--- Log opened Thu Aug 24 00:00:58 2017
00:07 < nmz787> http://www.sciencedirect.com/science/book/9780323313025
00:07 < nmz787> .title
00:07 < yoleaux> Microsystems for Bioelectronics - (Second Edition) - ScienceDirect
00:22 -!- jtimon [~quassel@173.29.134.37.dynamic.jazztel.es] has quit [Ping timeout: 276 seconds]
00:40 < nmz787> http://cidarlab.org/fluigi-project/
00:40 < nmz787> .title
00:40 < yoleaux> CIDAR at Boston University – Fluigi Project
00:40 < nmz787> "Fluigi is a Computer Aided Design (CAD) framework for creating microfluidic devices for use in synthetic biology."
00:53 -!- Malvolio [~Malvolio@unaffiliated/malvolio] has quit [Ping timeout: 260 seconds]
00:53 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has joined ##hplusroadmap
00:55 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has quit [Read error: Connection reset by peer]
00:55 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has joined ##hplusroadmap
01:00 -!- emeraldgreen [~user@92.222.68.248] has joined ##hplusroadmap
01:06 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has quit [Remote host closed the connection]
01:10 -!- emeraldgreen [~user@92.222.68.248] has quit [Ping timeout: 240 seconds]
01:12 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has joined ##hplusroadmap
01:16 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has quit [Ping timeout: 248 seconds]
01:27 -!- emeraldgreen [~user@188.227.115.178] has joined ##hplusroadmap
01:51 -!- TC [~talinck@cpe-174-97-113-184.cinci.res.rr.com] has joined ##hplusroadmap
01:52 -!- TC is now known as Guest33735
01:53 -!- Malvolio [~Malvolio@unaffiliated/malvolio] has joined ##hplusroadmap
01:55 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has quit [Ping timeout: 252 seconds]
01:55 -!- hehelleshin [~talinck@cpe-174-97-113-184.cinci.res.rr.com] has quit [Ping timeout: 252 seconds]
02:09 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has joined ##hplusroadmap
02:10 -!- justan0theruser [~justanoth@unaffiliated/justanotheruser] has quit [Ping timeout: 240 seconds]
02:15 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has quit [Ping timeout: 248 seconds]
02:33 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has joined ##hplusroadmap
02:48 -!- emeraldgreen [~user@188.227.115.178] has quit [Ping timeout: 248 seconds]
02:49 -!- emeraldgreen [~user@92.222.68.248] has joined ##hplusroadmap
03:02 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has joined ##hplusroadmap
03:06 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has quit [Ping timeout: 248 seconds]
03:11 -!- emeraldgreen [~user@92.222.68.248] has quit [Ping timeout: 240 seconds]
03:21 -!- bluebear_ [~dluhos@80.95.97.194] has joined ##hplusroadmap
03:26 -!- preview [~quassel@2407:7000:842d:4000::3] has quit [Ping timeout: 276 seconds]
03:29 -!- emeraldgreen [~user@92.222.68.248] has joined ##hplusroadmap
03:59 -!- preview [~quassel@2407:7000:842d:4000::3] has joined ##hplusroadmap
04:07 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has joined ##hplusroadmap
04:11 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has quit [Ping timeout: 246 seconds]
04:32 < kanzure> nmz787: their software doesn't seem to be available online. maybe you should email them?
04:50 -!- jtimon [~quassel@173.29.134.37.dynamic.jazztel.es] has joined ##hplusroadmap
05:16 -!- fltrz [d5d38b31@gateway/web/freenode/ip.213.211.139.49] has joined ##hplusroadmap
05:25 < kanzure> fltrz: welcome back.
05:26 < kanzure> fltrz: why you using weirdo web interface to freenode irc?
05:26 < kanzure> there are some irc client recommendations listed on this page http://diyhpl.us/wiki/hplusroadmap/
05:28 -!- preview [~quassel@2407:7000:842d:4000::3] has quit [Read error: Connection reset by peer]
05:28 < fltrz> yeah I should, lazy I guess
05:28 < fltrz> does it bother you guys?
05:30 -!- preview [~quassel@2407:7000:842d:4000::3] has joined ##hplusroadmap
05:30 < kanzure> merely seems impractical
06:08 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has joined ##hplusroadmap
06:10 < fltrz> "problem is overheating and engineers haven't looked at the problem yet.   should be easy to cool the inkjet printheads." what about gravity heat pipe? i.e. thermosyphon, say alcohol in contact with part of printhead, evaporates and condenses on a large cooled area, returns in liquid form by gravity
06:11 < fltrz> pressure in the heatpipe determines boiling point temperature
06:12 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has quit [Ping timeout: 255 seconds]
06:27 < kanzure> why specifically that technique?
06:32 < fltrz> I was assuming that the available area for cooling might be small with many nozzles together, so the highest heat conductivity
06:32 < fltrz> heat pipes can have ridiculous heat conductivities, because they use the latent heat in the phase transition
06:33 < kanzure> i am not sure where the cooling/heat transfer has to occur. might work even if it's entire centimeters away. i have no idea.
06:33 < fltrz> so with a small area it could still practically instantaneously boil a liquid, which is quickly replaced
06:33 < kanzure> i also don't know what temperatures are involved here. it could be "overheating" that is solved entirely by ambient cooling...
06:34 < fltrz> indeed, I don't have a real visualization of the setup you envision, just thought it might be cramped so I thought high heat flux per square cm would be desirable
06:34 < kanzure> well for the inkjet stuff it might be helpful to look at http://bioinformatics.org/pogo/
06:34 < kanzure> http://bioinformatics.org/pogo/POSAM_Man_Ch1_Assembly_v1-2_040601.pdf
06:35 < kanzure> http://bioinformatics.org/pogo/POSAM_Man_Ch2_User_v1-0_040414.pdf
06:36 < fltrz> your inkjet printheads will be thermal, or piezo?
