--- Log opened Fri Feb 21 00:00:59 2020 00:45 -!- saxo [~saxo@2001:19f0:6800:1102:5400:ff:fe11:39a1] has joined ##hplusroadmap 00:51 < archels_> juri_: I tried a bunch of CSG engines many years ago to make finite element models for neurons. Basically I was just unioning convex elements together (cones and spheres). Resulting meshes varied from broken to very broken. I believe there's no novelty left doing this for neurons (people have published alternative methods by now), but was wondering if things are still this bad? 01:00 -!- spaceangel [~spaceange@ip-89-177-23-46.net.upcbroadband.cz] has joined ##hplusroadmap 03:13 -!- hehelleshin [~talinck@98.29.27.253] has joined ##hplusroadmap 03:16 -!- CRM114 is now known as Urchin 03:18 -!- helleshin [~talinck@98.29.27.253] has quit [Ping timeout: 258 seconds] 03:26 -!- darsie [~kvirc@84-114-73-160.cable.dynamic.surfer.at] has joined ##hplusroadmap 03:31 -!- Human_G33k [~HumanG33k@62.147.242.8] has joined ##hplusroadmap 03:33 < juri_> archels_: my meshes are much better than they used to be. 03:35 -!- HumanGeek [~HumanG33k@62.147.242.8] has quit [Ping timeout: 260 seconds] 03:57 < nmz787> juri_: did you ever try my microfluidic model again? that one that was causing issues? 04:13 < juri_> honestly, there are too many errors in it, and i don't know the subject well enough to fix them. 04:14 < juri_> sadly, i have a talk next week, so can't really do implicitcadish things ATM. gotta pretend i know stuff. 04:21 < juri_> the good news is the engine now tells you about the errors, instead of eating all memory, and dieing. 04:22 < juri_> so, that's progreess. :) 04:34 -!- filipepe [uid362247@gateway/web/irccloud.com/x-hyhunjijjgopxkdb] has joined ##hplusroadmap 06:28 -!- CryptoDavid [uid14990@gateway/web/irccloud.com/x-hzsaepfhryugcklp] has joined ##hplusroadmap 06:43 -!- filipepe [uid362247@gateway/web/irccloud.com/x-hyhunjijjgopxkdb] has quit [Quit: Connection closed for inactivity] 07:26 -!- spaceangel [~spaceange@ip-89-177-23-46.net.upcbroadband.cz] has quit [Remote host closed the connection] 07:26 -!- filipepe [uid362247@gateway/web/irccloud.com/x-qkslwmcfshjmpxgw] has joined ##hplusroadmap 08:04 < kanzure> aaaaaaaaaaaaaaa 08:05 < juri_> https://bobkonf.de/2020/longtin.html 08:05 < juri_> aaaaaaaaaaaaaa 08:28 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=f2933d03 Michael Folkson: Add Rusty transcript from Boltathon >> http://diyhpl.us/diyhpluswiki/transcripts/boltathon/2019-04-06-rusty-russell-json-interface/ 08:28 -!- ##hplusroadmap Cannot send to nick/channel 08:28 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=ab7ca1e0 Bryan Bishop: Merge pull request #75 from michaelfolkson/add-rusty-boltathon >> http://diyhpl.us/diyhpluswiki/ 08:28 -!- ##hplusroadmap Cannot send to nick/channel 08:32 < lsneff> juri_: Does implicitcad run marching cubes or dual contouring or something on the sdf to render? 08:36 < juri_> marching cubes. 08:39 -!- shawwwn [uid6132@gateway/web/irccloud.com/x-xcsrteqicfvdcxio] has joined ##hplusroadmap 08:43 < lsneff> It's too bad there needs to be that intermediary for 3d printing. It'd be ideal to just slice the sdf directly, simpler too than slicing polygons. 08:44 < juri_> actually, i can probably do that.. just not yet. 08:44 < juri_> I've written a slicer. it sucks like a gravity well, but it's a thing. 09:05 -!- berndj [~berndj@azna.co.za] has quit [Read error: Connection reset by peer] 09:06 -!- berndj [~berndj@azna.co.za] has joined ##hplusroadmap 09:12 < kanzure> fenn: we should make a definitive document about CSG vs nurbs vs whatever else, and just get a final opinion and be done with it. 09:12 < kanzure> fenn: unfortunately i don't remember why i hated openscad; i remember the tessellation had to be re-done if you wanted to change the resolution of the output which seemed really stupid. 09:25 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=8a67732e Bryan Bishop: transcript: formal verification using constraint solving >> http://diyhpl.us/diyhpluswiki/transcripts/stanford-blockchain-conference/2020/smart-contract-bug-hunting/ 09:25 -!