--- Log opened Sat Dec 11 00:00:57 2021
01:19 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has joined #hplusroadmap
02:36 -!- nmz787 [~nmz787@user/nmz787] has quit [Ping timeout: 260 seconds]
06:04 < docl> .tw 1469512487846432771
06:04 < saxo> @lsparrish A stationary Earth space elevator is out, so it's (1) tidally locked tethers or (2) rotovator. In both cases you have to maintain rotation speed... how? Using propellant defeats the point. It would consume its own product plus other scarce resources (@AlanRominger, in reply to tw:1469440659253780495)
06:06 < docl> alan's response to my inquiry about what's bad about tethers
06:06 < docl> .tw 1469029553402171395
06:06 < saxo> Startup orbital rings might look like this. Two satellites in opposite orbits with magnetic coils that switch on as they get close to each other. // Along the length of the satellite there can be multiple coils to control spin and orbital inclination as well. https://pbs.twimg.com/media/FGMBtaAWYAAmgqw.png (@lsparrish)
06:06 < docl> thread he's referencing
06:17 < muurkha> so a thing I realized recently is that a swinging tether launching objects at some velocity v has exactly twice the acceleration of a simple linear accelerator mass driver, such as a gun barrel, that has a constant acceleration over the radius of the tether
06:19 < muurkha> like, if you want to launch something at 30 km/s fom a 50 km radius tether, the centripetal acceleration is twice the centripetal acceleration to launch them at 30 km/s from a 50 km long linear mass driver
06:20 < muurkha> gee forces do eventually matter for anything, it's just a matter of when.  even dumb masses can break
06:20 < muurkha> 23:27 < fenn> there's also this "two stage tether" concept which i don't really get
06:20 < muurkha> 23:27 < fenn> it's a spinning tether with a spinning tether on the end
06:20 < muurkha> 23:28 < fenn> and you can keep going with three tiers and so on
06:20 < muurkha> I used one of these when I was a kid to accidentally launch a supersonic piece of copper into my arm
06:21 < muurkha> the usual terminology is "a whip"
07:05 < docl> supersonic? ouch! did your arm survive?
07:11 < muurkha> it was a very small piece of copper, it stopped when it hit the bone
07:14 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0::f2f0] has joined #hplusroadmap
07:25 < kanzure> https://aseq.substack.com/p/roswell-revisited
07:26 < muurkha> the surgery to remove it took less than an hour, and it only pulverized a relatively small amount of tissue, so I still have the use of the arm
07:27 < kanzure> .tw https://twitter.com/cees_dekker/status/1461089656418807815
07:27 < saxo> Sad to hear that Ned Seeman passed away. Ned pioneered DNA nanotechnology across so many years and saw that field come to fruition. We’ll miss his talks with slides with corny humor that didnt change much over the decades. We’ll miss him as a person. / RIP Ned. https://twitter.com/swarupdeytweets/status/1461039224661037057 (@cees_dekker)
07:30 < muurkha> aw :(
07:31 < kanzure> evidently the protein ticker tape group and the boyden molecular ticker tape work was in unknowingly in parallel https://twitter.com/AdamEzraCohen/status/1448828657397608454
07:31 < kanzure> "Recording of cellular physiological histories along optically readable self-assembling protein chains" https://www.biorxiv.org/content/10.1101/2021.10.13.464006v1
07:33 < kanzure> "Time-tagged ticker tapes for intracellular recordings" https://www.biorxiv.org/content/10.1101/2021.10.13.463862v1
07:34 < kanzure> astrocyte computation through calcium waves https://www.biorxiv.org/content/10.1101/2021.10.20.465192v1
07:57 < fenn> docl in earth orbit you can use electrodynamic propulsion, it's just a long conductor with conductive spheres on each end, and current pushed through it. depending on how you look at it you're either accelerating the ionosphere as a fluid, or acting as a rotor in an electric rotor, with earth as the stator and earth's magnetic field as the field coils
07:58 < fenn> docl you can also use incoming mass from higher up the gravity well to regain momentum, or high Isp propulsion technologies of your preferred flavor
07:59 < fenn> all this is explained in those papers you still haven't read
08:03 -!- Llamamoe [~Llamamoe@178235178236.dynamic-4-waw-k-1-2-0.vectranet.pl] has joined #hplusroadmap
08:08 < muurkha> which papers?  pushing current through a long conductor with conductive spheres on each end sounds challenging; what discharges the spheres, the very thin top of the ionosphere?
