--- Log opened Thu Oct 22 00:00:54 2020
00:48 -!- SDr5 [~SDr@unaffiliated/sdr] has joined ##hplusroadmap
00:48 -!- midnight [~midnight@unaffiliated/midnightmagic] has quit [Ping timeout: 272 seconds]
00:49 -!- SDr [~SDr@unaffiliated/sdr] has quit [Ping timeout: 260 seconds]
00:51 -!- midnight [~midnight@unaffiliated/midnightmagic] has joined ##hplusroadmap
00:51 -!- saurik [saurik@carrier.saurik.com] has quit [Ping timeout: 260 seconds]
00:52 -!- faceface [~faceface@unaffiliated/faceface] has quit [Ping timeout: 260 seconds]
00:52 -!- saurik [saurik@carrier.saurik.com] has joined ##hplusroadmap
00:53 -!- faceface [~faceface@unaffiliated/faceface] has joined ##hplusroadmap
01:00 -!- SDr5 is now known as SDr
02:45 -!- spaceangel [~spaceange@ip-89-177-56-55.net.upcbroadband.cz] has joined ##hplusroadmap
03:29 < fltrz> L29Ah, my connection dropped after you said "<L29Ah> i guess surface tension is a bitch" ... so if you continued there please repaste
03:31 < fltrz> sorry my network kept dropping connections whats last thing I said?
04:08 < TMA> fltrz: last: "L29Ah, my connection..." before that "I wonderr what..." you can probably check the logs to answer this kind of questions
04:08 < fltrz> TMA, aah right, we still log
04:08 < fltrz> thanks
04:10 < TMA> bitte
04:42 < fltrz> why no scihub IRC channel?
06:07 < kanzure> well, she's on twitter, you could ask
06:31 -!- andytoshi [~apoelstra@unaffiliated/andytoshi] has joined ##hplusroadmap
06:46 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=03504e51 Murch: Fix typos in "advances in block propagation" >> http://diyhpl.us/diyhpluswiki/transcripts/gmaxwell-2017-11-27-advances-in-block-propagation/
06:46 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=fc2294b2 Bryan Bishop: Merge pull request #171 from Xekyo/typos_in_advances_in_block_propagation >> http://diyhpl.us/diyhpluswiki/
07:06 < fltrz> im twitterless, but she did post email and gpg key since last time I checked apparently
07:07 < fltrz> would like to contact her for running paper bloat analysis software
07:08 < fltrz> make her see that if we can compress dataset an order of magnitude or more, that it would start fitting on single consumer drives
07:08 < fltrz> and possibility impacts desirability, if something seems impossible or not applicable to a users hardware setup, (s)he will not consider it desirable
07:10 < fltrz> so as long as its ~100TB it will be hard to rally for support, but if we can make it ~10 or even ~1 TB suddenly average joe can visualize a single hard drive with whole of scihub as a desirable thing. perhaps for themselves, or for the family, or as a present to friends or family
07:12 < fltrz> also in single HDD regime, it will be easier to convince people to sign a petition forcing congress to entertain the idea of preloading HDD's with scientific literature
07:13 < fltrz> nobody wants to hear about complicated provably random subsamples of the literature, because lack of understanding is misinterpreted as lack of feasibility or possibility and hence lack of desirability
07:14 < fltrz> congress would reject the idea of course, but its 2020, and time to ventilate and force others to ventilate their true stance
07:28 < lsneff> I mean, the entire scihub as text would probably be tiny. Pdfs are way larger than the text they contain.
07:47 < fltrz> yes, also the fonts are repeated
07:49 < fltrz> its a relativeely small set of "papers" that are in fact full journal magazine (i.e. instead of the specific pages of the article), some pdfs that are just scanned, some that use insane images, extreme resolutions or raster for essentially vector graphics... also specific substreams in pdf are repeated
07:50 < fltrz> would need to run a counter on character definitions and count occurences
07:51 < fltrz> or first just hash the fonts in each pdf and count the occurences of each, if theres a long tail, investigate if its slight variations in fonts because of extra characters etc...
07:52 < fltrz> need some kind of tree breakdown by size for the streams in pdf
07:52 < fltrz> like the tools for inspecting HDD space usage
07:52 < fltrz> all the headers for a specific journal
07:53 < L29Ah> fltrz: poke her thru Tox
07:53 < fltrz> i have to look up what Tox is first
08:10 < lsneff> Any idea the size of scihub when compressed?
08:11 < lsneff> Any lossless compression should be able to deduplicate fonts, assuming all the papers are compressed into a small number of archives.