- ##hplusroadmap Cannot send to nick/channel 09:29 < kanzure> oh it might have been the custom language 09:30 < kanzure> if you have high-resolution constraints and very particular surfaces with particular curves, then it seems wrong to represent them as a collection of fine-grained triangles especially if the triangles don't actually correspond to the desired curvature 09:30 < kanzure> it's triangle worship 09:30 < kanzure> https://github.com/Coldcard/ckcc-protocol/blob/master/PROTOCOL.md#link-level-encryption 09:31 < lsneff> Doesn't openscad use cgal? 09:36 -!- berndj [~berndj@azna.co.za] has quit [Ping timeout: 260 seconds] 09:40 < lsneff> Ah, it looks like cgal uses triangle faces to represent shapes. 09:40 < lsneff> I think so, at least. 09:41 < juri_> from what i know, that's true. ImplicitCAD doesn't have to.. I'm still extracting the potential from it's engine. 09:49 < lsneff> kanzure: what's your opinion on being able to manipulate objects from the model render vs. only being able to do it progmatically? 09:50 -!- berndj [~berndj@197.242.93.82] has joined ##hplusroadmap 09:51 < kanzure> i like a graphical interface for quick sketching, concept design, ideation, tweaking, and testing, but personally i prefer to have a script that can generate the object and i can version control the script instead of version controlling a proprietary object file 09:53 < lsneff> So, automatic generation of the script? 09:55 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=25a92962 Bryan Bishop: CAD representation file >> http://diyhpl.us/diyhpluswiki/transcripts/stanford-blockchain-conference/2020/smart-contract-bug-hunting/ 09:55 -!- ##hplusroadmap Cannot send to nick/channel 09:55 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=177e5078 Bryan Bishop: more transcript >> http://diyhpl.us/diyhpluswiki/transcripts/stanford-blockchain-conference/2020/motoko-language/ 09:55 -!- ##hplusroadmap Cannot send to nick/channel 09:55 < kanzure> nah, as a user i've come to accept that if i have written a script and then switch into interactive design mode that while i should expect a parameteric CSG tree to be represented in memory, i should not necessarily expect that data structure to be efficiently translatable back into code (e.g. what about all my comments or other functions) 09:56 < kanzure> http://diyhpl.us/wiki/cad/representation/ 09:57 < kanzure> at most i would expect to be able to load an interactively-manipulated parameteric design back into code with a load("screw-drive-rail.cad") and maybe introspect the CSG tree, but i wouldn't expect (or maybe even i wouldn't want) automatic code generation 09:59 -!- jenelizabeth [~jenelizab@cpc155793-brmb11-2-0-cust474.1-3.cable.virginm.net] has quit [Remote host closed the connection] 10:04 < kanzure> "Bitcoin crypto-bounties for quantum capable adversaries" https://eprint.iacr.org/2020/186.pdf 10:12 -!- MarkOtaris [mark-otari@wikimedia/Mark-Otaris] has quit [Ping timeout: 248 seconds] 10:12 -!- kun0[m] [kun0matrix@gateway/shell/matrix.org/x-uecfhyvvswkilbum] has quit [Ping timeout: 245 seconds] 10:14 < kanzure> https://dominiccummings.com/2019/02/21/on-the-referendum-29-genetics-genomics-predictions-the-gretzky-game-a-chance-for-britain-to-help-the-world/ 10:23 -!- kun0[m] [kun0matrix@gateway/shell/matrix.org/x-brjclfvebooacnft] has joined ##hplusroadmap 10:30 -!- kun0[m] [kun0matrix@gateway/shell/matrix.org/x-brjclfvebooacnft] has quit [Quit: killed] 10:32 < lsneff> The thing that worries me about implicit/F-rep models is that there's very little ability for the software to actually reason about the shapes it's rendering. 10:35 < lsneff> You can get an arbitrarily precise approximation, but the actual information about what's a circle, where the edges are, which corners are sharp, etc is lost. 10:58 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=92d80391 Bryan Bishop: transcript: atomic multi-channel updates >> http://diyhpl.us/diyhpluswiki/transcripts/stanford-blockchain-conference/2020/atomic-multi-channel-updates/ 10:58 -!- ##hplusroadmap Cannot send to nick/channel 11:00 -!