08:30 < docl> I've heard of electrodynamic tethers
08:34 < docl> using a track like alan's suggesting lets you use them as braking only, which is less challenging
08:34 < fenn> tethers.com publications on the wayback machine several years ago
08:34 < docl> the acceleration is all localized to interaction between the east and west moving tracks
08:35 < fenn> http://web.archive.org/web/20180201231845/http://www.tethers.com/Bibliography.html
08:37 < docl> in my conception the tracks are made of many distinct units each with a net vertical shape (so they are stable), so hanging vertical conductors fits in well with that
08:38 < fenn> i wouldn't want to be anywhere near something going 16 km/s relative velocity
08:38 < fenn> someone farts and you're dead
08:41 < fenn> this is not exactly the same set of papers http://web.archive.org/web/20180216215847/http://www.tethers.com/TUPubs.html
08:41 < docl> tumbling tethers seem like a worse problem to me. they have to occupy a big chunk of orbit and when they snap they go unpredictable directions
08:42 < docl> 16 km/s is only 16 kHz per meter
08:42 -!- spaceangel [~spaceange@ip-78-102-216-202.net.upcbroadband.cz] has quit [Remote host closed the connection]
08:43 < muurkha> you mean 16 kHz meters
08:43 < muurkha> occupying space is the major drawback of tethers over linear mass drivers I think
08:43 < docl> alan did have a suggestion for longer range repulsion, use light reflected back and forth thousands of times
08:44 < docl> I like that magnets can both attract and repel though
08:45 < docl> yeah it's a linear mass driver where you only need to thrust at low gees at the interface since both are moving at close to orbital velocity
08:47 < docl> there's a different interface for launching stuff to orbit, where it uses only eddy current braking. this could be fairly distant and connected by tethers to the upper track (which I envision as the prograde / west-moving track)
08:55 < docl> .tw 1469712435112206348
08:55 < saxo> @AlanRominger Assuming we don't want mass in retrograde, the bottom track can be replaced by a braking mechanism that interacts with Earth's magnetic field. Current generated as it cuts through the field converts to thrust against the orbit. (Could also use aerobraking but ED seems better.) https://pbs.twimg.com/media/FGV3PgDWYAIU9Ot.png (@lsparrish, in reply to tw:1469337232855162885)
09:05 -!- Llamamoe [~Llamamoe@178235178236.dynamic-4-waw-k-1-2-0.vectranet.pl] has quit [Quit: Leaving.]
09:16 < juri_> yaay! https://f-droid.org/packages/com.wire/
09:16 < juri_> I did a thing! :)
09:31 < fenn> juri_: you built a package for f-droid?
09:32 < juri_> fenn: better, i got the company i work for to publish to f-droid. :)
09:32 < fenn> smart
09:32 < juri_> much more difficult, too. :)
09:32 < fenn> perhaps
09:33 < fenn> ah i missed "i work for"
09:33 < fenn> so it's basically you still
09:33 < juri_> yep. changing internal processes, business cases, etc.
09:34 < fenn> "what if those dirty homeless open source people come knocking on our door"
09:34 < fenn> "better keep it on google play only"
09:34 < juri_> "those dirty free software hippies are why you have a company. comply!" *runs*
10:34 -!- Croran [~Croran@71.231.214.173] has quit [Ping timeout: 256 seconds]
10:36 -!- Croran [~Croran@71.231.214.173] has joined #hplusroadmap
11:17 < muurkha> juri_: yay! thank you!
