?string=10.1080%2F15213269.2017.1328312
12:53 < fltrz> pay attention that its md5(DOI) where DOI is original, not with / -> @
12:53 < kanzure> H(DOI) wtf??
12:53 < kanzure> who the hell designs this shit?
12:53 < fltrz> kanzure: makes sense, the users enter DOI
12:54 < fltrz> so they don't need to do sql lookups
12:54 < kanzure> you are now leet hax0r
12:54 < fltrz> lol no
12:54 < kanzure> sorry man
12:54 < kanzure> it's mandatory
12:54 < fltrz> ahaha
12:54 < fltrz> none of them have captcha?
12:55 < kanzure> good question
12:55 < fltrz> now you need boatload volunteers running sw in exchange for pennies, requesting pdf's at 1/minute
12:56 < kanzure> now we setup server in seychelles hosting center
12:56 < fltrz> still I think better to design the sw such that it goes straight to siacoin or whatever
12:56 < kanzure> http://quasinetworks.com/readme.txt
12:56 < kanzure> this is who's hosting them (or maybe cloudflare? it's not clear)
12:57 < fltrz> kanzure: and dont suddenly increase the number of downloads, or theyll detect
12:57 < fltrz> ramp it up slowly
12:58 < fltrz> kanzure: what if captcha's is with cookies, and you recently did enough captcha and thats why not presented?
12:59 < fltrz> I also dont get captcha on the same one though, so probably just really no captcha there
12:59 < kanzure> just tried it, didn't need cookie
12:59 < kanzure> let's try pre-2017 row from your database
13:00 < fltrz> 10.1002/ajmg.1320090203 | 242b8cb5cd46bac3eb1a8a4c70e32244 | E. Magenis; M. G. Brown; J. Chamberlin; T. Donlon; D. Hepburn; N. Lamvik; E. Lovrien; M. Yoshitomi; John M. Opitz | 1981
13:01 < kanzure> both of these fail:
13:01 < kanzure> http://sci-hub.cc/downloads/48d8/10.1002@ajim.10045.pdf
13:01 < kanzure> http://sci-hub.cc/downloads/d48e/10.1002@ajim.10045.pdf
13:01 < fltrz> what did you use for DOI?
13:01 < fltrz> 644b
13:02 < fltrz> https://md5.gromweb.com/?string=10.1002%2Fajmg.1320090203
13:02 < kanzure> i used 10.1002/ajim.10045
13:02 < kanzure> $ echo "10.1002/ajim.10045" | md5sum  -
13:02 < kanzure> d48ef6eca9c391a5c2c9fd550aee48d8
13:03 < kanzure> $ echo -n "10.1002/ajim.10045" | md5sum  -
13:03 < kanzure> d2ee19c6a94207a594239a33972ac64f
13:03 < kanzure> ok i will try your row
13:03 < fltrz> https://md5.gromweb.com/?string=10.1002%2Fajim.10045
13:03 < fltrz> yep d2ee
13:03 < kanzure> fail http://sci-hub.cc/downloads/644b/10.1002@ajmg.1320090203
13:03 < fltrz> continue your row
13:04 < fltrz> http://sci-hub.cc/downloads/d2ee/10.1002@ajim.10045.pdf
13:04 < kanzure> fail http://sci-hub.cc/downloads/d2ee/10.1002@ajim.10045.pdf
13:04 < fltrz> hmm
13:04 < fltrz> I think its the "." ?
13:05 < kanzure> for future reference this server is 80.82.77.83
13:06 < fltrz> oh, I should continue cleaning the kitchen etc
13:06 < kanzure> welp maybe it's 2017-only
13:06 < fltrz> perhaps
13:07 < fltrz> really strange
13:08 < fltrz> perhaps captcha free cache?
13:08 < kanzure> oh wait i have idea
13:08 < fltrz> recent ones is what everyones referencing, and they don't want to tarnish their image
13:08 < kanzure> nope fail http://sci-hub.cc/downloads/644b/magenis1981.pdf
13:09 < fltrz> note that when it worked there was only numerals, / and .
13:10 < fltrz> and when it failed it had alphabetics?
13:11 < fltrz> kanzure: where did you get "10.1002/ajim.10045" ?