- o08oo [uid422206@gateway/web/irccloud.com/x-hkylqzueuyfrcjhw] has quit [Quit: Connection closed for inactivity] 11:03 -!- kun0[m] [kun0matrix@gateway/shell/matrix.org/x-rhxehmqvpbdcorsz] has joined ##hplusroadmap 11:04 -!- fox2p [~fox2p@cpe-66-108-32-173.nyc.res.rr.com] has quit [Ping timeout: 265 seconds] 11:05 -!- fox2p [~fox2p@cpe-66-108-32-173.nyc.res.rr.com] has joined ##hplusroadmap 11:24 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=460c82a5 Bryan Bishop: transcript: boomerang >> http://diyhpl.us/diyhpluswiki/transcripts/stanford-blockchain-conference/2020/boomerang/ 11:24 -!- ##hplusroadmap Cannot send to nick/channel 11:31 -!- MarkOtaris [mark-otari@wikimedia/Mark-Otaris] has joined ##hplusroadmap 11:32 < juri_> lsneff: that's why you render to the precision of your production infrastructure. for instance, i use the layer height of my print as the rendering resolution.. because i can't produce dimensions smaller than that. 11:33 -!- filipepe [uid362247@gateway/web/irccloud.com/x-qkslwmcfshjmpxgw] has quit [Quit: Connection closed for inactivity] 11:49 -!- shawwwn [uid6132@gateway/web/irccloud.com/x-xcsrteqicfvdcxio] has quit [Quit: Connection closed for inactivity] 11:52 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=6cfb1611 Bryan Bishop: transcript: brick >> http://diyhpl.us/diyhpluswiki/transcripts/stanford-blockchain-conference/2020/brick-async-state-channels/ 11:52 -!- ##hplusroadmap Cannot send to nick/channel 11:55 -!- shawwwn [uid6132@gateway/web/irccloud.com/x-rmcgutgaveafqgqn] has joined ##hplusroadmap 12:06 < lsneff> You might be able to control an extruder movement directly from the sdf without slicing it. 12:08 < lsneff> or rendering it 12:21 -!- Human_G33k [~HumanG33k@62.147.242.8] has quit [Remote host closed the connection] 12:22 -!- Human_G33k [~HumanG33k@62.147.242.8] has joined ##hplusroadmap 12:29 < maaku> Self-replicating nanosystems are probably very difficult to do in environments that aren't hard vacuums. <-- what about biology? 12:31 < maaku> also, if you want to talk about large cell sizes: https://en.wikipedia.org/wiki/Unicellular_organism 12:32 < lsneff> maaku: I didn't say they're impossible :P. I bet we could eventually do it with mechanical nanotech, but it'd be much more complex than reproducing in a vacuum. 12:35 < maaku> if you're talking about drexlarian diamandoid synthesis, yea just do it in a vacuum. but the vacuum chamber could be mm-sized 12:35 < maaku> for self-replicating systems, but I don't see the relevance of that outside of grey goo sci-fi stories 12:46 < lsneff> Yeah, I agree. Don't remember the context I made that comment. 12:46 -!- CryptoDavid [uid14990@gateway/web/irccloud.com/x-hzsaepfhryugcklp] has quit [Quit: Connection closed for inactivity] 13:32 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has quit [Ping timeout: 245 seconds] 13:42 -!- N-time [~Mark@212.225.172.60] has joined ##hplusroadmap 13:56 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=934c40c2 Bryan Bishop: transcript: optimistic VM >> http://diyhpl.us/diyhpluswiki/transcripts/stanford-blockchain-conference/2020/optimistic-vm/ 13:56 -!- ##hplusroadmap Cannot send to nick/channel 14:01 -!- CRM114 [~urchin@unaffiliated/urchin] has joined ##hplusroadmap 14:03 -!- Urchin [~urchin@unaffiliated/urchin] has quit [Ping timeout: 258 seconds] 14:16 -!- CRM114 is now known as Urchin 14:36 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=8efc0a50 Bryan Bishop: transcript: coded merkle trees >> http://diyhpl.us/diyhpluswiki/transcripts/stanford-blockchain-conference/2020/solving-data-availability-attacks-using-coded-merkle-trees/ 14:36 -!- ##hplusroadmap Cannot send to nick/channel 14:36 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has joined ##hplusroadmap 15:07 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=43476f2e fenn: subdivision surfaces (i.e. blender) >> http://diyhpl.us/diyhpluswiki/cad/representation/ 15:07 -!