14:39 < muurkha> another computational fabrication researcher to watch: https://amiraa.pages.cba.mit.edu/home/
14:49 < docl> .tw 1469556838412783626
14:49 < saxo> I really like the idea behind Open Source Ecology and the Global Village Construction Set, but it’s made a couple orders of magnitude too small. We need a massive catalogue of open source components and machines. For replicating industry on Mars. (@Robotbeat)
14:58 < muurkha> I'm skeptical that Twitter is a good place to find people doing groundbreaking work
15:01 < muurkha> it's a wonderful place to attack them for it, though
15:02 < kanzure> save your attacks for when you're on twitter
15:02 < kanzure> i think it's possible to find people almost anywhere
15:05 < muurkha> I'm not attacking anybody, just saying, it's not a very friendly environment for innovation and exploration
15:09 < muurkha> I mean William Kamkwamba grew up in Malawi, but Malawi is not a very friendly environment for innovation and exploration, as you can tell by the fact that (a) out of the 19 million people in Malawi there are two people listed on https://en.wikipedia.org/wiki/Category:Malawian_engineers and the other one is a politician whose engineering consisted of building roads, (b) that politician died in South 
15:09 < muurkha> Africa, and (c) Kamkwamba lives in the US now
15:18 -!- CryptoDavid [uid14990@id-14990.uxbridge.irccloud.com] has quit [Quit: Connection closed for inactivity]
15:19 < muurkha> it would be great to find more Kamkwambas but it's a lot easier if you look for them in Berkeley instead of Malawi or Twitter
15:22 < docl> hmm. a free online website is kind of easier to access than a university... although I don't know what berkeley offers in that regard
15:41 < muurkha> it offers a place where you can talk with other people who are doing interesting, innovative work and get feedback on your ideas without a flashmob forming to try to get you fired because you said "blacklist"
15:42 < docl> this chat is one of the best places I've found so far for people interested in what I consider the most important questions
15:42 < muurkha> yes, I'm a fan too
15:43 < muurkha> I am glad this chat is not Twitter
15:47 -!- helleshin [~talinck@108-225-123-172.lightspeed.cntmoh.sbcglobal.net] has joined #hplusroadmap
15:48 < muurkha> actually it looks like Kamkwamba has gone back to Malawi to try to fix precisely the problem I mentioned above:
15:48 < muurkha> > Right now what I’m working on is to put up an innovation center in Malawi. The innovation center is going to allow people to come and work with me to develop their ideas. Because I know there are so many talented young people all over the world but sometimes lack of space to help them build their ideas, it’s lacking sometimes so I want to enable that, making sure that anyone who has an idea 
15:48 < muurkha> they can have a space where they can think and build and be connected to professionals in the field that they are working on. I can be able to connect young people who have an idea with professionals in that field to help them to go through everything they have to know, what they have to do, in order to achieve their goals.
15:48 < muurkha> oh, only half time
15:50 < docl> ok now I gotta follow that guy on twitter... @wkamkwamba
15:51 < muurkha> haha
15:52 < docl> btw there seems to be a big famine ongoing/upcoming. they're saying it's caused by climate change, less cold air to condense the rain out :(
15:52 < muurkha> he hasn't tweeted since 02017, when he retweeted one tweet.  in 02016 he retweeted 12 tweets
15:53 < docl> here's his innovation center link https://movingwindmills.org/
15:53 < docl> .t
15:53 < saxo> Moving Windmills - Inspiring Innovation - Moving Windmills Project
15:53 < muurkha> yeah
15:54 < muurkha> I haven't been there but I imagine it dramatically improves the prospects for the next Malawian William Kamkwamba
15:54 < juri_> I'm still writing my slicer. :)
15:56 < muurkha> How's that going?
15:56 < juri_> it's slow work, but i'm confident in the result.
15:58 < muurkha> I find it hard to have confidence in Slic3r because I used to work at a Perl company
15:58 < juri_> I'm quite off the beaten path mathematically, but i've covered some ground that doesn't have pre-existing work. i think i get to write a math paper in the field of straight skeletons.
15:58 < juri_> yeah, i'm all haskell.. and using quickcheck to test my code. it's popped out so many errors.. and also convinced me i'm on the right path.
