| 03:52 |
crtdaydreams |
shinohai: rip |
| |
↖ |
| 03:53 |
crtdaydreams |
betcha SBF d00d has a few satoshis stashed away |
| 03:53 |
crtdaydreams |
"oh noes, am worth -$14.2 bil overnight" "anyways" |
| 03:55 |
* |
crtdaydreams proceeds to embark on USG funded ~~vacation~~ 'business trip' to the bahamas 16 mo from now |
| |
↖ |
| 03:56 |
crtdaydreams |
either that or like all other shitcoinistas before him; scrubbed off the map entirely |
| |
~ 4 hours 15 minutes ~ |
| 08:12 |
shinohai |
http://logs.bitdash.io/pest/2022-11-10#1015850 << Hey at least "Almadea research" researched how to keep the plates spinning for *this* long. Too bad that also didn't include how to dump your shit on another "major" exchange. |
| 08:12 |
bitbot |
Logged on 2022-11-10 03:52:42 crtdaydreams[jonsykkel|awt]: shinohai: rip |
| |
~ 2 hours 5 minutes ~ |
| 10:18 |
asciilifeform |
shinohai: context? |
| |
~ 1 hours 1 minutes ~ |
| 11:20 |
billymg |
$ticker btc usd |
| 11:20 |
busybot |
Current BTC price in USD: $17750.2 |
| 11:20 |
billymg |
according to my chart lines it needs to go above $18.6k and stay there to be out of the woods |
| |
↖ |
| |
~ 15 minutes ~ |
| 11:35 |
* |
shinohai has no chart lines, performs coin divination by tossing pickup stix on the floor ..... |
| 11:39 |
billymg |
place ipad with btc chart on table, toss pickup stix on ipad. now you've got chart lines! |
| 11:51 |
asciilifeform |
meanwhile: signpost , awt , jonsykkel , phf, billymg , et al : preview of 0xfa. a number of sections not done, and prolly over9000 typos but fwiw. |
| |
↖ ↖ |
| 11:51 |
* |
asciilifeform will include the html ver. once finished and vpatch |
| |
↖ |
| |
~ 41 minutes ~ |
| 12:33 |
billymg |
asciilifeform: nice! |
| 12:46 |
asciilifeform |
comments/bugreports invited. |
| 12:48 |
signpost |
reading now. the diagrams look very nice. will indeed look great on dead tree when appropriate. |
| 12:48 |
asciilifeform |
over9000 things missing (some marked, some not). e.g. not has illustration of chain mechanics (where notion is, in a gui client not need to 'detect forks', simply draw arrows to follow selfchain) |
| 12:49 |
asciilifeform |
in re new moving parts -- introduced 'multipart' message mark (in such way as back-compat. with existing pestrons) |
| 12:49 |
asciilifeform |
addrcast revised to burn less cpu (only attempt decode if originator handle matches a cold peer, given as is meaningless anyway if not in prev. spec) |
| 12:50 |
asciilifeform |
'inv' msg introduced (for use when folx start using netchain for threaded replies, and can no longer rely on netchain to represent 'tip' of conversation) |
| 12:50 |
asciilifeform |
'binary' msg type for eventual use w/ signpost's lubytron. |
| 12:51 |
asciilifeform |
just about errything else -- similar to previous, but organized in moar logical way so to be readable |
| 12:51 |
asciilifeform |
all discussion of 'forks' nixed, imho is job for frontend of client |
| 12:51 |
asciilifeform |
ditto reordering/interning of messages |
| 12:55 |
asciilifeform |
the 1 major change -- nomoar buffers, all valid msgs of the 'chained' (i.e. loggable) type go straight to db |
| 12:55 |
asciilifeform |
(if arrive outta order -- not problem in graphical frontend ) |
| 12:56 |
asciilifeform |
ditto if copies with lower bounce (incl. immediates) recvd. |
| 12:56 |
asciilifeform |
(updates log) |
| 12:56 |
asciilifeform |
i.e. most annoyingly gnarly, imho, piece of old spec -- vanished |
| 12:57 |
signpost |
glad to see UI implementation detail backflow dropped from the spec, yup. |
| 12:57 |
asciilifeform |
the 1 case where makes sense to 'sit on' a message is where bounce > lowest_historical(speaker). still need to stuff that into sect. 5. |
| 12:58 |
asciilifeform |
(generally one wants to propagate with min-bounce when possible, otherwise msg can run along longer-than-req'd path and end up cut off prematurely by max-bounce somewhere) |
| 12:59 |
signpost |
having my wife read too because I'm curious what an educated (semi-)nontechnical takes from it. |
| 12:59 |
asciilifeform |
in 5.4.10.5 , 'port hammering' introduced. (may need massage) |