- ##hplusroadmap Cannot send to nick/channel 15:11 < fenn> i'm putting in a good word for subdivision surfaces. despite requiring a lot of manual labor to craft, they're simple, easily understandable, robust, and easy to repair if somehow it gets messed up. also most game engines etc already support the workflow. if you need more precision to match a particular airfoil cross section or whatever, you just add more control points 15:13 < fenn> rendering/3D printing/etc would never show polygons because you can automatically determine the required level of subdivision to use 15:58 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=a94abae6 Bryan Bishop: transcript: ripple >> http://diyhpl.us/diyhpluswiki/transcripts/stanford-blockchain-conference/2020/no-incentive/ 15:58 -!- ##hplusroadmap Cannot send to nick/channel 15:58 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=c0eab5a9 Bryan Bishop: transcript: frontrunning >> http://diyhpl.us/diyhpluswiki/transcripts/stanford-blockchain-conference/2020/transparent-dishonesty/ 15:58 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=518a0457 Bryan Bishop: Merge remote-tracking branch 'origin/master' into 'master' >> http://diyhpl.us/diyhpluswiki/ 15:58 -!- ##hplusroadmap Cannot send to nick/channel 15:58 -!- ##hplusroadmap Cannot send to nick/channel 16:09 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has quit [Ping timeout: 248 seconds] 16:10 -!- turona [~quassel@2a01:c22:ac00:ef00:24c8:3493:8cdc:7e62] has quit [Ping timeout: 272 seconds] 16:11 -!- turona [~quassel@2a01:c22:b027:6700:2107:e7b6:3815:a446] has joined ##hplusroadmap 16:25 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=38fdb01f Bryan Bishop: transcript: clockwork for non-frontrunning >> http://diyhpl.us/diyhpluswiki/transcripts/stanford-blockchain-conference/2020/clockwork-nonfrontrunning/ 16:25 -!- ##hplusroadmap Cannot send to nick/channel 16:27 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has joined ##hplusroadmap 16:40 -!- yashgaroth [~ffffffff@172.58.30.153] has joined ##hplusroadmap 17:02 -!- filipepe [uid362247@gateway/web/irccloud.com/x-ftklnuvfvhfsazwn] has joined ##hplusroadmap 17:12 -!- superkuh_ [~superkuh@c-24-118-172-137.hsd1.wi.comcast.net] has joined ##hplusroadmap 17:12 -!- superkuh_ [~superkuh@c-24-118-172-137.hsd1.wi.comcast.net] has quit [Changing host] 17:12 -!- superkuh_ [~superkuh@unaffiliated/superkuh] has joined ##hplusroadmap 17:12 -!- superkuh [~superkuh@unaffiliated/superkuh] has quit [Read error: Connection reset by peer] 17:18 -!- superkuh_ is now known as superkuh 17:28 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has quit [Quit: leaving] 17:32 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has quit [Read error: Connection reset by peer] 17:33 < lsneff> https://i.redd.it/rzmsnenlg8i41.png 17:36 < lsneff> ^ grain of rice vs. smallest microcontroller 17:43 -!- fltrz [~fltrz@109.236.129.249] has joined ##hplusroadmap 17:44 < fltrz> kanzure, long time ago again, I know, ... but could UV-C inactivated virus particles in theory be inhaled as a form of vaccine? 17:46 < fltrz> instead of masks blocking 999 / 1000 particles, just inactivate so there is an N-log reduction: no hard breathing through a filter, and immune system gets used to recognizing and clearing this garbage, so that after 10^N particles at least one might be immune instead of get infected after all with a classic blocking filter 17:47 < fltrz> another possible advantage: I believe recognizing a threat is only one facet of an immune / emergency response, the context could be another parameter 17:48 < fltrz> by way of analogy: consider the good inoculation of children with fire drills: if there is fire, calmly evacuate the school building, and leave all your shit (your coat, your bag, yes even homework and your lunch bag...) 17:49 < fltrz> the same threat / signal (fire alarm), would later trigger a similar response in say a factory setting, just calmly vacate the building; this generalizes well from the school context to the factory building context 17:50 < fltrz> but sometimes, recognizing the same threat with accompanying drilled reflex / response is counterproductive: 17:50 -!