06:36 -!- preview [~quassel@2407:7000:842d:4000::3] has quit [Ping timeout: 246 seconds]
06:36 < kanzure> iezo
06:36 < kanzure> .. piezo..
06:37 < fltrz> perhaps not necessary at all, but am no expert
06:39 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has joined ##hplusroadmap
06:40 < kanzure> my current confusion is re: why so few inkjet research articles. was this all done in the 80s? i am mostly just seeing patents showing up in search results.
06:40 < kanzure> perhaps inkjet printing resolution reached some threshold and people stopped giving a crap?
06:40 < kanzure> where's the engineering limits characterization papers O_O
06:45 < fltrz> good question, I think a lot of research is hidden in companies, they only give the nuts they cant or dont have patience to crack to the academic world?
06:46 < fltrz> what limits are you most interested in? speed or resolution?
06:47 < fltrz> I guess your are right, that resolution was good enough for printing photographs, and there was no more demand for higher resolution?
06:47 < kanzure> resolution, throughput, compatibility with phosphoramidite chemistry reagents
06:48 < kanzure> speed is sorta important but really only i care about total throughput (so just bump up the number of nozzles if piezo fire rate is maxed out)
06:48 < kanzure> pogo/posam is a good overview but there's another one http://diyhpl.us/wiki/dna/synthesis/notes/
06:48 < kanzure> current status of this work is that i want to hire a chemist to review the chemistry and engineering compatibility issues (materials, humidity, temperature, other requirements)
06:50 < fltrz> the "dot"s in graphics terminology, I fear they don't correspond with single drops?
06:51 < kanzure> arrays in biology are really arrays
06:51 < fltrz> I see
06:51 < kanzure> you can buy 4096x4096 plates where each well is a physical indentation on the plate surface
06:51 -!- cluckj [~cluckj@static-173-59-27-112.phlapa.ftas.verizon.net] has joined ##hplusroadmap
06:52 < fltrz> I wonder how accurate the lateral positioning of printerheads is
06:52 < fltrz> or the heads will be static?
06:53 < kanzure> you can do both. also you can use autoencoders if you are bad at actuation/motion control.
06:54 < kanzure> in this project we were planning a stepper motion system using screw rod actuation with 1.5875 microns per step http://diyhpl.us/laser_etcher/laser_etcher/
06:55 < TMA> fltrz: about the same as stated dpi of the printer (not that you can stop at the precise position, but you can make the drop fly off at the correct position)
07:07 < kanzure> http://evobliss-project.eu/ "Liquid Handling Robots – You will build your own advanced liquid handling robot EvoBot based on our open-source design. You will only pay for materials (€1000) and will be instructed in how to assemble the system and apply it to simple tasks. After the summer school you take the robot with you."
07:10 < fltrz> would the diamond NV centers also be present in diamond coatings of  diamond coated glass (used in counter bar code readers in supermarkets, supposedly cheap because of mass production)
07:11 -!- Gurkenglas_ [~Gurkengla@dslb-178-008-180-030.178.008.pools.vodafone-ip.de] has joined ##hplusroadmap
07:17 < fltrz> youtube.com/watch?v=x7onZGqrYyY @0m20s or so
07:17 < kanzure> .title https://www.youtube.com/watch?v=x7onZGqrYyY
07:18 < yoleaux> Building Prototypes Dan Gelbart part 6 of 18 Coatings - YouTube
07:18 < kanzure> oh yes we have discussed these videos in the logs somewhere.
07:22 < fltrz> I was reading the text you mentioned https://arxiv.org/pdf/1708.06158.pdf where they use NV centers as magnetometers
07:23 < fltrz> which reminded me of the diamond coated panels at the start of this video
07:29 < fltrz> oh it IS zeeman splitting
07:36 < fltrz> "During chemical vapor deposition of diamond, a small fraction of single substitutional nitrogen impurity (typically <0.5%) traps vacancies generated as a result of the plasma synthesis. Such nitrogen-vacancy centers are preferentially aligned to the growth direction.[13]" yeah they should have NV-centers
08:40 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has joined ##hplusroadmap
08:41 < kanzure> .title https://news.ycombinator.com/item?id=15089476
08:41 < yoleaux> Ask HN: What is your all time favorite CS paper? | Hacker News
09:03 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has quit [Ping timeout: 240 seconds]
09:10 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has joined ##hplusroadmap
09:16 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has quit [Quit: Leaving.]
09:48 -!- gwillen [~gwillen@unaffiliated/gwillen] has joined ##hplusroadmap
09:52 -!- Gurkenglas_ [~Gurkengla@dslb-178-008-180-030.178.008.pools.vodafone-ip.de] has quit [Ping timeout: 255 seconds]
10:11 < kanzure> .title https://www.youtube.com/watch?v=bRcegZugqfY
10:11 < yoleaux> Time-Lock Puzzles from Randomized Encodings - YouTube
10:12 -!- andytoshi [~apoelstra@unaffiliated/andytoshi] has quit [Ping timeout: 240 seconds]
10:15 -!- andytoshi [~apoelstra@96.53.77.134] has joined ##hplusroadmap
10:18 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has quit [Remote host closed the connection]
10:28 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has joined ##hplusroadmap
10:29 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has quit [Remote host closed the connection]
10:32 -!- Urchin [~urchin@unaffiliated/urchin] has quit [Ping timeout: 258 seconds]
10:34 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has joined ##hplusroadmap
10:38 -!- augur [~augur@c-73-71-242-163.hsd1.ca.comcast.net] has quit [Ping timeout: 248 seconds]
10:52 -!- andytoshi [~apoelstra@96.53.77.134] has quit [Changing host]
10:52 -!- andytoshi [~apoelstra@unaffiliated/andytoshi] has joined ##hplusroadmap
11:14 -!- sachy [~sachy@nat.brmlab.cz] has joined ##hplusroadmap
11:44 -!- Darius [~quassel@66-215-89-229.dhcp.psdn.ca.charter.com] has joined ##hplusroadmap
12:15 -!- bluebear_ [~dluhos@80.95.97.194] has quit [Quit: Leaving.]
