--- Log opened Mon Jun 16 00:00:13 2025
00:13 < alethkit> can you get kleptoplasty in mammals?
00:21 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has joined #hplusroadmap
01:22 < hprmbridge> kanzure> "... Making it more confusing, they had a different number system for each type of thing they were counting. So they had one system for counting people, another for counting sheep, another for counting bushels of barley, another for counting years, etc. The concept of “numbers” had not been generalized as an abstract concept. Numbers existed only as “numbers of sheep” or “numbers of bushels”.  It
01:22 < hprmbridge> kanzure> is analogous to the English system of liquid measures, where the number of teaspoons, tablespoons, cups, gallons, etc., is a different count in each higher measure." https://x.com/DrPhiltill/status/1934132660886270453
02:49 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has quit [Ping timeout: 248 seconds]
04:54 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
05:03 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has quit [Ping timeout: 260 seconds]
05:09 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
06:38 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has left #hplusroadmap []
06:46 -!- L29Ah [~L29Ah@wikipedia/L29Ah] has joined #hplusroadmap
06:51 -!- justanot1 [~justanoth@gateway/tor-sasl/justanotheruser] has quit [Ping timeout: 244 seconds]
08:00 -!- TMM [hp@amanda.tmm.cx] has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
08:00 -!- TMM [hp@amanda.tmm.cx] has joined #hplusroadmap
09:55 -!- Malvolio [~Malvolio@idlerpg/player/Malvolio] has quit [Ping timeout: 248 seconds]
10:23 -!- justanotheruser [~justanoth@gateway/tor-sasl/justanotheruser] has joined #hplusroadmap
11:42 < kanzure> https://radicle.xyz/ does radicle store comments and code review in the same branch as the code? earlier ".bug/" projects in git put the issue tracker in the same branch as the code.
11:42 < kanzure> using a separate branch seems like a better idea
11:52 < kanzure> hm radicle.xyz might be more effective if it ran in the browser and the default landing page was itself a running local copy of radicle for the radicle project.
11:53 < RangerMauve> Can't do p2p in the browser so it'd kinda defeat the point IMO
11:54 < kanzure> RangerMauve: see https://docs.libp2p.io/guides/getting-started/webrtc/
11:55 < RangerMauve> That's not "true" p2p and is very bad for performance and relies on a bunch of centralized always online services
11:55 < kanzure> what, TURN stuff?
11:55 < RangerMauve> signaling too
11:56 < RangerMauve> Also lacks local network. also browsers aren't good at storing local user data without wiping it randomly or imposing arbitrary limits
11:56 < RangerMauve> Also webrtc coonnection counts are limited. Just establishing them is slow AF
11:57 < kanzure> local user data can be stored and backed up by file->save mechanics
11:57 < RangerMauve> I spent many years optimizing p2p stuff in browsers and it's severely limited ux / perf / resilience wise vs native
11:57 < kanzure> i want tor relay nodes in the browser. i know the performance will be awful but i want it anyway.
11:58 < RangerMauve> Make it :P
11:58 < kanzure> i think peerinfinity already did, but i could be wrong
11:58 < RangerMauve> Also this: https://snowflake.torproject.org/
12:00 < RangerMauve> A big part of it too is that it's expensive to detour into browsers while the core functions aren't sustainable yet
12:12 < RangerMauve> Looking at their protocol docs I'm not sure if they even do hole punching right now.