- N-time [~Mark@212.225.172.60] has quit [Quit: Leaving] 17:50 < fltrz> fire at sea: if one calmly vacates the ship and now swims safely at a distance of the fire... then you lose the ship and drown eventually 17:50 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has joined ##hplusroadmap 17:52 < fltrz> as far as I am aware vaccines are nearly universally introduced in the blood stream. and apparently they tend to generalize well to other points of entry of the live virus 17:53 < fltrz> but what if virus types for which designing vaccines are harder, are not necessarily because of the recognition problem, but because the bloodstream fire drill doesn't generalize well to the situation of a virus being detected on a mucous membrane, or the lungs? 17:54 < fltrz> sometimes leading to an over reaction of the immune system (which may be appropriate in the bloodstream, but might not be so in a different tissue type?) 17:55 < fltrz> these ideas came to me because I am trying to design a portable UV-C fluence dose light reactor, so you essentially breathe air from a long reflective chamber, which disinfects the air before you breathe it 17:57 < fltrz> also, looking up more about this, I suddenly had an idea for reagentless (but still with catalyst) DNA memory tape writing scheme 17:57 < fltrz> if you're interested I will write it out... 17:59 < darsie> lsneff: Which uC is this? 18:02 < lsneff> darsie: https://news.umich.edu/u-m-researchers-create-worlds-smallest-computer/ 18:28 < fenn> inactivated virus particles won't trigger an immune response without an adjuvant 18:30 < fenn> it would just look like random dirt to the immune system 18:31 < fenn> my (basic) understanding is that coronavirus is spread by contact 18:32 < fenn> the masks are to prevent patients from emitting droplets, and to prevent medical workers from touching their unprotected face 18:37 < fltrz> fenn, I thought all foreign particles ("dead or alive") were cleared by the immune system? 18:39 -!- shawwwn [uid6132@gateway/web/irccloud.com/x-rmcgutgaveafqgqn] has quit [Quit: Connection closed for inactivity] 18:39 < fenn> if you're trying to cause immunity you need to stimulate B-cell division (of the exact B-cell that produces antibodies against the virus) 18:39 < fenn> dirt and stuff is just dumb macrophage activity 18:40 < fltrz> fenn, (I am sure you know way more about this than me, I just had to look up adjuvant) wikipedia states among other things: "Adjuvants can act as a depot for the antigen, presenting the antigen over a longer period of time, thus maximizing the immune response before the body clears the antigen." 18:40 < fenn> it's like training a pit bull to attack a tire on a rope by beating the dog 18:40 < fenn> the dog won't attack the tire unless you beat it 18:41 < fltrz> but if you wear the UV-C inactivator for a prolonged period of time during an outbreak, you will be constantly breathing inactivated virus particles though, especially if someone works in the hospital say 18:41 < fenn> (and then hopefully the dog will attack the desired target later on when it counts) 18:41 -!- justanotheruser [~justanoth@unaffiliated/justanotheruser] has quit [Read error: Connection reset by peer] 18:43 < fenn> the wikipedia summary is not very clear 18:43 < fenn> adjuvants are nasty horrible chemicals that cause localized damage 18:43 < fenn> it's the damage signals emitted by local cells that cause the immune response 18:45 < fenn> also dead bacteria and interleukins 18:45 < fltrz> fenn, I was also recently wondering, I think the stressed cells must use at least 2 (or more) types of signal to provide feedback to the immune system: 18:46 < fltrz> a cell could be/end up in 3 states: 1) stressed 2) relieved 3) dead 18:47 < fenn> 4) malfunctioning 18:48 < fltrz> if we consider these as probabilities for a cell to reside in, then a single scalar number (or concentration) is insufficient, say stress signal, because a dead cell is no longer releasing stress signals 18:48 < fltrz> which could be interpreted by immune system as "good job" 18:48 < fltrz> so we need at least 2 signals (or 3 including your 4th cell state): stress signal, and relief signal 18:49 < fltrz> immune system can then assess not only what the concentration of stressed cells is, but also concentration of recovered cells, and implicitly of dead cells 18:50 < fltrz> (this is all quite analogous to the widespread armchair reasoning about infected, dead, recovered statistics) 18:52 < fltrz> fenn, do corona viruses produce toxins? 