12:46 < fenn> i thought barcode reader windows were alumina (sapphire)
12:49 < fenn> "methods of manufacturing sapphire today are variations of the Czochralski process, ... a tiny sapphire seed crystal is dipped into a crucible made of the precious metal iridium or molybdenum, containing molten alumina, and then slowly withdrawn upward at a rate of 1 to 100 mm per hour. The alumina crystallizes on the end, creating long carrot-shaped boules of large size up to 200 kg in mass."
13:09 < juul> yes i'm pretty sure they are
13:09 < kanzure> juul: why you using DHT
13:10 < juul> huh?
13:10 < kanzure> bionet
13:11 < juul> for discovering the other bionet nodes
13:11 < kanzure> the webapp has to discover other web servers?
13:12 < juul> yeah, think of it as decentralized github
13:12 < juul> we currently only have the global search working
13:12 < juul> but we're working on master-master replication between bionet nodes
13:13 < juul> so when you grant someone access to your project, and they happen to be on another node, the nodes will keep a continuous master-master replication for that project going
13:13 < kanzure> "decentralized github" so.. git?
13:13 < juul> well kinda, but git doesn't do what we need
13:14 < kanzure> ok sounds like you should have used a distributed database?
13:14 < juul> yes but the problem with all of those is how do you decide who stores what and what about access control
13:15 < juul> but i'm looking at the new stuff mafintosh added to hypercore for the dat project
13:15 < juul> looks pretty good
13:15 < juul> i wrote these a while back: https://github.com/biobricks/pastlevel
13:15 < juul> and https://github.com/biobricks/commitdb
13:16 < juul> which are generalized versions of the git data structure that can track key-value stores rather than just files, don't care if you keep multiple heads and allow simultaneous multi-user write access to a single repo
13:16 < juul> so that solves the the major shortcomings of git
13:17 < juul> but we might end up using hypercore and the other hyper* tools instead. i have to spend some time looking at the latest stuff they added
13:18 < kanzure> "git data structure" basically merkle dag stuff
13:18 < kanzure> "key-value store"  use file as key, file contents as value?
13:18 < kanzure> hm.
13:19 < kanzure> okay
13:20 -!- Darius [~quassel@66-215-89-229.dhcp.psdn.ca.charter.com] has quit [Remote host closed the connection]
13:22 < juul> yeah it's merkle dag stuff
13:23 < juul> yeah it's possible to make a simple key-value store that way, but there's a reason no-one does it
13:24 < juul> hypercore started as thing on top of the levelup api but ended up implementing their own key-value store since they ran into performance bottlnecks
13:39 < maaku> < juul> yeah, think of it as decentralized github <-- you mean git?
13:46 -!- preview [~quassel@2407:7000:842d:4000::3] has joined ##hplusroadmap
14:32 -!- augur [~augur@noisebridge130.static.monkeybrains.net] has joined ##hplusroadmap
14:48 -!- emeraldgreen [~user@92.222.68.248] has quit [Ping timeout: 246 seconds]
14:58 -!- emeraldgreen [~user@92.222.68.248] has joined ##hplusroadmap
15:19 < juul> maaku: yes kanzure already mentioned that and i replied. if you ever look into the crazy shit they had to do to get github working you'll start to realize that git has some serious limitations around multi user access
15:20 < kanzure> cc jrayhawk
15:21 < juul> so if anyone is looking at the bionet code right now you might realize that the UI code is a bit of a mess and the decentralized search isn't very scalable
15:22 < juul> we're in the middle of restructuring the UI and we hired someone to implement a gnutella-like scalable flooding search
15:25 < kanzure> juul: relatedly,
15:25 < kanzure> https://public-inbox.org/git/148971018136.2144.12683278043600094739@richardiv.omgwallhack.org/
15:25 < kanzure> 19:24 < jrayhawk> their statement of "of course --shared isn't completely uselessly insecure, that's what git-shell is for!" was followed up a month later with https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2017-8386
15:25 < kanzure> 19:24 < jrayhawk> words cannot describe the stupidity
15:29 -!- rpifan [~rpifan@136.55.14.48] has joined ##hplusroadmap
15:41 < juul> interesting
15:43 < juul> git is built around the concept of one copy of a repository per user though, and also around the idea that having multiple heads is bad and should be immediately fixed
15:44 < juul> so if people are actively editing something via a web interface in a decentralized environment you could definitely make that work but git would be working against you
15:45 < juul> merkle-dags are not complicated, so unless you're forced to work with git (as with github developers) it makes a lot of sense to roll your own or use a more generalized merkle-dag that does what you need
15:45 < juul> which is what dat, scuttlebutt and ipfs have done
15:46 < kanzure> but how do you explain the shared multi user features
15:46 < kanzure> besides "mistaken design" :-)
15:47 < jrayhawk> i just spent the last hour trying to get less dysfunctionally infuriated and here you go bringing up git
15:48 < kanzure> too much stormfront?
15:48 < jrayhawk> nah, too much tech support
15:48 < kanzure> the worst.
15:57 < juul> looks like it just sets the unix permissions correctly for multiple users to commit to one repository. if we're not talking about a bare repository then that could potentially end up being problematic since writing to a file and then committing is not an atomic operation, so someone could write, then someone else could write, then the first commit happens... if it's a bare repo then you end up either
15:57 < juul> having one non-bare repo per user anyway and then locally pushing to the repo, or you end up talking to the bare repo directly using more low-level git operations. If you use low-level git operations then git really isn't doing much for you anymore, except limiting you by shoehorning your data into files and dirs and disallowing new commits every time you end up with multiple heads and forcing you to
15:57 < juul> buffer incoming commits until you can merge (which sometimes requires human intervention)
15:58 < kanzure> what's wrong with having one non-bare repo per user? that's how git works-- you commit locally, then you inspect the remote branch, then you do any necessary merge work (locally).