08:16 < L29Ah> 18:10:29]<lsneff> Any idea the size of scihub when compressed?
08:16 < L29Ah> see libgen
08:17 < L29Ah> you can download some chunks of it as torrents and try your compression innovations
08:18 < L29Ah> but i doubt you fit it into any commercially available single HDD
08:19 < lsneff> You can buy a 16tb hard drives on amazon
08:24 < fltrz> ignoring the mongoloid scanned papers, and the few extreme sized ones, I think it can be compressed to single HDD size losslessly. once we accept lossy encodings, we can shave off a few bits of resolution on each vector coordinate etc
08:24 < lsneff> Does scihub actually store anything? Or does libgen mirror all the papers for them.
08:25 < fltrz> thats very obscure to me
08:26 < fltrz> I mean we can probably find a compromise between lossy and lossless, scanning is an inherently lossy process
08:26 < L29Ah> lsneff: afaiu the latter
08:27 < lsneff> I don't think it's a good idea to lossy compress the entirety of human scientific research
08:27 < L29Ah> fltrz: yeah if you drop the scanned papers, 16TB should be doable
08:27 < fltrz> proper vectors graphics could be compressed losslessly except perhaps some of the insane vector graphics with moleccular orbitals 3d rendered and shaded to beyond typical printer resolution
08:28 < fltrz> L29Ah, we should try to "OCR" the the scanned papers to proper vector graphics though
08:29 < fltrz> just accept that its not lossless
08:29 < lsneff> The obvious end goal for libgen is to host from orbit.
08:30 < fltrz> perhaps they do and this knowledge is censored in the west :P
08:31 < fltrz> I can't shake the impression that sci-hub .se and .tw are actually western middlemen between sci hub and the user. sweden, taiwan are western aligned
08:31 < L29Ah> libgen has multiple official and an unknown number of unofficial mirrors so shutting it down won't achieve much anyway
08:32 < fltrz> also hesitant to contact "elbakyan" for this reason
08:33 < L29Ah> i've seen her in person and i must say she isn't the easiest person to communicate with, but she knows her shit well
08:34 < fltrz> if the site was supported by foreign nation states, it'd seem like a trivial exercise to host from space and to mention this fact on sci-hub .se and .tw
08:34 < L29Ah> i doubt it's supported by any state
08:34 < L29Ah> it's more like cat'n'mouse games like thepiratebay
08:35 < lsneff> Is that still up?
08:35 < fltrz> so the relative lack of the articles of chinese origin is more because people there are more afraid of leaking repercussions?
08:35 < lsneff> There's probably not much interest
08:35 < L29Ah> lsneff: it's walking dead
08:36 < L29Ah> today everyone seem to have shifted to rutracker.org
08:37 < fltrz> lsneff, why would that be? because it feels like an 1800 costume drama where they are expected to communicate their findings in english and it offends many of them?
08:38 < lsneff> I mean from the other side
08:38 < L29Ah> i'm wondering tho why sci-hub .onion is long dead
08:38 < lsneff> My guess is there isn't much interest in those who consume scihub to have access to chinese papers.
08:38 < lsneff> That is strange
08:38 < lsneff> You'd think it would be on tor first
08:38 < L29Ah> i don't understand chinese so i don't have much interest in chinese papers
08:39 < fltrz> L29Ah, I am talking about english language papers by chinese
08:39 < L29Ah> and i got an impression that worldwide scientific communication language today is English
08:39 < L29Ah> fltrz: i've seen lots of these
08:39 < lsneff> I have as well
08:39 < L29Ah> from scihub
08:40 < fltrz> L29Ah, I mean not by chinese people working in the west
08:40 < fltrz> I mean from english language non western journals
08:40 < L29Ah> me too: chinese people working in chinese institutions
08:40 < fltrz> like the local university journals etc
08:40 < L29Ah> ah, not sure about the journals
08:40 < L29Ah> say, in russia there aren't many non-russian-language journals
08:41 < fltrz> the same authors often have papers in some western journals and then a load on eastern journal
08:41 < fltrz> so its written by the same (non) intelligent person
08:41 < fltrz> so its not that the quality is much lower
08:42 < fltrz> its just that those articles will not be available in sci hub as far as I find the case :(
08:42 < fltrz> I never tried to compare earlier when .onion was still up
08:43 < lsneff> <fltrz> so its written by the same (non) intelligent person
08:43 < lsneff> ?
