asciilifeform: !w poll
watchglass: Polling 10 nodes...
watchglass: 192.151.158.26:8333 : Could not connect!
watchglass: 205.134.172.6:8333 : Alive: (0.073s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=620173
watchglass: 208.94.240.42:8333 : Alive: (0.097s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=620173
watchglass: 108.31.170.3:8333 : Alive: (0.096s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=620173 (Operator: asciilifeform)
watchglass: 205.134.172.27:8333 : Alive: (0.083s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=620173 (Operator: asciilifeform)
watchglass: 205.134.172.4:8333 : Alive: (0.143s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=620173
watchglass: 143.202.160.10:8333 : Alive: (0.228s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=620173
watchglass: 213.109.238.156:8333 : Alive: (0.343s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=620173
watchglass: 103.36.92.112:8333 : Alive: (0.506s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=620173
watchglass: 188.121.168.69:8333 : Alive: (0.356s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=620173
asciilifeform: !w probe 192.151.158.26
watchglass: 192.151.158.26:8333 : Could not connect!
asciilifeform: BingoBoingo ^ iirc yours ?
BingoBoingo: ty
asciilifeform: BingoBoingo: would be interesting to hear what brought it down
asciilifeform: ( 'wedge' dun normally give 100% stone-cold 'connection refused' )
BingoBoingo: tail of debug.log http://paste.deedbot.org/?id=auS2
asciilifeform: BingoBoingo: what type of node was this ?
asciilifeform: BingoBoingo: and this loox like an orderly shutdown, rather than crash..?
asciilifeform: (i.e. exactly what you'd get if e.g. kill (not -9) , or if shutdown executed by root )
asciilifeform: 'received block 0000000000000000000e' << evidently dun have the hash untruncation patch, not to mention unwedger?
asciilifeform: it wasn't wedged tho. someone gave it a clean shutdown
jurov: asciilifeform: FYI, I just brutally symlinked both /dev/random and /dev/urandom to /dev/ttyUSB0 , ran strace to verify it really reads fuckgoats juice,...
jurov: ..and it seem to generate rsa keys faster!
jurov: both openssl and ssh-keygen
asciilifeform: jurov: stock /dev/random is abominably slow , typically, so not surprise. the tricky part (as described in detail by bvt) is to make this behave sanely during a) boot b) any other condition where the device is unavailable
asciilifeform: ( as well as c) make the FG do useful work even when not being actively polled )
asciilifeform: linking urandom (if you have it on the machine at all) to a 7kB/s rng is imho painful (when do you use urandom? typically it's when you need coupla MB of liquishit , ~quickly~, and quality dun matter much )
jurov: in fact, both openssl and ssh-keygen read from urandom
asciilifeform: jurov: keep in mind also that if you dun initialize the uart correctly, FG will still appear to work, but output will be in no sense entropic
asciilifeform: afaik w/out bvt's kernelisms, there is no way to make it happen a) on boot b) if uart is unplugged/replugged immediately
jurov: i see
asciilifeform: jurov: see 'flow control' section in the manual
asciilifeform: it's a serial device, and sends rng bytes. which requires a 8-bit-clean tty, which linux does ~not~ by default give.
asciilifeform at one time considered to make it shit uuencoded binary, but decided against -- waste half the bw, and would still be susceptible to catastrophic misconfig
asciilifeform: the pc arch simply offers no 100% satisfying place to plug in rng. imho i picked the least retarded one (100% of the gear in my machine room, whether from 1980s or present day, offers serial)
asciilifeform: the reason why not built usbism into fg itself, is that usb inherently demands 2-way comm. while FG is emphatically a 1-way device.
asciilifeform: this was much more important than e.g. ability to use w/ 486 lappy that lacks usb.
asciilifeform: !w poll
watchglass: Polling 10 nodes...
watchglass: 192.151.158.26:8333 : Could not connect!