| 12:59 |
signpost |
at least the first few sections |
| 12:59 |
asciilifeform |
signpost: nifty. asciilifeform fed sect1 to pet, similarly |
| 13:05 |
asciilifeform |
1 remaining q is wat to do with malformed msgs that are pointed to by a chain. (current text: 'discard') but potentially results in infinite fetches via getdata. prolly oughta store'em but mark 'broken heart' to not display in console |
| 13:06 |
asciilifeform |
(malformed -- not 'martian' but break some rule, e.g. invalid utf, or circular chain, or direct but bounce not zero, or 1 of several dozen other boojums) |
| 13:07 |
asciilifeform |
ideally nobody will send these but can't be ruled out, so gotta handle'em. |
| 13:08 |
asciilifeform |
iirc phf pointed out this gotcha in old days when 'gossip' etc discussed |
| 13:09 |
signpost |
seems useful to surface these so bugs are immediately obvious to users. |
| 13:10 |
signpost |
orphanage log or something |
| 13:12 |
asciilifeform |
well, same log, but with flag 'not goes to console' |
| 13:13 |
asciilifeform |
(ideally would show as specially-marked turd in gui, show hexdump if clicked or somesuch) |
| 13:13 |
asciilifeform |
at some pt there likely will be pestnets with multitude of not necessarily 100% compat. clients; then will be useful |
| 13:14 |
asciilifeform |
the important part is to gracefully handle malformed msgs, in particular -- not to infinitely attempt to getdata'em |
| 13:15 |
signpost |
asciilifeform: wife points out "adddressless" typo in 1.7 paragraph 2 |
| 13:15 |
signpost |
and yup, makes sense |
| 13:15 |
asciilifeform |
signpost: ty. prolly over9000 typos, asciilifeform not yet made any attempt to comb for'em |
| 13:15 |
signpost |
nbd |
| 13:15 |
asciilifeform |
broken links, too, likely |
| |
~ 23 minutes ~ |
| 13:39 |
shinohai |
"Tether reportedly begins freezing $USDT addresses owned by FTX at the request of law enforcement" <<< LMAO! |
| 13:42 |
asciilifeform |
next yr: 'blather begins freezing tether addrs at req of lol enforcement' |
| 13:45 |
asciilifeform |
$ticker btc usd |
| 13:45 |
busybot |
Current BTC price in USD: $17325 |
| 13:45 |
asciilifeform |
notbad waterfalling |
| |
~ 24 minutes ~ |
| 14:10 |
billymg |
guys, he's sorry! https://twitter.com/SBF_FTX/status/1590709166515310593 |
| 14:11 |
asciilifeform |
iirc karpeles also 'was sorry'(tm) |
| 14:16 |
signpost |
asciilifeform: what brings in the 0xF after N in chained multipart msgs? byte alignment? |
| 14:17 |
asciilifeform |
signpost: 'Of' rather than 0f |
| 14:17 |
asciilifeform |
click the link, it's a named field |
| 14:17 |
signpost |
oh derp |
| 14:17 |
asciilifeform |
lol |
| 14:17 |
signpost |
3 of 10 etc |
| 14:17 |
asciilifeform |
aha |
| 14:18 |
* |
signpost shakes face, continues |
| 14:18 |
asciilifeform |
(4.2.2.8) |
| 14:31 |
signpost |
close paren possibly misplaced in 5.4.5.3 |
| 14:31 |
asciilifeform |
indeed, ty signpost |
| 14:45 |
signpost |
the spec is beautiful, very clear. |
| 14:45 |
* |
signpost will want to get his hands into his own implementation at some point after his encoder's done |
| 14:46 |
asciilifeform |
ty signpost . imho could be made clearer, still needs (in addition to the obv. missing chapters) pseudocoad algos for msg eating/shitting etc, illustration of chain mechanics, coupla other things |
| 14:46 |
asciilifeform |
a naive reader may not immed. understand why messages are chained, and why some but not others, for instance |
| 14:47 |
signpost |
looking forward to when the spec itself can sit in asciilifeform's chain of messages, either as multipart message or hash-reference to pdf. |
| 14:47 |
signpost |
yup. that said my wife was able to explain to me why this was better than signal, so that's pretty cool. |
| 14:47 |
asciilifeform |
neat |
| 14:47 |
signpost |
written as is. and she doesn't code, though obviously listens to me discuss it all the time. |
| |
~ 29 minutes ~ |
| 15:16 |
jonsykkel |
nice will read |
| 15:17 |
jonsykkel |
ohw does one generate circular chain without breaking sha257 |
| 15:19 |
asciilifeform |
jonsykkel: can't presently think of a way |