13:12 < kanzure> earlier item today
13:13 < fltrz> http://sci-hub.io/downloads/d2ee/10.1002@ajim.10045.pdf just trying
13:13 < fltrz> nope
13:13 < kanzure> it was on one of the other subdomains :\
13:13 < kanzure> through the DOI base64 stuff
13:16 < fltrz> crap now even hideproxy shows me infinite captcha loop
13:18 < fltrz> its really annoying if something seems to work but then doesnt?
13:21 < fltrz> kanzure: can an article have multiple DOI strings?
13:21  * fltrz gets dizzy
13:24 < kanzure> i sure hope not. DOI would be pointless.
13:26 < fltrz> true
13:26 < fltrz> cant we ask sci-hub to beam it at gbps to the retroreflector we put on the moon?
13:28 < fltrz> no we'd need huge telescopes for sending and receiving, and easily jammed by whoever has the biggest telescope
13:29 < fltrz> did retroreflective cloth exist when they put the cube retroreflectors on the moon? perhaps less efficient per square meter, but easily much larger area per volume or weight
13:31 < fltrz> regarding sunshades for climate control, do they take into account the optical pressure?
13:31 < fltrz> it would systematically be pushed down, and not up when in the shade of the earth
13:34 < kanzure> subdomains: cyber, dabamirror, dacemirror, moscow, ocean, ololo, secure, twin, www, zeze
13:34 < kanzure> "secure" is probably for https
13:36 < kanzure> cloudflare says this thing is offline http://moscow.sci-hub.ac/b299b1a9b91f0707061aeb7351be5786/kapustin2014.pdf
13:36 < kanzure> but it works here http://ocean.sci-hub.ac/b299b1a9b91f0707061aeb7351be5786/kapustin2014.pdf
13:36 < kanzure> b299b1a9b91f0707061aeb7351be5786 is md5(DOI)
13:37 < kanzure> DOI = 10.1021/cg5000183
13:38 < kanzure> hmm.
13:40 < fltrz> so ocean is like cyber
13:40 < fltrz> oh wow .ac
13:48 -!- Anokhi [d83197f9@gateway/web/freenode/ip.216.49.151.249] has joined ##hplusroadmap
14:00 -!- Anokhi [d83197f9@gateway/web/freenode/ip.216.49.151.249] has quit [Quit: Page closed]
14:06 < fltrz> kanzure: I can imagine an alternative reason why she wants to keep sci-hub the central gateway... suppose theres fast mirrors / everyone has a local copy at some point,.. good for the group we don't need sci-hub anymore... but the legal cases against her don't disappear, and the local government will no longer suffer humiliation by extraditing or whatever
14:07 -!- Urchin [~urchin@unaffiliated/urchin] has quit [Ping timeout: 258 seconds]
14:08 < fltrz> while as long as she holds the torch for free information the local governments would loose face by extraditing her..
14:08 < fltrz> it's another reason for going directly to decentralized system, even if you intend to eventually, you don't want to end up in her position
14:09 < fltrz> she didn't see that coming, she probably thought first things first, liberate the data
14:09 < fltrz> then became dependent on it
14:11 < fltrz> the group is greedy, and will forget heroes (but not the heroism) as soon as they have what they need
14:11 -!- jaboja [~jaboja@jaboja.pl] has quit [Remote host closed the connection]
14:13 -!- Urchin [~urchin@unaffiliated/urchin] has joined ##hplusroadmap
14:15 < fltrz> kanzure: "I know almost nothing about Python but I've produced an API for sci-hub in R using Selenium (so the same approach should be possible in python). In R, the CAPTCHA issue is currently resolved by using selenium to drive a web browser & displaying a cropped screenshot of each CAPTCHA as it arises, and then prompting a user input to enter the code (you can then extract the link to the PDF and download). This isn't ideal, 
14:16 < fltrz> (I have also heard of people using python to bypass CAPTCHAs)
14:20 < fltrz> can't AI solve the captcha's
14:22 < fltrz> do we know what captcha software is being used?