15:59 < muurkha> yeah, pbt tends to find a lot of bugs
15:59 < docl> hslice right?
15:59 < juri_> yeah, i'd like to see some other slicers trying it. :)
15:59 < juri_> yepyep.
16:00 < muurkha> and geometry code is full of edge cases, and I don't mean the modern degraded sense of the term "cases I'm too lazy to cover" but rather boundary cases
16:00 < juri_> I'm currently on a 3 month fork to severely change the engine.
16:00 < muurkha> why did you end up using straight skeletons?
16:00 < muurkha> I didn't know what they were
16:01 < juri_> because that's how it's done?
16:02 < juri_> I ended up using some new papers to design my engine.. and the "cases i'm to lazy to cover" in the paper turn out to be "how to write a meta-algorithm to allow you to test your problem so you know which one of the other algorithms or these new algorithms will solve your problem fastest."
16:02 < juri_> now i'm implementing that.
16:02 < muurkha> haha, that's awesome!
16:03 < juri_> it'll be awesome when i get it working. i'm just about 2 years in. :P
16:04 < juri_> with a good quickcheck framework, at least i know i am making progress.
16:05 < muurkha> I have this tendency to think of Haskell as having a great deal of Perlish whipupitude, despite being strongly statically typed
16:05 < muurkha> but I suppose you're more bottlenecked on manipulexity than whipupitude
16:06 < juri_> it lets me scafold bad solutions and solidify them well. thank god for that. "maybe this works?" has a series of straight forward transforms to actual algorithm.
16:07 < muurkha> I guess I'm holding it wrong
16:08 < muurkha> my Haskell is kind of kindergarten level
16:08 < juri_> apply stan, and hlint religiously. stan will keep you out of "head: empty list", and hlint will show you where to simplify
16:09 < juri_> I have mostly switched away from lists, to using slist because of stan. now if it only had a NonEmpty version.
16:10 < muurkha> how much parallelism are you able to wring out of the problem?
16:11 < juri_> All of it. :D
16:11 < juri_> I program on a machine with 56 threads. when i measure compute time consumed, it easily accounts for 48 threads, at full boar.
16:15 < muurkha> nice!
16:15 < muurkha> I've been thinking it would be important to integrate FEM modeling and in particular topopt into slicers
16:16 < muurkha> do you know of anyone who's doing that?
16:17 < juri_> no, i'm going in a different direction. instead of a bajillion switches and knobs, i'm integrating a SCAD engine, and exposing slicing primitives to that. so you can write code for "how to slice".
16:17 < juri_> maybe next year. :)
16:18 < muurkha> that sounds interesting!
16:18 < muurkha> you've developed an algebra of slicing?
16:18 < juri_> we've always had one. it's just not been very good. :)
16:19 < juri_> or exposed to the users.
16:19 < juri_> I tend to write my code as one big library, which lends itsself to using a scripting engine to operate it. i had a scripting engine in implicitcad. i'm just re-using it.
16:24 < muurkha> yeah, that makes sense
16:25 < muurkha> what's the central kernel of the algebra?  hmm maybe the hslice documentation will tell me
16:25 < juri_> I'm using projective geometric algebra, rather than linear algebra. things get complicated. :)
16:26 < muurkha> what, no code of conduct?  problematic!
16:26 < muurkha> ;)
16:27 < juri_> you need a community before you have community problems. right now, i have an incomplete slicer, and a lot of math. :)
16:29 < muurkha> yeah, agreed!
16:31 < muurkha> I think it goes further: you need a project structure that isn't just "I'm Julia and this is my project, you can fork it and/or send me code if you want"
16:33 < juri_> If you want to poke at the code, i suggest you look at https://github.com/Haskell-Things/HSlice/tree/handle_more_cells , it's just about to be merged, and is a lot more sane, or insane, depending on how you look at the whole thing. :)
16:33 < muurkha> because that form of community can scale up by a few orders of magnitude
16:33 < muurkha> I will look!  but I did say my Haskell was kind of kindergarten-level
16:34 < juri_> I still comment like a C programmer. hopefully, it's useful. :)
16:39 < docl> I'm also looking, also kindergarten level at haskell
16:44 < juri_> the place to look might just be at the test suite then. check out tests/Math/PGA.hs
16:44 < muurkha> aha, thanks
16:44 < muurkha> heh, Ganja.hs
16:45 < juri_> I'm not responsible for the naming. :)
16:46 < juri_> there's a website for visualizing the output of that. i use that file when i want to see what geometry looks llike, rather than.. yeah. :)
16:46 < muurkha> did you want to ask the original authors what they were smoking when they wrote that code?