15:59 < kanzure> (trivially you could argue that your target audience should not be expected to be computer users.)
16:01 < juul> yeah, that's fine for git's intended use-case, but if you're implementing a piece of web-based collaborative software then you're expecting multiple users to be editing a repos simultaneously, and you can't predict which repos so you end up implementing a "clone on write" strategy which is slow, or you do something so low-level that you're not actually using the git codebase anymore (what github does)
16:01 < juul> or you force everyone to explicitly fork every time they want to change something... which might be fine for code projects but imagine if wikipedia did that
16:02 < kanzure> wikipedia should be doing that, in fact the wikipedia-git thing was the best idea they had
16:02 < kanzure> or rather, they didn't have that idea, i think that was from outside wikimedia foundation
16:03 < juul> i agree that this workflow should be an option
16:03 < juul> but i'm writing a web app for biologists. have you met biologists?
16:04 < juul> also, the core functionality of this early version of the bionet is inventory management
16:04 < juul> it would be awkward to have to fork a repo as part of the workflow of specifying a location for a new item
16:06 < fenn> what is the point of decentralized inventory management?
16:06 < fenn> does each inventory have its own database node?
16:06 < juul> yes
16:07 < fenn> so if there's one inventory then it's centralized right?
16:07 < kanzure> fenn: probably for biobrick part sharing
16:07 < juul> yeah the bionet is for sharing physical biomaterials and its associated data
16:07 < rpifan> who was it that told me to eat salmon
16:08 < fenn> me
16:08 < juul> you do a global BLAST across the federated/decentralized bionet
16:08 < kanzure> captain sleepyhead
16:08 < rpifan> i went and got that and sardine from TJs and butter lettuce
16:08 < rpifan> and feta cheese with spinach
16:08 < fenn> good job
16:09 < fenn> now go exercise instead of hanging out on IRC
16:09 < rpifan> i bike 3 hours a day for work
16:09 < rpifan> ill go do some lifting later
16:10 < jrayhawk> yeah, i have pretty given up on e.g. ikiwiki and i am pretty sure the only way to scale that model is by reimplementing large portions of the plumbing to remove the entire concept of a checkout
16:10 < juul> find what you want in someone else's lab. if it's not within walking distance then you hit "request" and the owner and/or lab manager sees the incoming request which shows where in the lab the sample is located. the person walks over, pulls a sample, puts it into an envelope and clicks "fulfill" to print a pre-paid shipping label
16:10 < juul> i'm working on automating the "walk to the sample and manually pull it out" using an automated -80 storage system
16:10 < jrayhawk> i have been meaning to poke at other DVCS wikis to see if any of them actually implemented something like that the right way
16:11 < fenn> juul this sounds great, now we just need open source enzyme-making plasmids (pre-transformed into e. coli or whatever)
16:11 < juul> the bionet could be (and probably will be) generalized for non-bio things and non-shipped things so you could imagine doing a query for "where's the closest waterjet cutter i can access"
16:13 < kanzure> fenn will you be going to biohacktheplanet with juul's free ticet
16:13 < kanzure> .. ticket.
16:13 < kanzure> you don't have to tell me but you should arrange for that if you're going.
16:14 < fenn> oh whew it's not like right now at least
16:14 < juul> fenn: yes, we have just made a deal last month with Twist for 20 megabases of synthesis. This will be used as the community directs based on the reddit.com/r/bionet posts and upvotes. It will all be made available for requests via the bionet using the Open MTA in a plasmid that has been cleared of third party rights. The synthesized sequences will be auto-cleared of third party rights to the best of
16:14 < juul> our abilities by BLASTing the sequences against the lens.org international patent database
16:14 < kanzure> will you accept my ubermensch mini chromosome as a submission
16:15 < juul> the biggest problem with making that deal with Twist (or anyone) was that none of the U.S. synthesis companies actually give you the right to redistribute the DNA that you pay them to make
16:15 < juul> this is an insane situation but Drew somehow managed to convince Twist
16:15 < fenn> you're not redistributing the DNA, you're redistributing a copy of it :P
16:15 < juul> kanzure: haha, i haven't heard ot that
16:15 < fenn> but seriously WTF kind of terms is that
16:15 < kanzure> that's just further reason to make dna synthesizers instead of dna synthesis companies
16:16 < kanzure> juul: mostly small mutations to important proteins http://diyhpl.us/wiki/genetic-modifications/
16:16 < juul> fenn: :) it's an open question how big a part of a thing is the thing itself when it comes to DNA and material transfer agreements
16:16 < fenn> is DNA a material covered by MTAs?
16:17 < juul> an MTA is a contract between the sender and receiver of biomaterials. most places that send you DNA force you to agree to either the UBMTA or something more restrictive
16:17 < kanzure> prepare for disappointment
16:17 < fenn> this all seems really weird to me
16:17 < juul> the UBMA does not allow redistribution
16:17 < kanzure> fenn: it's all just crap. it needs to go away.
16:17 < juul> and even Addgene forces you to sign the UBMTA
16:17 < fenn> i can order nuts and bolts from a catalog and there's no covenant with the manufacturer
16:17 < fenn> why is biology so different?
16:18 < kanzure> yes but can you use nuts and bolts to build a nuclear bomb???
16:18 < fenn> yes
16:18 < juul> it all stems from the fact that the bio world is ruled by patents and contracts (no copyright) and overly litigous large companies
16:18 < kanzure> yep.
16:18 < juul> we're launching the OpenMTA next month
16:18 < fenn> but your DNA is not even covered by any patents
16:18 < fenn> by definition
16:18 < kanzure> it's probalby because it's all expensive stuff
16:18 < juul> and we've gotten a few large universitites to sign on
16:18 < kanzure> if it was cheap then they wouldn't be able to enforce any of this
16:18 < juul> you never know if the DNA you make is covered by someone elses patents
16:19 < fenn> you never know anything at all
16:19 < kanzure> deep
16:19 < fenn> (my standard snark for overly-philosophical rationalist wankery)
16:20 < kanzure> juul: would drew be able to find me a chemist to work on a dna synthesizer
16:20 < kanzure> i'm willing to pay
16:21 < fenn> so if i steal DNA from someone who ordered from an evil DNA synthesis company, and give it to you, are you then legally able to distribute that DNA?