14:25 -!- augur [~augur@noisebridge130.static.monkeybrains.net] has joined ##hplusroadmap
14:25 -!- g0d355__ [~lmao@104.131.75.159] has quit [Max SendQ exceeded]
14:25 < fltrz> is libgen older than sci-hub? perhaps what she meant with wanting to be the gateway is more literal: she doesn't want to store or host the pdf's just maintain the sci-hub portal to libgen
14:45 -!- jaboja [~jaboja@jaboja.pl] has joined ##hplusroadmap
14:48 < fltrz> kanzure: they are using https://github.com/dapphp/securimage/blob/master/securimage_show.php
14:49 < fltrz> so it should be possible to train a NN (since we have unlimited stimuli & ground truth pairs)
14:51 < fltrz> oh its actually http://www.phpcaptcha.org/
15:08 -!- p0nziph0ne [~p0nziph0n@unaffiliated/p0nziph0ne] has quit [Quit: Leaving]
15:14 < fltrz> kanzure: they seem to use this font? https://cooltext.com/Download-Font-Xtrusion
15:14 < fltrz> and a wordlist, and securimage configured to use single words from the wordlist
15:20 < fltrz> pretty sure the captchas can be automated away, so only this and diverse set of IP's (volunteers?)
15:27 < fltrz> LOLwut! the random number to generate the captcha is generated ... USER SIDE!
15:29 < kanzure> was transcribing presentation on signature aggregation for libsecp256k1
15:29 < fltrz> go to: https://www.phpcaptcha.org/try-securimage/ then right click the "recycle icon" for new captcha, then select "inspect element" in firefox, then notice the url to the captcha generator php & Math.random()!!
15:29 < kanzure> selenium is overkill, please don't recommend it
15:30 < fltrz> I substituted Math.random with 10, then clicked to regenerate captcha, it gave a new one, then tried regenerating again and again, but it stayed the same
15:30 < kanzure> yes it's likely to be phpcaptcha because she is php person
15:30 < fltrz> then I changed 10 to 11 and clicking changed the captcha once again...
15:31 < fltrz> so unless the captcha is simply cached, it seems the randomness is generated user side...
15:31 < kanzure> yes you usually have to seed the RNG
15:31 < kanzure> but i wouldn't want to do crypto-secure RNG stuff in php that sounds like a recipe for disaster
15:31 -!- Malvolio [~Malvolio@unaffiliated/malvolio] has quit [Ping timeout: 240 seconds]
15:32 < fltrz> so the user loads sci-hub, and proceeds until the first captcha, then modifies the regenerate button, replaces Math.random() with 10, solves the captcha once, and every time he encounters a captcha find and replace the Math.random() with 10 and always pass the same answer?
15:33 -!- gene-hacker [~tetrapod@CMU-943103.ANDREW.CMU.EDU] has quit [Ping timeout: 240 seconds]
15:34 < fltrz> no it doesn't work
15:35 < fltrz> I saw the captcha momentarily show the old captcha and then switch to the new one, so it was simply being cached :(
15:35 -!- Malvolio [~Malvolio@unaffiliated/malvolio] has joined ##hplusroadmap
16:03 -!- preview [~quassel@2407:7000:842d:4000::3] has joined ##hplusroadmap
16:31 -!- gene-hacker [~tetrapod@c-24-131-17-249.hsd1.pa.comcast.net] has joined ##hplusroadmap
17:37 -!- mrdata [~mrdata@unaffiliated/mrdata] has joined ##hplusroadmap
17:46 < fltrz> kanzure: why do you consider time travel (backwards) impossible?
17:53 -!- atrus6 [~atrus6@72.241.82.247] has quit [Read error: Connection reset by peer]
18:00 < kanzure> fltrz: i have only figured out forward time travel
18:01 < fltrz> you mean where the observer remains younger?
18:02 -!- danfox [~danfox@vulpinedesigns.co.uk] has quit [Ping timeout: 240 seconds]
18:02 < kanzure> stuff like that
18:05 -!- danfox [~danfox@159.134.107.235] has joined ##hplusroadmap
18:07 -!- aeiousomething [~aeiousome@gateway/vpn/privateinternetaccess/aeiousomething] has joined ##hplusroadmap
18:10 -!- darsie [~username@84-113-55-42.cable.dynamic.surfer.at] has quit [Ping timeout: 248 seconds]
18:17 < maaku> fltrz: what evidence do you have for time travel not being impossible?