08:44 < fltrz> if elbakyan was forced to cooperate with western authorities, this could be a honey pot, and being served chinese IP could be used in court as "patriotic" as opposed to "traitorish"
08:45 < fltrz> lsneff, yes, so its not that "sci hub users shouldn't be interested in lo quality foreign journal" because its the same person continuing research in chinese journals
08:45 < fltrz> the safest move would be to complain about the low amount of chinese journal material :)
08:48 < fltrz> getting involved with compression seems benign too if current sci-hub turns out honeypot: in the end we just help the journals store their papers more compactly, perhaps it results in custom minipdf format
08:55 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has joined ##hplusroadmap
09:11 < fltrz> anyone here familiar with EDM or specifically wire-EDM?
09:18 < lsneff> You making one?
09:18 < lsneff> I'm thinking about whether I need a micro-edm to make flextures for nano positioners for an AFM
09:21 < fltrz> lsneff, i'm on a general EDM reading binge, I want to make a micro EDM (point milling, not yet wire)
09:22 < fltrz> its just that I had an idea for wire EDM that I don't see in the literature
09:22 < lsneff> Oo, do share
09:22 < fltrz> : with wire EDM they use a fine metal wire which unspools from the fresh spool and onto the used spool, which is usually returned as scrap base material
09:23 < fltrz> if you spool to fast you waste your expensive wire (for non micro EDM typically meters per second), if you spool to slow the wire breaks occasionally and you need rethreading...
09:24 < fltrz> so the wire is both the tool electrode and a mechanical positioner due to its ductility and tensile strength
09:24 < fltrz> the thinner the wire, the faster / less energy the cutting because you cut a smaller width
09:25 < fltrz> so I thought, why not take some 12K continuous filament carbon fiber tow, electrodeposit copper faradayically in copper sulfate solution
09:26 < fltrz> then the CF servers the mechanical strength, while the copper layer serves as the conductive electrode
09:26 < fltrz> measuring resistance indicates where on the spool the copper is getting thin, and you can selectively reapply copper in the thin regions of the wire
09:27 < L29Ah> sounds expensive
09:27 < fltrz> carbon fibre is not expensive? the copper wear is in solution, so can be recovered chemically
09:28 < fltrz> the problem is having the wire serve both for tensioning to a straight line and as electrode that wears out
09:28 < fltrz> separate the functions in a way you can monitor wear and reapply the copper
09:29 < fltrz> currently the used wires are perfectly usable for large sections of their length
09:30 < fltrz> yet they return the old wire and buy expensive new wire because they can't afford the downtime of rethreeading after the broken section too often
09:30 < L29Ah> copper is much cheaper than carbon fiber'
09:31 < fltrz> carbon fibre (with a layer of copper) will be much thinner say 10micron than whatever wire is used currently
09:31 < fltrz> so machining is more efficient
09:33 < fltrz> the used spools of carbon fiber with irregular copper clad can be rejuvenated (multiple in parallel if necessary)
09:35 < fltrz> theres 2 options for rejuvenating the wire: actively sense which sections need more or less copper, or alternatively simply faradayically remove all the copper on the CF, then reapply uniformly by electrodeposiiting copper at a constant feed rate of carbon fiber
09:37 < fltrz> a cheap spool of carbon fiber contains miles and miles of carbon fiber easily
09:38 < fltrz> this also reduces dependency on the EDM wire suppliers
09:53 < fltrz> lsneff, I was thinking of using CF as electrodes for micro EDM (then discovered it was done before, but I still intend to try that path since its essentially infinite electrode at ~5 micron diaeeter)
09:53 < fltrz> while thinking of this I had the copper on CF wire EDM idea, even though I don't intend to build a wire EDM immediately
09:59 < fltrz> prior art here: https://www.sciencedirect.com/science/article/pii/S2212827113001455 (open access)
10:47 < fltrz> theres a *lot* of literature on micro EDM apparently
11:42 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=7cc48479 Adam Jonas: add dev++ 2017 Peer-to-Peer Network and Mempool with jnewbery >> http://diyhpl.us/diyhpluswiki/transcripts/scalingbitcoin/stanford-2017/edgeplusplus/p2p-john-newbery/
11:42 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=2303ed5f Bryan Bishop: Merge pull request #172 from adamjonas/add-p2p-jnewbery >> http://diyhpl.us/diyhpluswiki/
12:11 < kanzure> someone in here was building a wire-EDM machine
12:11 < kanzure> heath: for fltrz do you know who was doing the wire-EDM stuff?
12:11 < kanzure> maybe phreedom?
12:12 < kanzure> fltrz: btw someone has already converted the scihub data set to plaintext
12:12 < kanzure> or maybe it was the libgen dataset.
