IRC channel logs

2022-10-12.log

back to list of logs

<stikonas>oriansj: hmm, I'm thinking that maybe libc*.M1 and headers in https://github.com/oriansj/M2libc/tree/main/amd64 should go to linux subdirectory...
<stikonas>but probably not the defs file
<stikonas>hmm, I think libc-core.M1 would be quite big for UEFI
<stikonas>we have to restrict ourselves to cc_amd64 subset of C and it's tricky to call UEFI functions from it (e.g. no support for &)
<stikonas>so possibly I'll just steal the whole startup sequence from cc_amd64.m1
<stikonas>anyway, I don't expect M2.efi to be ready before the ned of this month...
<oriansj>stikonas: probably as you'll be creating a UEFI folder to mirror the linux folder
<oriansj>(and we have to figure out how we want M2-Mesoplanet to chose which folder to collect files from)
<oriansj>and minus the stack setup, everything else can be done in C
<oriansj>civodul: please do share the error messages and/or logs so that we can help if you need/want it.
<oriansj>ilmu[m]: Hail is for greeting, not thanking but I love the spirit ^_^
<oriansj>pabs3: well provided there is low bandwidth communication, global software development could continue. It only takes a few hundred HAMs around the world to provide such a network over packet switch radio.
<oriansj>assuming developers can adapt to minimal network communication methods such as IRC and usenet
<muurkha>why a few hundred?
<muurkha>I mean, why not two?
<oriansj>muurkha: well the question is the size of the network region around them that they can provide bandwith for
<muurkha>I guess it depends on where the developers are
<oriansj>^bandwith^bandwith^
<muurkha>heh
<oriansj>^bandwith^bandwidth^
<oriansj>and what network is available to power between the developers and the network relays
<oriansj>for example there may be cables between towns and cities that continue to function, so even whole countries may remain connected after collapse
<oriansj>but in a worst case scenerio; the relays may only be able to service a couple square miles (if less the number grows into the thousands)
<muurkha>I was crossing the pedestrian bridge over the Sarmiento Line Train the other day, and I realized it would be possible to drop a magnetically anchored radio transponder onto the top of the train from the bridge
<muurkha>that train goes back and forth between Station 11 in downtown and Moreno Station about 18 hours a day, making about 12 trips
<muurkha>those are about 150 km apart
<oriansj>assuming the train keeps running
<muurkha>right
<muurkha>if the transponder has a range of 500 m, that covers about 150 square km, with a round trip latency of up to an hour and a half
<muurkha>typically a handheld GMRS radio has a kilometer or three of range for voice traffic; they use UHF
<muurkha>but that's mostly limited by surface topography; if you can get line of sight to a repeater you can get tens of km range out of them
<muurkha>they cost about US$6: https://articulo.mercadolibre.com.ar/MLA-617285621-handy-walkie-alcance-5km-real-handies-manos-libres-par-x2-_JM
<oriansj>and there are things like: https://github.com/Lora-net/LoRaMac-node which enable cheap nodes
<muurkha>oh neat, I haven't tried out my LoRa radios yet
<muurkha>ultimately if you can integrate a signal over ten times as long a time period, you can receive the same number of bits at a ten times lower power level
<muurkha>some of my notes on this from 9 years ago are in https://dercuano.github.io/notes/ultraslow-radio.html
<muurkha>there's a lot of military research on signals with low probability of detection, not just interference robustness, which might be important if you're living in an environment where telecommunication is considered subversive
<oriansj>muurkha: good stuff
<muurkha>without a common time source, clock drift limits feasible integration times, but as long as you can passively receive GPS or GLONASS, you can use those to synchronize your clocks, getting arbitrarily long integration times
<muurkha>so you could plausibly communicate at one bit per week
<oriansj>muurkha: definitely not enough bandwidth for downloading the Linux kernel in one's lifetime