16:21 < juul> The reason the synthesis companies enforce these crazy rules is this: If they make a piece of DNA for you then they may be violating someone elses patent. They can automate the BLASTing of the sequence against the U.S. patent database, but that only works for checking against sequence-based patents, and there are patents on use as well. It's completely unrealistic for them to really know if they're
16:21 < juul> making something illegally (violating a patent) so instead they try to keep the entire transaction with their customer secret. They figure if you can't redistribute then the patent holders are unlikely to find out who originally made it in some court-ordered disclosure down the line, which would open them up to lawsuits
16:22 < juul> kanzure: i can ask him. do you have more specifics about what you need? where are they going to work etc?
16:22 < kanzure> juul: don't know where they will work. that's a good question. project overview: http://diyhpl.us/wiki/dna/synthesis/notes/ (inkjet)
16:22 < juul> fenn: patent law states that you cannot make, sell or give away something patented by someone else unless you have a license
16:22 < juul> with some exceptions for reselling something you bought from someone with a license
16:23 < fenn> let's just assume it's in fact not (yet) patented DNA
16:23 < kanzure> i think it's not just whether it's patented, it's "infringement" risk
16:23 < kanzure> and then every action has some sort of net infringement score
16:23 < fenn> infringment happens only if it's patented
16:24 < kanzure> that's up to the court to decide
16:24 < fenn> no that's wrong
16:24 < kanzure> infringement happens even if you didn't know about the other patent or didn't know that the claims were related
16:24 < juul> fenn: well i broke the law by accepting a stolen good, but let's assume the person who ordered it from the evil synthesis company actually gave it to me (in violation of their contract) then you could redistribute it legally
16:24 < fenn> the court decides whether what you did is the same as what was patented
16:24 < juul> then *i could redistribute it legally
16:25 < juul> yes patents are super vague. much vaguer than copyrights
16:25 < juul> the only good thing about patents is that they expire
16:25 < juul> like the GFP patents
16:25 < juul> they last of them expired in january
16:25 < fenn> can you copyright DNA sequences?
16:26 < kanzure> patents overrule copyright law
16:26 < juul> the general consensus of u.s. lawyers is no
16:26 < kanzure> so that wont wor
16:26 < juul> there are a few who argue that you can
16:26 < kanzure> this is all so silly. let's just directly work to invalidate their model.
16:26 < juul> the question isn't as general as that though
16:26 < juul> if i encode the harry potter books as dna then i am obviously violating copyright
16:26 < kanzure> no, only if you distribute or something
16:27 < juul> the more relevant question is: Does the DNA being constructed by most people nowadays constitute a "creative work" which would allow it to be protected by copyright law
16:27 < juul> the consensus right now is: not yet
16:27 < fenn> oops i accidentally contaminated every square mile of the planet with a complete library of human knowledge, my bad
16:27 < juul> it's too much copy-past of natural sequences and artificial selection pressures, not enough design and engineering
16:28 < juul> *copy-paste
16:29 < fenn> somehow photography is copyrightable
16:29 < fenn> i really don't see the logic in any of this
16:29 < kanzure> didn't now about the recent expiration of GFP patents
16:29 < kanzure> *know
16:29 < juul> yes, though only if humans trigger the shutter
16:29 < fenn> does pressing "enter" count?
16:29 < juul> we're working on a little side-project that shows all the latest and upcoming expired DNA patents
16:29 < fenn> i mean i can write a script to take a photo at such and such parameters, or a series of such photos
16:30 < juul> fenn: yes, even setting up a script counts
16:30 < fenn> but where does it end
16:30 < juul> but that was that photo taken by a monkey a while back
16:30 < juul> not copyrighted
16:30 < kanzure> if you invent the button, but you only cause it to be pressed by virtue of selling the camera, how is that not the same thing? you pressed enter on the assembly line.
16:30 < fenn> i write an AGI that designs a set of genes that grow into a sentient race... are they copyrighted?
16:30 < juul> so you just have to get subsentient beings to do your work for you
16:30 < juul> i guess sentient beings would work too as long as the law doesn't recognize them as such
16:32 < fenn> we need to hurry up and get mrdata on trial so we can hammer out these legal problems
16:33 < fenn> if i go to mars and give up any citizenship do i have to deal with this crap?
16:34 < juul> mars is under maritime law
16:34 < juul> so international waters work as well
16:34 < fenn> mars is covered by the outer space treaty, so it's not quite the same
16:35 < juul> does the outer space treaty talk about IP law?
16:35 < juul> i was under the impression the ownership of anything other than real estate was covered by maritime law on mars
16:35 < juul> as famously referenced in the Mars movie
16:36 < juul> (possibly the book too... can't remember)
16:36 < juul> kanzure: i will read that before my next meeting with drew
16:36 < fenn> yes yes space pirates and all that
16:36 < kanzure> juul: your question about space is a good one. i don't have an answer.
16:36 < kanzure> juul: i mean the unfortunate answer might be that maybe i need to go find someone with an academic lab who would be willing to host such a project.....
16:37 < fenn> "The agreement governing the ISS makes it clear (in Article 5) that the applicable laws, including those governing IP rights, depend on which part of it an astronaut is in." so presumably it depends which mars colony you're in
16:38 < kanzure> fenn: what about financial transactions executed on satellites in geosynchronous
16:38 < kanzure> orbit.
16:38 < juul> kanzure: that person might be drew
16:38 < juul> i'll run it by him
16:38 < kanzure> juul: welp i can help fund it if he's interested in taking that on.
16:38 < juul> ok
16:38 < kanzure> juul: also possible gp-write tie-ins (although they are still a bit disorganized?)