12:13 < kanzure> hrm
12:13 < fltrz> kanzure, where can I learn more about scihub plaintext? libgen also interesting
12:18 < fltrz> I wonder what size it was and if already compressed or not
12:18 < fltrz> kanzure, if it happened, sci-hub sounds more likely as the percentage of scans is even higher for books
12:46 < fltrz> kanzure, phreedom is not an active nick, do you know if he's just offline, or if its a slightly different spelling?
12:46 < fltrz> or how phrequently he joins ##hplusroadmap
13:03 -!- darsie [~kvirc@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 240 seconds]
13:07 < fltrz> Researchers at the Harvard Nanopore group have engineered solid state pores with single walled carbon nanotubes across the diameter of the pore.[26] Arrays of pores are created and chemical vapor deposition is used to create nanotubes that grow across the array. Once a nanotube has grown across a pore, the diameter of the pore is adjusted to the desired size. Successful creation of a nanotube coupled with a pore is an important step towards
13:07 < fltrz> identifying bases as the ssDNA translocates through the solid state pore.
13:07 < fltrz> ^ from wikipedia
13:08 < fltrz> what is meant with "Successful creation of a nanotube coupled with a pore is an important step towards identifying bases as the ssDNA translocates through the solid state pore." ?
13:14 -!- justan0theruser [~justanoth@unaffiliated/justanotheruser] has joined ##hplusroadmap
13:14 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has quit [Ping timeout: 260 seconds]
13:27 < fltrz> I'm trying to find some actual numbers for nanopore dimensions required to sequence DNA?
13:34 < fltrz> seems 2-10 nm
13:40 -!- filipepe [uid362247@gateway/web/irccloud.com/x-emtuulvqpffsbvme] has joined ##hplusroadmap
14:26 < fltrz> kanzure, I found http://gnusha.org/logs/other/phreedom_edm.log thats 10 years ago
14:34 < kanzure> ah, well there you go
14:36 < fltrz> no clue how or where to contact the person though.. im still perusing this log, but don't actually see edm mentioned yet. unless hes using EDM to make his setups
14:39 < kanzure> processing
14:41 < fltrz> " < Phreedom> gene: the same way they make it at a fab: you need PZT powder(made by mill or hydrothermal), press(carjack will do) and a furnace"
14:41 < fltrz> what happens to oil in carjack when you put it in a furnace?
14:48 -!- Sir_Alexei [uid348072@gateway/web/irccloud.com/x-hlzierrwcyaudkug] has joined ##hplusroadmap
14:55 < fltrz> does anyone know how usable the piezo cylinders in lighters are for positioning? I'm assuming they are properly poled otherwise they couldn't generate a voltage to ignite the flame. i'm hoping they behave enough like a transducer so voltages result in reproducible translations
14:55 < fltrz> if so thats dirt cheap micro actuators
14:58 < lsneff> fltrz: https://dberard.com/home-built-stm/scan-head/
14:59 < fltrz> lsneff, yeah I know about the piezo quadrant trick, but was thinking more for positioning stages
15:00 < fltrz> the range and load for the piezo quadrants is too low
15:00 < fltrz> (thinking of microstepper)
15:02 < fltrz> not sure if phreedom only wanted to use interferometers for the axes, or within the stepper to detect correct touchdown distance to reduce wear and noise.
15:02 < lsneff> Ah, a nanopositioner
15:02 < fltrz> yeah
15:06 -!- spaceangel [~spaceange@ip-89-177-56-55.net.upcbroadband.cz] has quit [Remote host closed the connection]
15:08 < fltrz> lsneff, you think the piezo cylinders from disposable lighters would work?
15:08 < lsneff> No idea
15:09 < heath> kanzure: yeah, that'd be phreedom
15:09 < lsneff> Whatsit look like?
15:09 < heath> unfortunately, when russian invaded ukraine, he had to leave his workshop behind
15:09 -!- shawwwn [uid6132@gateway/web/irccloud.com/x-rcxjbuvqnyybwzhf] has joined ##hplusroadmap
15:10 < fltrz> lsneff, a cerammic cylinder, with mmetalliized top and bottom iirc (it was hard to pry open the plastic casing and my memory is hazy, but that was without proper tools, so now would be much easier
15:10 -!- Sir_Alexei [uid348072@gateway/web/irccloud.com/x-hlzierrwcyaudkug] has quit []
15:10 < fltrz> heath, what does phreedom do now?