18:18 < fltrz> fltrz: not evidence, the complement of "having evidence against somthing" is not "having evidence for something" but simply "not having evidence against something"
18:19 -!- atrus6 [~atrus6@72.241.82.247] has joined ##hplusroadmap
18:19 < fltrz> I find backwards time travel not evident at all, but I don't rule it out, and I also consider Bell's paradox as containing a crucial loophole hardly ever discussed
18:20 < fltrz> *bell's theorem
18:21 < fltrz> (it assumes spatial part of spacetime to be topologically cartesian, i.e. no handles in math terminology, or wormholes/portals lay man terminology)
18:25 < fltrz> I believe particles are wormholes (its not a new idea), and there are some indications for it (in fact Bell's theorem & EPR paradox, the results of which have been confirmed can be viewed as such an indication)
18:26 < fltrz> i.e. an entangled spatially separated part could have a large distance between them, and synchronized clocks would confirm the simultaneous measurement of complementary quantum states, such that quantum state seems to travel faster than light
18:27 < fltrz> but if those entangled particles imply a wormhole, the distance through the wormhole might be minuscule
18:28 < fltrz> such that the signal is in fact not traveling faster than light
18:28 < fltrz> *quantum signal
18:30 < fltrz> another indication are dirac strings, in the calculation for what magnetic monopole charges are admissible, one comes to notice what is termed a dirac string, it should be there, but you can place the string arbitrarily, connecting the monopole to some point at infinity
18:31 < fltrz> however if you consider multiple monopoles of opposing magnetic charges, you instead of terminate the string at infinity terminate it at a monopole of opposite charge
18:31 < fltrz> or all of them at infinity, or given 2 positive and 2 negative magnetic monopoles, connect them in one pairing, or another pairing
18:32 < fltrz> this indicates that the fields on R^3=RxRxR do not fully specify the system
18:34 < fltrz> also, when viewing particles as wormholes, there is no need (in the complicated manifold) for the concept of charge, charge is the effect of electric fields sourcing from/into the wormhole
18:34 < fltrz> charge is then emergent from electric fields on a non-factorizable space
18:34 < fltrz> so the loophole is imho quite literal!
18:36 < fltrz> another indication is our "arbitrary" 3 spatial + 1 time dimensions, why not 2+1 or 7+1?
18:37 < maaku> kudos to you for having a real response to that question.
18:38 < fltrz> suppose we create some time travel, then different histories can affect each other, i.e. probe histories at the time machine deflecting the original course of history
18:38 < fltrz> so you get new kinds of conservation laws in 4D+1T
18:39 < fltrz> to put another way, we can move in 3 dimensions, and are dragged in time, with time travel we would be able to move arbitrarily (only somewhat initially) in 4 dimensions, but in doing so create a new time dimension
18:41 < fltrz> i.e. in sci-fi stories, you can perceive 2 kinds of causility, the causality within a timeline, along what we perceive as time, but then a second kind of causality introduced by feedback by timetravel, this second kind of causality becomes a weird poorly understood time axis
18:41 < fltrz> while the old time axis slowly becomes a navigable spatial axis
18:42 < fltrz> if true, the 3-ness of our 3+1 dimensions could be an indication that the 2D+1T beings invented time travel, with our perception of the universe as a result
18:42 < fltrz> and before them 1D+1T -> 2D+1T
18:42 < fltrz> so time travel might have been discovered 3 times already
18:44 < fltrz> to what extent these patterns(creatures?) would have been conscious is hard to find out, but if we invent time travel we would still have a poor understanding or view of 4D spacetime, we would only be measuring it through experimental machines etc...