12:18 < RangerMauve> Everything is using ip+port so getting WebRTC into their transports would be a major PITA https://github.com/radicle-dev/heartwood/blob/master/crates/radicle/src/node.rs#L546
12:20 < RangerMauve> Actually no I'm full of shit. Their core is built on top of cyphernet so it could be possible to use some sort of proxies or alternative addresses. https://docs.rs/cyphernet/latest/cyphernet/ Right now they use HostName though
12:25 -!- stipa_ [~stipa@user/stipa] has joined #hplusroadmap
12:25 -!- stipa [~stipa@user/stipa] has quit [Read error: Connection reset by peer]
12:25 -!- stipa_ is now known as stipa
15:39 -!- stipa_ [~stipa@user/stipa] has joined #hplusroadmap
15:41 -!- stipa [~stipa@user/stipa] has quit [Ping timeout: 252 seconds]
15:41 -!- stipa_ is now known as stipa
15:56 < fenn> why couldn't a p2p node also run a web server? this isn't a hard problem guys
15:58 < RangerMauve> The hardest part there is getting incoming connections. Bypassing NAT is easiest using UDP but most web stuff is TCP
15:59 < fenn> oh, bleh
15:59 < RangerMauve> The state of the art is using QUIC HTTP3 and using UDP hole punching over your existing UDP socket for the DHT
16:00 < RangerMauve> IIRC IPFS folks are migrating to nodes using HTTP to connect to each other for the exact reason you're thinking
16:00 < RangerMauve> Sadly most p2p protocols need to optimize the wire protocol a lot to make replication viable when you're making loads of connections to dozens of peers
16:09 < fenn> perhaps the rest of the solar system will start out with ipv6 even if earth is fucked
16:21 < fenn> all the files on app.radicle.xyz say "File not found." i'm pretty sure that's not supposed to happen
16:24 -!- darsie [~darsie@84-113-82-174.cable.dynamic.surfer.at] has quit [Ping timeout: 252 seconds]
16:29 < fenn> oh i see, you want to seed the radicle network just by viewing the webpage
16:29 < fenn> that seems kinda like a virus
16:30 < fenn> people should have to opt in to participating in a p2p network
16:30 < fenn> if i were designing a browser that should not be possible
16:34 < hprmbridge> kanzure> no, you'd have to press a button to run 'git push' or 'git host'
16:45 -!- Malvolio [~Malvolio@idlerpg/player/Malvolio] has joined #hplusroadmap
17:16 < fenn> .t https://www.sciencedaily.com/releases/2014/11/141113152916.htm
17:16 < saxo> The science behind total recall: New player in brain function and memory | ScienceDaily
17:16 < fenn> "FXR1P (Fragile X Related Protein 1), was responsible for suppressing the production of molecules required for building new memories"
17:17 < fenn> dangerous stuff. what happens when the CIA and friends are able to reliably block memory formation
17:21 < fenn> .t https://www.cell.com/cell-reports/fulltext/S2211-1247(14)00882-1
17:21 < saxo> Just a moment...
17:21 < fenn> "FXR1P Limits Long-Term Memory, Long-Lasting Synaptic Potentiation, and De Novo GluA2 Translation"
17:42 <+gnusha_> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=38398b12 fennwiki: add "Increased Translation of GluA2 in FXR1P cKO Mice" to mouse tables >> http://diyhpl.us/diyhpluswiki/genetic-modifications/
19:54 < fenn> $80 portable specturm analyzer https://www.tinysa.org/wiki/
19:54 < fenn> spectrum*
19:57 < fenn> would be fun to walk around with two of them strapped to your head and stereo audio
20:26 < fenn> oh maybe it's not realtime
20:37 < fenn> "The fastest scanning speed occurs with an RBW or 300kHz or wider and is about 2 scans per second. But with increase of the frequency span and decrease of the RBW the scanning speed decreases quickly. A scan from 0MHz to 350MHz with an RBW of 10kHz takes about 2 minutes."
20:38 < fenn> (resolution bandwidth, the size of each frequency bin)
20:38 < fenn> 2 fps is not great for AR audio perception
20:45 < superkuh> rtlsdr with rtl_power does about 80 MHz/s frequency hopping at 2.4 MS/s. hackrf does about 8 GHz/s (@ 20MS/s). Typical frontend for this use case is qspectrumanalyzer.
20:51 < fenn> superkuh: i'm having trouble conceptualizing this since i know very little about radio concepts. how do i convert these numbers into scans per second a some resolution bandwidth?
20:52 < fenn> at*
20:53 < fenn> i would want ~10 scans per second for good immersion
20:54 < fenn> and can you dither the binning frequency bands to better trade off resolution vs speed?
20:59 < superkuh> Since they both do complex sampling the instantaneous bandwidth is equal to the megasamples per second.
20:59 < superkuh> The resolution of the FFT is arbitrary.
21:00 < fenn> i would like to smile and nod but i actually want to know the answer so please bear with me
21:01 < fenn> so we have some 40MHz window we can move around the spectrum right? and it can be expanded to cover more of the spectrum but at lower resolution?
21:01 < superkuh> Take the frequency hopping rate and divide it by the frequency range.
21:02 < superkuh> For a hackrf if you want to cover 2 GHz then at 8 GHz/s you cover that range 4 times a second.
21:03 < superkuh> The frequency hopping rate is the instantaneous bandwidth times the number of rehops/sec, but I'm just calling it frequency hopping rate here.
21:04 < superkuh> For the hackrf hopping at 8 GHz/s covering 40 MHz of frequency range, (that's just 2 tunings) you'd get updates about 200 times/s.