18:54 < fltrz> how do we know that UV-C inactivated virus particles don't produce a stress response? they might still achieve entry into cells but just have damaged RNA, damaged proteins, and some free radicals 18:54 < fenn> yes that's plausible 18:55 -!- darsie [~kvirc@84-114-73-160.cable.dynamic.surfer.at] has quit [Ping timeout: 260 seconds] 18:55 < fenn> you might end up selecting for UV-resistant RNA sequences 18:57 < fltrz> yes, if the UV dose is too low I can imagine this happening eventually, just like with finishing an antibiotic regimen prematurely 18:57 < fltrz> perhaps it would generate melanin or something... 18:57 < fenn> .wik pyrimidine dimer 18:57 < saxo> "Pyrimidine dimers are molecular lesions formed from thymine or cytosine bases in DNA via photochemical reactions. Ultraviolet light (UV) induces the formation of covalent linkages between consecutive bases along the nucleotide chain in the vicinity of their carbon–carbon [...]" - https://en.wikipedia.org/wiki/Pyrimidine_dimer 18:58 < fenn> this is the molecular mechanism of DNA damage with the lowest energy 18:58 < fltrz> fenn, yes, I was incidentally wondering what was known about the UV-C damage to protein and RNA? 18:59 < fltrz> because we know that RNA viruses also get inactivated by UV-C radiation 19:01 < fenn> it happens in RNA as well 19:03 < fenn> there's a zillion possible reactions that could be catalyzed by UV 19:03 < fltrz> fenn, also, the pyrimidine dimer is what I was hoping could be used for DNA data storage, a spot of UV-C light could be focused on a stretch of DNA to put DNA into the pyrimidine dimer state ("resetting a sector to 0") and then use photolyase enzymes (caged in a transparent localized cage around the DNA) to repair the dimers at specific locations to set individual "bit" regions 19:03 < fltrz> the photolyase enzuymes are activated by visible light (violet, blue) 19:04 < fenn> nice 19:04 < fltrz> so reset a long sector with UV-C (all dimers), then set individual bits by highly localized enzymes 19:04 < fenn> presumably there's a nanopore involved somewhere 19:05 < fenn> to feed the DNA through at a constant rate while writing bits 19:05 < fltrz> yes, the photolyase enzyme cage has 2 nanopores so DNA can pass through, reading is open ended, eather another nanopore, and seeking along the DNA is faradayic conduction of the long electropositive DNA just like in nanopores 19:06 < fltrz> I was hoping that since the photolyase enzymes "recognize dimers", that it's optical spectrum (in the green or red region) would still be markedly different depending on dimer or not 19:07 < fenn> probably not 19:08 < fltrz> :( probably not 19:08 < fltrz> perhaps raman spectroscopy of same enzymes? 19:08 < fenn> just copy the DNA with dimer-induced errors and sequence it 19:09 < fltrz> perhaps conventional nanopores can sense the difference between dimer or not? 19:09 < fltrz> does DNA amplification even reproduce dimers? 19:11 < fenn> sorry, the internet has become useless for me over the past year or so 19:11 < fenn> can't answer any question i don't already know the answer to 19:11 < fltrz> the reason I was keeping this in mind for DNA tape memory, was because it seemed rather reagent less, UV-C to erase the state to dimers, and violet/blue + photolyase to set to undimer 19:12 < fltrz> fenn, in what sense has the internet become useless? the internet itself having become walled garden fest, or you mean because you changed (old age, disabilities)? 