18:44 < fltrz> if we do and we conquer the universe, we don't really perceive but would cause some unknown dynamics in 4D+1T
18:44 < fltrz> the new time axis being how histories affect each other
18:48 < fltrz> causality within one such history corresponds to our normal causality along our normal time axis, causality between histories would be the 'impossible' kind of causality along some hard-to-fathom time axis
18:55 < fltrz> but I have no evidence, no
18:58 < fltrz> Bell's theorems were in a sense a response to EPR paradox, and EPR paradox was first in taking the simplyfing stance of assuming (for simplicity) spatial part of spacetime to be cartesian
18:59 < fltrz> so Bell's theorems, being a response, used the same assumption
19:02 < fltrz> EPR paradox published in 1935 Physical Review 47, Einstein Rosen bridge Physical review 48 same year
19:03 < fltrz> so Einstein et al kept this possibility in the back of their mind
19:04 < gene-hacker> FTL, relativity, causality, pick two
19:05 < fltrz> in fact Weyl already considered wormholes... and if you ever find the time even maxwell in his main treatise on electromagnetism (which I highly recommend for reading) keeps the possibility open, maxwell is more like a mathematician scrutinizing every assumption, its amazing
19:06 < fltrz> theres whole sections where he describes the problem of monodromy, and then he derives his treatise on the assumption of cartesian space, but this whole section is his literal reminder that it is just that, an assumption, he clearly in his mind leaves open the possibility of wormholes
19:08 < fltrz> he heavily suggests them, without explicitly claiming either way
19:10 < fltrz> if you are afraid of reading such a text, it is in fact rather entertaining, with random remarks that tell you things about the times he lived in
19:13 < fltrz> what I mean with he keeps the possibilities open: whereas most authors might consider different possibilities, but if they only elaborate on one they simply don't mention the other possibilities, while maxwell is extremely careful and honest: first enumerates some possibilities, and then continues with one of them, but makes sure for posterity that he has listed other possibilities he can't exclude
19:14 < fltrz> to the point he finds it necessary to prove that spherically symmetric electric potentials of particles can only take 2 specific forms/laws
19:14 < fltrz> these things were not shown in my electromagnetism courses at university, which I still hold in high regard
19:16 < fltrz> gene-hacker: define FTL travel?
19:16 < gene-hacker> any FTL
19:16 < gene-hacker> travel
19:17 < fltrz> i.e. suppose there's a portal connecting New York to alpha centauri, and the distance along the portal is 1m, and you walk at 3miles an hour through the portal, is that faster than light?
19:17 < fltrz> I agree that locally nothing can travel FTL
19:18 < gene-hacker> how do you prevent the roman rings from occuring? https://en.wikipedia.org/wiki/Roman_ring
19:18 < gene-hacker> if you can make travel only unidirectional then it's ok
19:18 < gene-hacker> but things get weird if we allow bidirectional wormholes
19:19 < fltrz> I don't think they need to be prevented, i.e. I believe we can make backward time machines
19:19 < fltrz> constantly splitting histories
19:20 -!- preview [~quassel@2407:7000:842d:4000::3] has quit [Ping timeout: 246 seconds]
19:21 < fltrz> suppose I have a inverse memory time machine (i.e. I can read out some bits that I write in later)
19:21 -!- Gurkenglas [~Gurkengla@dslb-178-000-222-214.178.000.pools.vodafone-ip.de] has quit [Ping timeout: 240 seconds]
19:22 < gene-hacker> well then you don't have causality and things get weird
19:23 < fltrz> before I turn the device on, I commit to a number N, swear to myself i will read the number from the time machine, and then enter the number 2*N/(a+b)
19:23 < fltrz> what will I read?