16:39 < fenn> kanzure it's bad enough already with stuff like btc-e owned by a russian and servers in new zealand (where the owner has never set foot)
16:39 < juul> hm possibly. there was that whole falling out between drew and george though
16:39 < kanzure> i am not aware of this
16:39 < juul> i think they made up but i don't think drew has gone to any of their events
16:39 < kanzure> what was the nature of the dispute
16:39 < juul> he got the secret invite to the original gp-write meeting, thought it was fucked up to do it in secret and immediately tweeted it out to the world
16:39 < juul> he == drew
16:40 < kanzure> and they were upset about drew bringing publicity to it?
16:40 < juul> it was supposed to be a secret meeting for VIPs only
16:40 < kanzure> juul: so i was in gp-write audience and i was typing real-time transcripts. they physically tapped me on the shoulder and told me to stop. it got into the news: http://www.vocativ.com/428192/scientists-human-genome-struggle-transparency/
16:40 < juul> the fact that it was even being discussed was secret
16:40 < juul> haha cool
16:41 < kanzure> .tw https://twitter.com/kanzure/status/861953430369259520
16:41 < yoleaux> Womp womp. @GP_write has asked me to not tweet top secret #GPwrite transcripts. Oh well. ¯\_(ツ)_/¯ (@kanzure)
16:42 < kanzure> small teams can outcompete consortiums like gp-write, so...... :-)
16:43 < fenn> you should have gotten an official OK for delayed publishing of the transcripts, since the argument was that presenters didn't want to get scooped
16:43 < fenn> personally i think it was bullshit though
16:44 < fenn> a fake excuse
16:45 < kanzure> there wasn't much to scoop except maybe the vascularized organoids stuff
16:45 < fenn> financial transactions happening in satellites would be in the jurisdiction of whatever company owns the satellite
16:46 < kanzure> well blockstream doesn't own the satellites
16:46 < fenn> i wonder if you can just "give away" a satellite, to the world
16:46 < fenn> yes but eutelsat or whatever does own it
16:47 < fenn> now you have to have end-of-life plans to de orbit the satellite, so maybe they'd say you have to destroy it if you went out of business (or sell it)
16:48 < kanzure> what happened to those satellite hijackers
16:48 < kanzure> re: old unused satellites still transmitting
16:48 < juul> the comsat satellites are still up
16:48 < juul> they just cracked down on a bunch of the south-american "hijackers"
16:49 < juul> if you can call it hijacking when you use an open relay
16:49 < juul> sorry satcom
16:49 < juul> they're supposedly super easy to use
16:49 < juul> since they're just analog relays
16:50 < juul> and geostationary? i think so
16:50 < kanzure> juul: https://blockstream.com/2017/08/15/announcing-blockstream-satellite.html
16:51 < juul> haha
16:55 < kanzure> 16:53 <+xentrac> this joint is interesting because as far as I know it's the only known example of this joinery technology from 5000 BC, when it was made, until like 1 AD
16:55  * kanzure nods
17:12 < juul> huh?
17:13 < maaku> kanzure: https://www.indiegogo.com/projects/gemini-pda-android-linux-keyboard-mobile-device-phone#/
17:13 < maaku> .title
17:13 < yoleaux> Pardon Our Interruption
17:13 < maaku> "Gemini is an ultra-thin clamshell mobile device with fully integrated tactile QWERTY keyboard, that fits in your pocket. Designed for Android, it also features a dual boot Linux option. Gemini is fully equipped with 4G, WiFi & Bluetooth enabling both data communications and mobile phone calls."
17:29 < kanzure> looks like a tiny keyboard.
17:30 < kanzure> i think my personal rule is that a keyboard must be at least as large as my feet
17:44 < rpifan> fenn, so i got the salmon and im hungry
17:44 < rpifan> what do i do with it
17:49 -!- jtimon [~quassel@173.29.134.37.dynamic.jazztel.es] has quit [Ping timeout: 255 seconds]
18:10 < fenn> put it in your mouth
18:10 < fenn> swallow
18:11 < fenn> chewing is advised
18:15 < rpifan> but do u just eat it from the tin
18:15 < fenn> i put it in a bowl with bbq sauce
18:16 < rpifan> oh
18:16 < rpifan> hmm
18:16 < rpifan> bbq sqauce is carbs
18:17 < fenn> well nothing's perfect
18:26 -!- maaku [~mark@173.234.25.100] has quit [Quit: Lost terminal]
18:28 < rpifan> true
18:29 -!- maaku [~mark@173.234.25.100] has joined ##hplusroadmap
18:29 < fenn> https://singularityhub.com/2017/06/14/forget-police-sketches-researchers-perfectly-reconstruct-faces-by-reading-brainwaves/
18:29 < fenn> .title http://www.cell.com/cell/fulltext/S0092-8674(17)30538-X
18:30 < rpifan> well then i just added some spinach to my protein shake
18:30 < rpifan> and my is it good
18:30 < yoleaux> fenn: Sorry, that doesn't appear to be an HTML page.
18:30 < fenn> https://www.scribd.com/document/350455135/The-Code-for-Facial-Identity-in-the-Primate-Brain-Cell
18:30 < rpifan> it adds a nice minty almost flavor
18:30 < rpifan> im surprised
18:30 < fenn> ugh everything is always broken
18:32 -!- maaku [~mark@173.234.25.100] has quit [Client Quit]
18:32 -!- maaku [~mark@173.234.25.100] has joined ##hplusroadmap
18:33 -!- maaku [~mark@173.234.25.100] has quit [Client Quit]
18:33 -!- maaku [~mark@173.234.25.100] has joined ##hplusroadmap
18:52 < kanzure> wow pdf is a highly resilient format.   if you decompress all the FlateDecode objects and replace them with empty values, serialize the document, and then append the entire document to the end of the file, the entire pdf file will render perfectly fine. O_O
19:01 -!- yashgaroth [~yashgarot@2606:6000:cd4d:3300:f5e0:f867:a11d:8d52] has joined ##hplusroadmap
19:11 -!- jaboja [~jaboja@jaboja.pl] has joined ##hplusroadmap
19:21 -!- mrdata- [~mrdata@unaffiliated/mrdata] has joined ##hplusroadmap
19:25 -!- Dumuzi [Dumuzi@Dont.Blink.PanicBNC.org] has quit [Ping timeout: 255 seconds]
19:29 -!- Dumuzi [Dumuzi@Dont.Blink.PanicBNC.org] has joined ##hplusroadmap
19:31 < kanzure> "macrospatial sky canopy dominion... some variations include an organic maglev shinkansen." what?