19:12 < fenn> probably search engines changing their heuristics 19:13 < fltrz> yes, search has become a nightmare too for me 19:13 < fenn> also more and more sites with cloudflare and BS webservers that don't respond 19:13 < fltrz> and anal retentive javascript requirements 19:14 < fenn> death by a thousand cuts 19:15 < fltrz> the reason I don't think DNA copying will preserve the UV-C dimerized state,... it's a bistable state, I am not sure if it will even succeed in copying, ... I think the complementary strand will sequence fine, but possibly the strand with the dimer would fail to copy? or perhaps result in 2 disconnected strands of DNA? 19:19 < fltrz> anyway, the UV-C disinfector would use LED's, because trying to make it with mercury lamps... would cause a person to inhale ozone for prolonged periods of time... which causes... pneumonia! 19:19 < fenn> and your UV-C LED wouldn't generate ozone? 19:20 < fltrz> no, the UV-C leds are 260-275nm... ozone is sub 200nm 19:21 < fltrz> theres a mercury line there 19:21 < fenn> "the molecule falls apart to O + O2 after absorbing a photon. The most important absorption is the Hartley band, extending from slightly above 300 nm down to slightly above 200 nm." 19:21 < fltrz> glass can block some, but over time your still breathing significant amounts of ozone 19:21 < fltrz> fenn, thanks for pointing that out 19:21 < fltrz> fenn, where did you find this? 19:21 < fenn> https://en.wikipedia.org/wiki/Ozone 19:22 < fenn> i had previously researched UV sterilization of mold 19:22 < fltrz> fenn, thats not the wavelength for generating ozone, but light it emits as it dissociates back to oxygen 19:22 < fltrz> which is necessarily of a longer wavelength that for generating ozone 19:23 < fenn> oh 19:24 < fenn> "Ozone in the stratosphere is mostly produced from short-wave ultraviolet rays between 240 and 160 nm." 19:24 < fltrz> fenn, but I still do want to find out the exact ozone production rates as a function of wavelength & intensity in air 19:24 < fltrz> all I find are vague quotes saying below 200nm 19:25 < fltrz> but not sure how hard that limit is, or if it is still producing ozone at say 240nm but just at a much lower rate... 19:25 < fenn> there would still be two-photon absorption 19:26 < fenn> the low wavelength UV would break down ozone 19:26 < fltrz> "Oxygen starts to absorb weakly at 240 nm in the Herzberg bands, but most of the oxygen is dissociated by absorption in the strong Schumann–Runge bands between 200 and 160 nm where ozone does not absorb. " 19:27 < fltrz> so really, I just should find the optical spectrum for oxygen 19:28 < fltrz> I wonder if 265nm is far enough from 240nm to prevent formation of ozone? most UV-C led manufacturers tout the ozone-freeness 19:33 < fltrz> fenn, did you see the videos of people in china "recycling" used masks, from dumpsters at hospitals? reselling them? 19:35 < fltrz> I really hope that was fake, but official media channels relating the story of one city/hospital intercepting a shipment of masks intended for another city/hospital located more centrally in the epidemic center is not exactly hope inspiring 19:36 < fltrz> lot of 1) panic 2) social proximity could be avoided if people could recharge a reusable mask 19:37 < fltrz> instead of regularly gathering at the local pharmacist 19:38 < fenn> put the mask in the sun for a day? 19:40 < fltrz> sometimes its not sunny, and there is little UV-C at sea level 19:41 -!- filipepe [uid362247@gateway/web/irccloud.com/x-ftklnuvfvhfsazwn] has quit [Quit: Connection closed for inactivity] 19:43 < fenn> bake it in an oven? 19:43 < fltrz> sooner or later filters get clogged 19:44 < fenn> the masks i've seen all have giant gaps around the sides :\ 19:44 < fltrz> yes, I also think there has been very little innovation on durability and ergonomics for masks 19:44 < fltrz> theres specialized full face masks intended to be reused after disinfection 19:45 < fltrz> but the mass surgical masks,... I consider it mostly like ink jet printer cartridges 19:45 < fenn> unfortunately full face masks cost 100x the crappy ones 19:45 < fltrz> theres no incentive to make a durable and ergonomic device 19:45 < fltrz> and even the expensive ones don't see much innovation on the masks 19:46 < fltrz> I was thinking small semi-inflated "bicycle tires" along the rim, so that it conforms to your specific face shape... 