19:23 < fltrz> most probably sqrt(N)
19:23 < fltrz> quite possibly a number close to it
19:24 < fltrz> and very improbably a number far from it
19:25 < fltrz> and an infinitesimal probability (the "first" probe history, not yet affected by "previous" histories where I write in a result) I will read random quantum noise
19:25 -!- mrdata [~mrdata@unaffiliated/mrdata] has quit [Read error: Connection reset by peer]
19:25 < fltrz> where "first" and "previous" are not along our usual time axis, but along the emergent new time axis
19:26 < fltrz> ugh i used the wrong formula
19:26 < fltrz> 2*N(N/x+x)
19:26 < fltrz> where x was the number you read
19:26 < fltrz> so you read x, enter 2*N(N/x+x)
19:27 < fltrz> the "first" history reads noise and this initial value will start the usual approximation formula
19:28 < fltrz> you can enter this in spreadsheet for example to see the effect of iteration
19:28 < fltrz> the sqrt is always between N/x and x, for arbitrary x
19:29 < fltrz> probably (x+N/x) works just as well
19:29 < fltrz>  divided by 2
19:32 < fltrz> suppose you want to find a "preimage" nounce for a hash of a certain bitcoin difficulty, you start the machine, read out the previous nonce, check if it matches IF NOT increment the nonce and write IF IT DID write the same nonce you read
19:32 < fltrz> probe history gets random quantum noise
19:33 < fltrz> with overwhelming probability you get a valid nonce, with low probability you are not that far from it
19:33 < fltrz> but instead of going through all the nonces, just run the machine again
19:40 < fltrz> oh I forgot to mention the reason why overwhelming probability your result will be a fixed point of the map (like x'=(N/x+x)/2 or x'=validnonce(x)?x:x+1; ) you use, is because of the infinite iterations performed the majority just passes the fixed point of the map
19:41 < fltrz> only a minority of infinite iterations are not yet at a fixed point of the map
19:44 < fltrz> consider some map on 8 bits -> 8 bits, but a map such that any input eventually reaches a fixed point after sufficient iterations, there could still be multiple fixed points
19:47 < fltrz> suppose there are 2, then consider how many of the 256 possibly inputs reach fixed point A (say 200) and how many reach B (say 56), then with overwhelming probability you will reach A or B (with probabilities 200/256-epsilon and 56/256-epsilon) and only with probability 2 epsilon would you get a non fixed point
19:47 < fltrz> where epsilon is infinitesimal
19:48 < fltrz> (this is upon activating the machine, reading some x, and writing x'=f(x), doing this just once)
19:49 < fltrz> but having read x, you must compute f'(x), one iteration, and write actually enter it
19:50 < fltrz> if you don't commit to actually entering the result of a single interation, you screw up your probabilities
19:50 < fltrz> and it wont work
19:51 -!- preview [~quassel@2407:7000:842d:4000::3] has joined ##hplusroadmap
19:51 < fltrz> so regarding the roman rings, that would be "chronology violation" according to others, which I think would occur, however paradoxical it may sound
19:56 -!- yashgaroth [~yashgarot@2606:6000:cd4d:3300:f5e0:f867:a11d:8d52] has joined ##hplusroadmap
19:57 < fltrz> but then again, im probably just crazy :)
20:26 < fltrz> ugh should have never toyed with sci-hub
20:35 < fltrz> kanzure: do you know of a publicly downloadable database of article crossreferences? i.e. such that I can find all the references and cites given an article DOI or smth?
20:59 -!- aeiousomething [~aeiousome@gateway/vpn/privateinternetaccess/aeiousomething] has quit [Ping timeout: 248 seconds]
21:20 -!- mrdata__ [~mrdata@unaffiliated/mrdata] has joined ##hplusroadmap
21:20 -!- sachy [~sachy@nat.brmlab.cz] has quit [Ping timeout: 248 seconds]
21:21 -!- maaku [~mark@173.234.25.100] has quit [Ping timeout: 248 seconds]
21:21 -!- maaku [~mark@173.234.25.100] has joined ##hplusroadmap
21:21 -!- mrdata__ is now known as mrdata
21:21 -!- sachy [~sachy@nat.brmlab.cz] has joined ##hplusroadmap
21:21 -!- mrdata_ [~mrdata@unaffiliated/mrdata] has quit [Ping timeout: 248 seconds]
21:23 -!- aeiousomething [~aeiousome@gateway/vpn/privateinternetaccess/aeiousomething] has joined ##hplusroadmap