19:52 < fenn> i'll have whatever they're having
19:52 < kanzure> canned salmon
19:52 < fenn> just ate it
19:56 -!- rpifan [~rpifan@136.55.14.48] has quit [Quit: Leaving]
20:27 -!- mrdata- [~mrdata@unaffiliated/mrdata] has quit [Read error: Connection reset by peer]
20:51 -!- rpifan [~rpifan@136.55.14.48] has joined ##hplusroadmap
21:12 -!- helleshin [~talinck@cpe-174-97-113-184.cinci.res.rr.com] has joined ##hplusroadmap
21:13 -!- Guest33735 [~talinck@cpe-174-97-113-184.cinci.res.rr.com] has quit [Ping timeout: 252 seconds]
21:26 -!- jaboja [~jaboja@jaboja.pl] has quit [Remote host closed the connection]
21:28 < fltrz> so, I can describe a point in 3 dimensions in a cartesian coordinate system (x,y,z) or in say spherical coordinates (r,theta,phi). suppose I have some function of space which happens to be linear in the coordinates of one system, then they wont in the other
21:29 < fltrz> in the article http://www.cell.com/cell/fulltext/S0092-8674(17)30538-X mentioned above, they used an already existing parametrization of face space, call its coordinate system x_i, I can arbitrarily use any nonlinear bijection and use a coordinate system x'_i
21:29 < fltrz> which maps one on one to x_i coordinate system
21:30 < fltrz> they claim to have a linear space response of the upper levels face recognition cells, how accidental is that?
21:31 < fltrz> if the original authors of the face parametrization had used different coordinate system/parametrization, this response would not have been linear
21:32 < fltrz> either: 1) the face space region of human faces is small enough within the macaque face space, that the monkeys effectively use a linear tangent space in the small region of human faces
21:33 < fltrz> or: 2) the monkeys had previous exposure (or learned during the experiment) the face space representation of the face generator
21:33 < fltrz> or 3) the article is bogus
21:35 < fltrz> i.e. suppose I make a face generator 2; using the exact source code of the original face generator, but modified such that the new face parameters are not linearly but still bijectively related to the old ones
21:35 < fltrz> so I still generate faces in the same space, my parametrization is not linear with respect to the original one
21:36 < fltrz> if I repeat the experiment then, according to this paper I should get a nonlinear response of the neurons with respect to the new parametrization
21:37 < fltrz> so somehow the designers of the face generator found the macaque face representation space before it was measured?
21:38 < fltrz> if however the measured response would be measured to be linear even to the new parameters (which are nonlinearly related to the old parameters) then the macaque are simply getting familiar with the face space parametrization during the experiment
21:42 < fltrz> or 4) the face space representation of humans and macaques is similar enough that the creators of the face generator, by testing the orthogonality of the parameters on human judgements, good parameters were those that unknowingly back then resulted in linear response in our face recognition cells as well
21:50 < fltrz> considering the large range of firing rates over which linearity is reported 1) can be excluded
21:51 < fltrz> reporting linearity over such a wide range is dubious in and of itself, considering the sigmoid function inherent in neurons
21:51 < fenn> the linearity was in rate of firing
21:52 < fltrz> true
21:53 < fltrz> so I retract the dubiousity of that, but still 1) is excluded because of the wide range of firing rates
22:01 < fenn> their decoding accuracy was only like 75%
22:02 < fltrz> true, but figure 4 H is very convincing (unless falsified)
22:04 < fltrz> fenn: you agree its very coincidental that the authors of the face generator ended up using a parameter space that turns out to be linear with these neurons firing rates?
22:04 < fenn> re 4H they could have cherry picked the particular cell to make the graph from
22:05 < fenn> o
22:05 < fltrz> fenn: I'd consider that a form of falsification, but then only had 200 cells to cherry pick from
22:05 < fenn> well i dunno what "68% of variance" looks like when plotted
22:05 < fltrz> so thats still very linear... unless they also cherry pick the datapoints that weren't as linear
22:07 < fltrz> ok, so basically "the measurement was so vague we could just as well consider the response linear (without invalidating our measurements)" ?
22:08 < fenn> yes i think that is closer to the truth
22:08 < fltrz> I should make a vague measurement of the orbits of planets, and then report they might actually move in perfect circles after all
22:08 < fltrz> :)
22:08 < fenn> i'm having a hard enough time just reading this paper and understanding what they did, much less criticizing it
22:08 < fltrz> I do think option 4) is interesting though
22:09 < fenn> i don't really get where the 50 axes come from
22:09 < fenn> i think they just picked a number and did PCA on that many axes
22:11 < mrdata> speaking of mars and treaty, a man sued the ontario government for discriminating against martians, circa some time this past decade, iirc. the case was dismissed by ontario supreme court justice Epstein because there was no treaty with mars giving martians rights.
22:11 < fenn> huh
22:12 < mrdata> if it existed, such a treaty would be mutual, so earthlings would get rights under martian law as well
22:12 < fenn> what do treaties have to do with it
22:13 < fenn> you're saying illegal aliens from bumfuckistan have no rights because there's no treaty with them?