watchglass: 205.134.172.4:8333 : Alive: (0.083s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=620223
watchglass: 205.134.172.6:8333 : Alive: (0.088s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=620223
watchglass: 205.134.172.27:8333 : Alive: (0.084s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=620223 (Operator: asciilifeform)
watchglass: 108.31.170.3:8333 : Alive: (0.161s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=620223 (Operator: asciilifeform)
watchglass: 143.202.160.10:8333 : Alive: (0.157s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=620223
watchglass: 208.94.240.42:8333 : Alive: (0.207s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=620223
watchglass: 213.109.238.156:8333 : Alive: (0.403s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=620223
watchglass: 103.36.92.112:8333 : Alive: (0.527s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=620223
watchglass: 188.121.168.69:8333 : Alive: (0.346s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=620223
jfw: http://logs.nosuchlabs.com/log/asciilifeform/2020-03-03#1008322 - precisely an off-the-shelf optoisolator. Perhaps if I get to baking FGs, can do some of those too...
snsabot: Logged on 2020-03-03 14:10:24 asciilifeform: jfw (supposing you're tuned in) -- can't resist to ask, what is meant by data diode in this context ? is off-the-shelf optoisolator, or sumthing else ?
asciilifeform: jfw: at one time i tried to buy commercial 'send rs232 over fiber 1way', i.e. with separate power supplies ; did not find on market, had to make (pictured in linked thrd.)
asciilifeform: was wondering whether this is what jfw had been selling.
jfw: asciilifeform: hm, which part of the linked pictures was what you made? the fiber optics, USB converters?
asciilifeform: jfw: the pictured set was receiver & transmitter , on separate breadboards
asciilifeform: jfw: was for experiment where looked for possible diff. in quality of output of FG in rf-shielded enclosure, w/ battery power inside, and 1 fiber for transmission, vs. ordinary . (did not find detectable effect.)
asciilifeform: sewn from off-the-shelf components, as i think is obv., 2 visible-light led/detector pairs (manufacturer samples) , 1 plastic fiber ('corning' iirc), 1 usb-uart (cp2102 iirc), 1 FG, 1 battery holder, 2 standard breadboards, a qty of wire.
jfw: gotcha. Yep that's about what we're doing, just USB converter on both ends.
asciilifeform: makes sense, ty
jfw: And since I was poked about it: http://logs.nosuchlabs.com/log/asciilifeform/2020-02-05#1006048 - doesn't seem to describe what MP has in mind, though I could see it as the first guess from initial description.
snsabot: Logged on 2020-02-05 17:08:54 asciilifeform: shinohai: i sincerely hope that mp et al actually try to bake this, it'll be hilarious to watch ( where signing e.g. linux kernel will need coupla 10,000s of rsa invocations, auto-recognition of 'what lang' so to get the comment & demarcation syntax, and coupld 10MB of entropy.. )
asciilifeform: jfw: unless i misread, idea was to put a sig in ea. file.
jfw: there's clearly an intent to support source files directly, but also http://billymg.com/2020/01/mp-wp-patch-viewer-and-code-shelf/comment-page-1/#comment-83
asciilifeform: this means, to sign project, N rsa invocations, where N files. and to press tree, N * n_s where n_s is # of signers.
asciilifeform: jfw: if (as in linked comment) signing only patches, could do w/out the 'mandate same comment syntax in all langs' tho. in fact could do right nao, simply prepend the detached sigs to top of patch. (why to do this, remains to me a mystery, but could w/out rewiring much of anyffin)
shinohai: Just shut up and do what MP says, it is not for vassals to know mysteries of the king.
asciilifeform: shinohai: i invited jfw specifically to comment re subj, have some patience wontcha.
shinohai: ^.^
jfw: lolz. asciilifeform: I hear ya, am uncertain why exactly he wants this and can't say I'm hugely motivated to run and do it at present, but I've started with seeking clarification on what's actually proposed.
asciilifeform: jfw: i was curious re whether it made sense to you (and whether aware of the hilarity of 'weeks+ press' hidden in the 1st proposed variant, will admit.)
asciilifeform: rsa only seems 'instant' when happens 1x, when manually pressed key in e.g. gpg. but if you want coupla 10k rsaings, suddenly becomes obv that no, not 'instant'.
jfw: right.
asciilifeform: ( fwiw mine 2.3x faster than koch's, when long public exponent. but still ultimately too slow for many uses that'd be practical if were 'instant'.
asciilifeform: anyways ty jfw for comment. i'ma bbl.
jfw: yw asciilifeform.