21:23 -!- redlegion [~x@unaffiliated/redlegion] has quit [Ping timeout: 260 seconds]
21:24 -!- streety_ [~streety@li761-24.members.linode.com] has quit [Ping timeout: 240 seconds]
21:24 -!- streety_ [~streety@li761-24.members.linode.com] has joined ##hplusroadmap
21:25 -!- danfox [~danfox@159.134.107.235] has quit [Ping timeout: 240 seconds]
21:26 -!- danfox [~danfox@vulpinedesigns.co.uk] has joined ##hplusroadmap
21:26 -!- redlegion [~x@104.219.54.200] has joined ##hplusroadmap
21:26 -!- redlegion [~x@104.219.54.200] has quit [Changing host]
21:26 -!- redlegion [~x@unaffiliated/redlegion] has joined ##hplusroadmap
21:27 -!- Rmesil8O4b[m] [rmesil8o4b@gateway/shell/matrix.org/x-oqzgcluhwdilxsgy] has quit [Ping timeout: 240 seconds]
21:27 -!- gwillen [~gwillen@unaffiliated/gwillen] has quit [Ping timeout: 252 seconds]
21:29 -!- gwillen [~gwillen@unaffiliated/gwillen] has joined ##hplusroadmap
21:38 -!- gene-hacker [~tetrapod@c-24-131-17-249.hsd1.pa.comcast.net] has quit [Ping timeout: 240 seconds]
21:41 -!- Rmesil8O4b[m] [rmesil8o4b@gateway/shell/matrix.org/x-oqzijkfgwbfswhnx] has joined ##hplusroadmap
21:48 -!- aeiousomething [~aeiousome@gateway/vpn/privateinternetaccess/aeiousomething] has quit [Ping timeout: 248 seconds]
21:49 < fltrz> it seems there exists no service to look up the citations, its all on the specific publisher pages
21:53 -!- vikraman [~vh4x0r@gentoo/developer/vikraman] has quit [Ping timeout: 246 seconds]
21:59 -!- aeiousomething [~aeiousome@gateway/vpn/privateinternetaccess/aeiousomething] has joined ##hplusroadmap
22:04 -!- vikraman [~vh4x0r@2605:2700:0:5::4713:952b] has joined ##hplusroadmap
22:05 -!- vikraman is now known as Guest41452
22:13 -!- jaboja [~jaboja@jaboja.pl] has quit [Remote host closed the connection]
22:27 < kanzure> fltrz: pubmedcentral might have a dowlnoadable database of citations and cross-references...
22:28 < kanzure> there should be a commoncrawl data set of open access papers, but i don't think anyone has created it yet, just need a ~1 GB dump of stuff
22:29 < fltrz> hmm, not sure where to get the ~1GB dump of stuff...
22:30 < kanzure> arxiv used to have a torrent of thigns
22:30 < kanzure> maybe biorxiv could be convinced to do torrents
22:38 -!- yashgaroth [~yashgarot@2606:6000:cd4d:3300:f5e0:f867:a11d:8d52] has quit [Quit: Leaving]
22:38 < kanzure> .title https://news.ycombinator.com/item?id=15184541
22:38 < yoleaux> Interactive Demo of Bloom Filters | Hacker News
22:38 < kanzure> https://www.jasondavies.com/bloomfilter/
22:46 < kanzure> had weird conversation with anselm earlier today
22:46 < kanzure> i asked him what twist is using and he says traditional chemistry
22:47 < kanzure> oh wait i misread, wow
22:47 < kanzure> they are using inkjets
22:47 < kanzure> and someone else totally not anselm says "They may skip the capping step altogether as modern reagent quality and process control makes it somewhat redundant"
22:53 < kanzure> wat "Everyone involved in pushing for GP-write is smoking crack and they're only doing it because it resonates with the memory of the original HGP and they don't know enough about vertebrate bio to suggest an actually creative idea."
22:53 < kanzure> but the goal is dna synthesis tech.... not vertebrate biology science.
22:59 < kanzure> anyway, if twist really is doing inkjet stuff, then why haven't they brought that up when i have been discussing inkjetting at gp-write? hmm.
23:04 < kanzure> and they are not doing electrochemical detritylation as i was wondering the other day.
23:05 < kanzure> w
23:05 < kanzure> ejriewqjqiejr0201rd
23:10 < nmz787> hmm
23:11 < kanzure> .title https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2017-September/014932.html
23:11 < yoleaux> [bitcoin-dev] Merkle branch verification & tail-call semantics for generalized MAST
23:17 -!- aeiousomething [~aeiousome@gateway/vpn/privateinternetaccess/aeiousomething] has quit [Ping timeout: 246 seconds]
23:19 -!- Urchin [~urchin@unaffiliated/urchin] has quit [Ping timeout: 258 seconds]
23:34 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has joined ##hplusroadmap
--- Log closed Thu Sep 07 00:00:11 2017