22:13 < fenn> i find it hard to believe
22:13 < mrdata> the plaintiff in this particular case was not arguing that human rights were violated
22:13 < mrdata> had that been the case, it may have had merit
22:14 < mrdata> based on universality of human rights
22:17 < mrdata> but let's say for sake of argument that the plaintiff was steadfastly insisting he is not human; rather than spend time having expert witnesses declare him to be mentally ill, the court chose this route
22:18 < fltrz> lol
22:18 < fltrz> does that set a precedent thought?
22:19 < mrdata> could be
22:19 < fenn> No person shall ... be deprived of life, liberty, or property, without due process of law ... nor shall any State deprive any person of life, liberty, or property, without due process of law
22:19 < fenn> i'm not familiar with the canadian legal system
22:19 < mrdata> it is similar
22:20 < mrdata> based on british legal system except we do have a constitution
22:20 < fltrz> are martians considered persons?
22:20 < fenn> apparently the english courts have less power and can be overruled by legislature in decisions of whether due process has been fulfilled
22:20 < fenn> the equivalent english wording says "man" not "person"
22:21 < fltrz> I guess if cows can be deprived of life without due process of law, so can martians
22:21 < fenn> "The due process clauses apply to "legal persons" (that is, corporate personhood) as well as to individuals"
22:22 < fenn> so theoretically a martian non-person could incorporate themselves as property of their corporation
22:22 < mrdata> if this is a precedent, then it is something aliens should be very aware of if they decide to visit
22:22 < fenn> it seems like a bad precedent
22:23 < fltrz> oh, but is there a precedent if the case was dismissed?
22:23 < fenn> every case sets a precedent
22:24 < mrdata> the precedent seems to be that a treaty should be negotiated between respective parties, human and alien, to confer appropriate rights mutually
22:24 < mrdata> because what if these are tasty aliens?
22:24 < mrdata> i should think they probably dont like being eaten
22:24 < mrdata> and vice versa
22:25 < mrdata> if they find us to be tasty, well. i want a guarantee of protection if i board the space craft
22:25 < fenn> Precedent Set: Case Dismissed Against “Drone Slayer” For Shooting Down Spy Drone
22:25 < fenn> seen "batteries not included"
22:26 -!- night is now known as Adifex
22:26 < fenn> "the ruling appears to be a victory for property owners" unless your property happens to be a drone
22:26 < mrdata> the drone violated privacy
22:27 < fenn> so do spy satellites, do i get to shoot those down too?
22:27 < mrdata> the space above your land is protectible, i guess, to a degree
22:27 < mrdata> altitude matters
22:27 < fenn> bah
22:27 < mrdata> air space does use the concept of altitude
22:28 < mrdata> some air space is rather controlled
22:28 < mrdata> commercial air craft fly over my place from time to time
22:29 < mrdata> they're high enough, though, that i probably couldnt hit them; and they probably would need a really expensive camera lens to get close up shots of my deck
22:30 < fltrz> what if an endangered bird violates your privacy above your property?
22:30 < mrdata> trump wants you to shoot eagles
22:31 < mrdata> but you shouldnt
22:31 < fltrz> I don't live in US
22:31 < fenn> so i don't know how to get an answer whether dismissal actually sets a legal precedent
22:31 < mrdata> good, then
22:31 < mrdata> somewhat
22:33 < mrdata> trump wants you to shoot eagles with lead bullets; why else would he hobble the EPA?
22:35 < fenn> .wik void for vagueness
22:35 < yoleaux> "In American constitutional law, a statute is void for vagueness and unenforceable if it is too vague for the average citizen to understand." — https://en.wikipedia.org/wiki/Void_for_vagueness
22:35 < fenn> this would seem to wipe out a huge swath of existing law
22:37 < fenn> a divided court found that "annoying" was not unconstitutionally vague
22:39 < mrdata> lol
22:41 < fltrz> does .chair TLD exist? arm.chair: sci-hub like DOI entry field & go button => SE or redditlike thread with comments (both positive and variation ideas and negative criticisms)
22:41 < fenn> are you pitching a lame startup
22:41 -!- Adifex is now known as night
22:42 < fltrz> like the macaque face discussion above, there could be many places and channels where a paper is discussed, but its the reviewing is not gathered together
22:42 < fltrz> fenn: yes, but I dont want to make it
22:43 < fenn> omg who do i bomb
22:43 < fenn> i don't like what happened to the internet, i'm angry! rawr!
22:45 < fltrz> nothing happened to the internet, there never was a DOI=>open review map?
22:45 < fenn> not that i know of
22:45 < fenn> i'm angrily reading the list of TLDs
22:45 < fltrz> where?
22:46 < fenn> https://en.wikipedia.org/wiki/List_of_Internet_top-level_domains
22:47 < fenn> this section is especially awful https://en.wikipedia.org/wiki/List_of_Internet_top-level_domains#Brand_top-level_domains
22:51 < fenn> so now when i enter 'foo.zip' into a super magic whammy bar it will go to a domain owned by "charleston road registry" instead of searching for 'foo.zip'?
22:51 < mrdata> yeah
22:51 < mrdata> unless you put another word in there, and a space
22:53 < mrdata> depending on your browser
22:54 < fenn> there's no consistency for plural/singular either
22:55 < fenn> foo.watch is operated by Donuts, but foo.watches is Richemont DNS
22:55 < fenn> the point of DNS was to be memorable
22:56 < fenn> foo.watch is about watches, not videos
22:57 < fenn> of course you know its main use will actually be crap like trump.watch
22:58 < fenn> huh dns can't resolve trump.sucks
23:00 < fltrz> can resolve 2.0nm either
23:00 < fenn> 0nm isn't a TLD
23:01 < fltrz> thats why it can't be resolved then I guess
23:02 -!- yashgaroth [~yashgarot@2606:6000:cd4d:3300:f5e0:f867:a11d:8d52] has quit [Quit: Leaving]
23:06 -!- poppingtonic [~brian@unaffiliated/poppingtonic] has joined ##hplusroadmap
--- Log closed Fri Aug 25 00:00:59 2017