19:46 < fenn> i'd rather have a huge filter with a battery pack and air blower mounted on the hip or back of the head, and a hose going to a silicone rubber mask 19:46 < fenn> anyway it seems that the masks just stop you from touching your face 19:47 < fltrz> fenn, thats roughly how I see my device: minus air blower 19:47 < fenn> currently talking about cyclonic dust filters on another channel so may be getting my signals crossed 19:47 < fltrz> fenn, what is the role of the air blower? 19:47 < fenn> so you don't have to suck air through it with your lungs 19:47 < fenn> also supposedly higher velocity works better for hepa filters 19:48 < fenn> the particle impinges on the filter strand instead of going around it 19:48 < fltrz> fenn, thats also a reason why I prefer to use UV-C leds, you don't need to inhale through a resistive filter 19:49 < fltrz> so with UV-C leds its like breathing through a wide pipe (I would use valves so air goes in one direction systematically) 19:49 < fltrz> so theres almost no resistance to breathing 19:49 < fltrz> and I don't need to power a motor, just the LED(s) 19:49 < fltrz> so lighter battery requirements 19:51 < fltrz> (keep in mind if you try my UV-C led idea, that you need light baffles to prevent UV-C from reaching the user, as it is cataractogenic and carcinogenic) 19:52 < fltrz> between the light baffles on input and output, you want a long but reflective device, since even aluminum will absorb around 10% per reflection 19:53 < fltrz> you want the rays to be mostly parallel to the pipe 19:54 < fltrz> but if it does reach the wals of the device, you want to reflect as much as possible, since every reflection increases the illumination power ... 19:57 < fenn> which godzilla monster has laser breath 19:58 < fltrz> :) 19:59 < fltrz> yes, I was also thinking about UV-C lasers, which would be much easier to design the reflective cavity: use the device cavity with 2 mirrors as the external cavity for a UV-C lasing medium 19:59 < fltrz> but thats probably not within budget for this outbreak 20:00 < fltrz> the UV-C leds can be ~ $7 for 40mW at bulk amounts 20:01 < fenn> progress 20:01 < fenn> last i checked they were >$100 each 20:01 < fltrz> yes, was also seeing like $20 dollars minimum each (at low quantities) and that for jut 3mW 20:02 < fltrz> until I sent a mail to one of the more promising manufacturers and I now have their pricing 20:03 < fltrz> I even suspect their 40,50,60 mW to actually be the same UV-C led, but I suspect each with a different "protection" zener added-on, for price differentiation 20:03 < fltrz> I have to get rid of the zener anyway, because it wastes energy and consumer thermal budget 20:04 < fltrz> *consumes 20:04 < fltrz> then I will characterize the LED myself to verify if the LED on 40mW is actually less efficient than on 60mW or not 20:04 < fltrz> and compare the zener diode breakdown voltages... 20:06 < fltrz> I don't really mind price differentiation, but if this is what they're doing, its 1) at the cost of energy efficiency, 2) at the cost of losing ability to sense the LED diode junction temperature for lifetime 20:06 < fltrz> fenn, if you were searching before, do you want to know which ones I'm considering? 20:08 < fltrz> its Klaran, from CIS 20:27 < fenn> thanks, but i don't even remember what project it was for 20:35 < fltrz> perhaps it was for " i had previously researched UV sterilization of mold" ? 20:37 < fltrz> (sounds like that should work, with some precaution for preventing human exposure, if its on a wall, perhaps a Wifi controlled device, or with a timer, so you can exit the room and close the door) 20:37 < fltrz> theres also a handy table of UV fluences for different pathogens 20:38 < fltrz> "Fluence (UV Dose) Required to Achieve Incremental Log Inactivation of Bacteria, Protozoa, Viruses and Algae" 20:39 < fenn> nah i was using CFLs for the mold 20:39 < fltrz> https://iuvanews.com/stories/pdf/archives/180301_UVSensitivityReview_full.pdf 20:43 -!- aeiousomething [~aeiousome@unaffiliated/aeiousomething] has joined ##hplusroadmap 20:51 -!- yashgaroth [~ffffffff@172.58.30.153] has quit [Quit: Leaving] 21:17 < nmz787> juri_: where to get the latest implicitcad code? --- Log closed Sat Feb 22 00:00:00 2020