16:47 < juri_> no, i was thankful they wrote something, so i could build on it. i'm strongly anti-new-project.
16:47 < juri_> ... I'm the one that converted it to PGA.
16:47 < muurkha> I just meant Ganja.hs!  I didn't intend that as an actual criticism
16:48 < juri_> hah! :)
16:49 < muurkha> I am not in a position to be able to offer actual criticism
16:50 < muurkha> so if anything I said above sounded like a criticism, that wasn't my intent
16:50 < juri_> no offense taken. i'm used to it. i write slicers in haskell. :P
16:50 < juri_> and this is what i do to *relax*.
16:51 < muurkha> sure, what else would someone do to relax?
16:51 < muurkha> I mean I guess you could watch a movie or something, but where's the fun in that?  ;)
16:54 < muurkha> hmm, I'm finding https://en.wikipedia.org/wiki/Geometric_algebra difficult to understand because I don't know what quadratic forms and Clifford algebras are.  is there a textbook you can recommend?
16:55 < juri_> there's a talk that was given at siggraph that's good. other than that, there's some good papers.
16:56 < juri_> haven't read any textbooks on the subject. it's too new, and what i have found is too theoretical.
16:58 < muurkha> am I looking at the wrong geometric algebra?
16:58 < muurkha> I'm pathologically pro-new-project myself
17:02 < juri_> https://bivector.net/2DPGA.pdf https://arxiv.org/pdf/1101.4542.pdf and https://arxiv.org/pdf/1411.6502.pdf are most of what i work from.
17:03 < muurkha> thanks!
17:03 < juri_> charles gunn also did a talk at siggraph. go watch it.
17:05 < juri_> if you read my PGA code, you'll find some more funny operators. i also spotted another 'crack in the math' that lets you take context into your operation choices, letting you avoid half of the math in some cases.
17:07 < muurkha> this is looking really interesting
17:08 < muurkha> I like algebraic approaches to things because they fit my brain
17:08 < muurkha> looks like the Wikipedia page is about the right geometric algebra
17:08 < muurkha> but I feel like I need to know what Grassmann algebras and Clifford algebras are first
17:09 < juri_> yeah, i'm just working in PGA 2,0,1*.
17:10 < juri_> clifford algebras are the set of possible algebras. for PGA 2,0,1*, it means i've got a dual, i've got two positive vectors, and one zero vector.
17:10 < juri_> I'll probably graduate to higher algebras when i start laying down curved surfaces.
17:18 < muurkha> thanks!
17:28 < lsneff> gonna be giving a talk about rust in aerospace and what my student rocketry team has been up to in February at the flight software workshop
17:39 < muurkha> cool!  what have you learned?
17:40 < kanzure> shamir secret sharing on paper https://github.com/roconnor-blockstream/SSS32
17:40 < muurkha> that's a good idea
17:42 < muurkha> I wonder if you can use it without knowing the Greek alphabet :)
18:44 -!- yashgaroth [~ffffffff@2601:5c4:c780:6aa0::f2f0] has quit [Quit: Leaving]
19:03 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 256 seconds]
21:30 -!- andytoshi [~apoelstra@user/andytoshi] has quit [Remote host closed the connection]
21:30 -!- soundandfury [~soundandf@user/soundandfury] has quit [Remote host closed the connection]
22:27 -!- spaceangel [~spaceange@ip-78-102-216-202.net.upcbroadband.cz] has joined #hplusroadmap
22:56 < maaku> lsneff: will it be recorded?
--- Log closed Sun Dec 12 00:00:58 2021