15:11 < lsneff> fltrz: might work, only way is to try
15:12 < heath> related https://patents.google.com/patent/US5449313A/en
15:18 -!- yashgaroth [~ffffffff@c-73-147-55-120.hsd1.va.comcast.net] has joined ##hplusroadmap
15:23 < lsneff> An open source nanopositioner would be awesome
15:30 < fenn> a system of levers made from flexures can have a stupidly high leverage ratio
15:35 < lsneff> Well, you can do sub-nm movements with raw piezoelectrics without leverage
15:35 < fenn> no position feedback tho
15:36 < fenn> i guess that's a problem with most systems
15:36 < fenn> mounting a piezo stage on the end of a micro-manipulation stage could have benefits?
15:37 < fenn> i don't know what people want nano positioning for in the first place
15:37 < lsneff> Yeah, that's true, you need the feedback from some other, application specific, source
16:00 -!- filipepe [uid362247@gateway/web/irccloud.com/x-emtuulvqpffsbvme] has quit [Quit: Connection closed for inactivity]
16:20 -!- filipepe [uid362247@gateway/web/irccloud.com/x-tojmjcodklukdxjx] has joined ##hplusroadmap
16:38 < docl> maybe the car jack stays outside the furnace and pushes a (steel e.g) piston that extends into the furnace
16:39 < docl> dan berard's site also has some repurposed capacitors that work as actuators. not sure if they are good for nano scale
16:40 < docl> he polls them by running current through to heat them up
16:42 -!- acertain_ [uid470584@gateway/web/irccloud.com/x-nuijhjibtvepkxua] has joined ##hplusroadmap
16:43 < docl> https://dberard.com/2015/08/16/mlcc-piezo-actuators/
16:43 -!- acertain [~acertain@unaffiliated/fread2281] has quit [Disconnected by services]
16:43 -!- acertain_ is now known as acertain
16:43 -!- acertain- [~acertain@unaffiliated/fread2281] has joined ##hplusroadmap
16:43 < fenn> http://youtu.be/Ky5l9ZxsG9M the starship SN8 nosecone stacking in progress
16:46 -!- acertain- [~acertain@unaffiliated/fread2281] has quit [Client Quit]
16:46 < fenn> another perspective http://youtu.be/c5Mir9LC3PY
16:48 < lsneff> The size of the people standing on the lifts really put it into perspective.
16:48 < lsneff> Absolutely massive amount of cargo space
16:49 < lsneff> docl: Dan Berard eventually ended up using piezo tubes as an actuator
16:51 < lsneff> video compression having a hell of a time of it on the labpadre stream
16:53 < fenn> it's more of an automatic denoising setting problem that happens around sundown
17:25 -!- yashgaroth [~ffffffff@c-73-147-55-120.hsd1.va.comcast.net] has quit [Quit: Leaving]
17:38 < lsneff> Ah, I bet I could order SLA or jetted 3d printed parts to make flexures for nanopositioners
17:43 -!- Sir_Alexei [uid348072@gateway/web/irccloud.com/x-rznyukzqqxfokitq] has joined ##hplusroadmap
17:49 -!- shawwwn [uid6132@gateway/web/irccloud.com/x-rcxjbuvqnyybwzhf] has quit [Quit: Connection closed for inactivity]
19:20 -!- filipepe [uid362247@gateway/web/irccloud.com/x-tojmjcodklukdxjx] has quit [Quit: Connection closed for inactivity]
19:42 -!- justan0theruser [~justanoth@unaffiliated/justanotheruser] has quit [Ping timeout: 272 seconds]
20:32 -!- shawwwn [uid6132@gateway/web/irccloud.com/x-lqgvqbypcwazudbz] has joined ##hplusroadmap
20:36 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has joined ##hplusroadmap
20:39 -!- darsie [~kvirc@84-113-55-200.cable.dynamic.surfer.at] has joined ##hplusroadmap
21:12 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has quit [Ping timeout: 272 seconds]
22:42 -!- shawwwn [uid6132@gateway/web/irccloud.com/x-lqgvqbypcwazudbz] has quit [Quit: Connection closed for inactivity]
22:44 -!- Sir_Alexei [uid348072@gateway/web/irccloud.com/x-rznyukzqqxfokitq] has quit []
23:31 -!- aztec [~Aztec03@unaffiliated/aztec03] has quit [Ping timeout: 260 seconds]
23:31 -!- Aztec03 [~Aztec03@aztec.dog] has joined ##hplusroadmap
23:31 -!- Aztec03 [~Aztec03@aztec.dog] has quit [Changing host]
23:31 -!- Aztec03 [~Aztec03@unaffiliated/aztec03] has joined ##hplusroadmap
--- Log closed Fri Oct 23 00:00:55 2020