Show Idle (>14 d.) Chans


← 2020-07-28 | 2020-07-30 →
feedbot: http://verisimilitudes.net/2020-07-28 << A Syndication of Verisimilitudes -- An Implementation of SHA256 in Ada
zzz999: Logged on 2020-07-21 21:08:19 asciilifeform: fixed an unclosed 'table' tag in logotron; will cut an errata patch at some pt.
lobbes: Yeah I'm still logging this chan and others. Fwiw I'm sometimes consider re-cutting auctionbot once more to sit on the alfbot rather than the ircbot, but then again, this constant recoding of the thing is getting ridiculous
lobbes: But, over the last few months the alfbot has proven that it can stay connected with ~0 intervention from myself, which I like!
asciilifeform: ohey wb lobbes !
asciilifeform: lobbes: i haven't published any changes to the bot per se since sept. '19. updated www end strictly. presently satisfied that when it discons, is from fleanode & isp weather, rather than any logic defect.
asciilifeform: lobbes: on one occasion, however, it was unable to autoreconnect (somehow 'socket in use', even tho theoretically socket gets gc'd on connect...) but was unable to reproduce. so possibly still needs some sharp edge sanded off.
asciilifeform: !w poll
watchglass: Polling 12 nodes...
watchglass: 205.134.172.4:8333 : (172-4.core.ai.net) Alive: (0.069s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=641350
watchglass: 205.134.172.6:8333 : (172-6.core.ai.net) Alive: (0.082s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=641350
watchglass: 205.134.172.27:8333 : Alive: (0.084s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=641350 (Operator: asciilifeform)
watchglass: 205.134.172.26:8333 : Alive: (0.151s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=641350
watchglass: 192.151.158.26:8333 : Alive: (0.178s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=641350
watchglass: 143.202.160.10:8333 : Alive: (0.161s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=641350
watchglass: 208.94.240.42:8333 : Alive: (0.231s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=641350
watchglass: 108.31.170.3:8333 : (pool-108-31-170-3.washdc.fios.verizon.net) Alive: (0.334s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=641350 (Operator: asciilifeform)
watchglass: 213.109.238.156:8333 : Alive: (0.367s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=641350
watchglass: 188.121.168.69:8333 : (rev-188-121-168-69.radiolan.sk) Alive: (0.328s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=641350
watchglass: 176.9.59.199:8333 : Busy? (No answer in 20 sec.) (Operator: jurov)
watchglass: 103.36.92.112:8333 : Busy? (No answer in 20 sec.)
asciilifeform: !w uptime
watchglass: asciilifeform: time since my last reconnect : 6d 18h 49m
asciilifeform: !q uptime
snsabot: asciilifeform: time since my last reconnect : 3d 1h 17m
trinque: asciilifeform: I've seen this very same "socket is still connected, yet bot is offline" very infrequently with deedbot
trinque: I half suspect it's some bug on the freenode side
trinque: even considered having a pair of twins bot deployment which chatter to each other to confirm connection, rather than trusting that someone on the server side PONGed
trinque: but who gives a fuck
trinque: I'd rather have a socket connection to my WoT than to maintain this
trinque: and let the bots be a bridge to w/e toilet
trinque: *irc bots
asciilifeform: trinque: tcp -- and in particular, the standard stack for it -- per se, is (as discussed prev.) a quite sad kludge. for instance i still dunget why there's any trace of a socket left after process terminates in linuxen, for there to later 'socket in use'
asciilifeform: trinque: i also considered twin bots btw (even wrote a sync mechanism, but never made it automatic) . is slightly moar complicated than seems at 1st , because of (for instance) message reorderings and netsplits
asciilifeform tangentially, at one pt -- before realized that fleanode is run by clowns -- wondered wai they dun offer reliable (i.e. with message buffering) connections for $ . it'd be relatively simple to implement, on their end.
asciilifeform: trinque: at the risk of pedantry -- fleanode doesn't 'pong', they 'ping', client is expected to 'pong'.
asciilifeform at one time considered to host a fleanode noad in his rack. but then learned that they expect, for 100% phree, an ~entire~ 1u, w/ exclusive root, running whatever shitware, and offer absolutely nuffin in exchange
asciilifeform: ... so in light of this it aint the least bit surprising that the entire thing is nao owned by usg. cuz who else is willing to host fat boxen simply in exchange for getting to sniff packets directly from their arses.
asciilifeform: http://logs.nosuchlabs.com/log/asciilifeform/2020-07-29#1017475 << fixed the generics thing after all ? what was the mistake there, verisimilitude ?
snsabot: Logged on 2020-07-29 00:50:36 feedbot: http://verisimilitudes.net/2020-07-28 << A Syndication of Verisimilitudes -- An Implementation of SHA256 in Ada
verisimilitude: I was tired of fiddling with it, asciilifeform, but had a hunch renaming would work to get what I wanted from generics. A kind #ada user told me that, yes, renaming would work, and so I did. I'd more issues, solving several before publishing, but now it works well, and I'll update the documentation if I give it a comprehensive review before moving to the general SHA library.
verisimilitude: The documentation isn't yet to my standards, piggybacking from what I wrote for the SHA1, but the specification is very well documented, and the body isn't poor in this either.
asciilifeform: verisimilitude: have you compared performance to that of the sha2 that comes w/ gnat ?
verisimilitude: With this done, moving to the general SHA should be finished soon, but I don't always finish work on the schedule I'd prefer.
verisimilitude: No, asciilifeform.
verisimilitude: I've not looked at that at all, actually.
verisimilitude: I suppose that would be worth an article, however, since I was already planning something similar.
asciilifeform: 1, 2, possibly elsewhere (they actually did an ok job cutting the thing into orthogonal pieces for reuse in the various sha* variants)
verisimilitude: I'll take a look.
asciilifeform still needs to bake a civilized ada keccak...
verisimilitude: I wanted to be finished with SHA-3 by now, but obviously ain't.
verisimilitude: I'm going to go in the same APL to Common Lisp to Ada progression, likely.
verisimilitude: Common Lisp is going to be more difficult, since I'll be inclined to have separate codepaths for bits and octets, since the language has specializations for bits.
feedbot: http://mvdstandard.net/2020/07/drug-czar-moving-to-kill-cannabis-registry-for-home-growers-and-increase-sales-channels-for-weed/ << The Montevideo Standard -- Drug Czar Moving To Kill Cannabis Registry For Home Growers And Increase Sales Channels For Weed
← 2020-07-28 | 2020-07-30 →