21:04 < superkuh> With the hackrf_sweep backend and qspectrumanalyzer.
21:05 < superkuh> For an rtlsdr it'd be about 2 times per second.
21:05 < fenn> ah i should not have said 40MHz that was just an error
21:06 < superkuh> The instantaneous bandwidth "window" you can move around (retune) on the hackrf is 20 MHz (20MS/s complex sampled).
21:06 < superkuh> For rtlsdrs it's generally about 2.4 MHz max (though 2.5 is usable).
21:07 < fenn> that is really tiny if we want to cover the whole spectrum
21:07 < fenn> is there a way to "zoom out"?
21:08 < superkuh> No. The instantaneous bandwidth is a function of the ADC and how many times it samples the voltage per second. More bandwidth means an ADC that samples faster which gets expensive. High end oscilloscopes interleave multiple coherently clocked ADCs.
21:10 < fenn> like 1000 of them?
21:10 < superkuh> Like 4.
21:12 < superkuh> Commercial spectrum monitoring equipment can (signalhound, etc) can cover ~single THz/s with ~low hundreds MS/s.
21:30 < hprmbridge> nmz787> Fenn are you interested in walking around and listening to the world with your ears, and watching the frequency distribution plots on screen (or AR)? Or are you trying to RF scan your brain, and then later correlate the audio recording with the spectral recording?
21:30 < hprmbridge> nmz787> Or something else?
21:30 < hprmbridge> nmz787> What's the input to the spectrum analyzer?
21:35 < fenn> i want to map the spectrum analyzer output to an audio spectrum and listen to that while walking around turning my head to point at various radio sources that may be in the environment that most humans are blissfully ignorant of
21:35 < fenn> for funsies
21:35 < hprmbridge> nmz787> So directional antennas... Big bunny ears
21:36 < fenn> so say a 100MHz signal will be a bass tone and a 5 GHz will be a high pitched tinnitus noise
21:36 < superkuh> Going down to 100 MHz directional will be rather large.
21:36 < fenn> yes bunny ears will probably be necessary for the long wavelength signals but if it's impractical then we could do something else with that like run the antenna along the legs
21:37 < superkuh> A tapered slot/vivaldi hat is probably the best solution. Or muti-arm achimedian spiral on the front or back of shirt.
21:38 < fenn> i'd like to compare the relative phase to determine directionality but unclear how to do that since the relative phase changes with frequency and at low frequencies you'd need crazy precision
21:38 < fenn> so instead the cheap shitty solution is signal strength and directional antennas
21:39 < superkuh> You could make a little vector antenna have receivers for all 6 components of the electromagnetic wave (and your multiple combined antennas for each) and compute the poynting vector.
21:39 < superkuh> There are also polarimetry tricks but I think they'd be not so useful with body worn antenna.
21:41 < superkuh> Hard to do that broadband though.
21:41 < superkuh> And you'd have multiple strong signals making things confusing.
22:42 < fenn> nmz787: what's the thingy called where you spin coat UV resist onto a blank copper clad PCB and then use a blu-ray laser to expose it and get a {mask pattern for chemical etching, solder mask, silk screen}
22:44 < fenn> i'm imagining this as a 3d printer attachment
23:05 < fenn> Laser Direct Imaging
23:05 < fenn> apparently the UV resin temperature matters a lot, fortunately 3d printers have bed heaters
23:34 < hprmbridge> nmz787> yeah that's just lithography to me
23:34 < hprmbridge> nmz787> but direct-write, sure
23:34 < hprmbridge> nmz787> that term also applies to DMD/DLP/LCD projectors too
23:35 < hprmbridge> nmz787> (DLP may be an antiquated term, sorry)
23:36 < fenn> i think you need a different resin with those unless you modify the projector
23:36 < fenn> 405nm laser pointers are everywhere
23:37 < fenn> everyone seems to use laser printed transparency sheets as a mask
23:37 < hprmbridge> nmz787> I ripped the bulb out of a projector once and tricked it into not shutting off
23:53 < hprmbridge> nmz787> I have a little laser etcher that I made a PCB that could handle SSOP pitch, as I recall
23:55 < hprmbridge> nmz787> https://www.ebay.com/itm/315665166179
23:55 < hprmbridge> nmz787> (or that lineage, from probably 6-8 years ago) it was basically made from two CD-ROM drives, in terms of the gantry
23:56 < hprmbridge> nmz787> the part that moves the laser across the radius of the discs
--- Log closed Tue Jun 17 00:00:14 2025