01:16 |
asciilifeform |
lol! |
01:16 |
asciilifeform |
iirc was pocket change, too |
01:16 |
asciilifeform |
'illegal bitcoin'(tm)(r)(c) |
| |
~ 2 hours 14 minutes ~ |
03:31 |
bonechewer |
speaking of "illegal bitcoin", ru seems to now want to stamp it out in favor of a digital ruble |
| |
↖ |
| |
~ 11 hours 7 minutes ~ |
14:39 |
asciilifeform |
$ticker btc usd |
14:39 |
busybot |
Current BTC price in USD: $38717.43 |
14:39 |
asciilifeform |
$3 when!111 |
14:39 |
asciilifeform |
!w poll |
14:39 |
watchglass |
Polling 14 nodes... |
14:39 |
watchglass |
205.134.172.27:8333 : Alive: (0.023s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=719768 (Operator: asciilifeform) |
14:39 |
watchglass |
205.134.172.4:8333 : (172-4.core.ai.net) Alive: (0.082s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=719768 |
14:39 |
watchglass |
54.39.156.171:8333 : (ns562940.ip-54-39-156.net) Alive: (0.112s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=719768 |
14:39 |
watchglass |
71.191.220.241:8333 : (pool-71-191-220-241.washdc.fios.verizon.net) Alive: (0.098s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=719768 (Operator: asciilifeform) |
14:39 |
watchglass |
205.134.172.6:8333 : (172-6.core.ai.net) Alive: (0.081s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Return Addr=0.0.0.0:8333 Blocks=719768 |
14:39 |
watchglass |
205.134.172.26:8333 : Alive: (0.149s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Return Addr=0.0.0.0:8333 Blocks=719768 |
14:39 |
watchglass |
205.134.172.28:8333 : Alive: (0.082s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Return Addr=0.0.0.0:8333 Blocks=719768 (Operator: whaack) |
14:39 |
watchglass |
208.94.240.42:8333 : Alive: (0.204s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=719768 |
14:39 |
watchglass |
54.38.94.63:8333 : (ns3140226.ip-54-38-94.eu) Alive: (0.201s) V=88888 (/therealbitcoin.org:0.8.88.88/) Jumpers=0x1 (TRB-Compat.) Blocks=719768 |
14:39 |
watchglass |
94.176.238.102:8333 : (2ppf.s.time4vps.cloud) Alive: (0.323s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=718917 |
14:39 |
watchglass |
82.79.58.192:8333 : (static-82-79-58-192.rdsnet.ro) Alive: (0.328s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=719768 |
14:40 |
watchglass |
103.36.92.112:8333 : (terebe.ns01.net) Alive: (0.333s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=719768 |
14:41 |
asciilifeform |
http://logs.nosuchlabs.com/log/asciilifeform/2022-01-20#1075322 << seems they forgot to unban from last time before 'banning' again ? |
14:41 |
dulapbot |
Logged on 2022-01-20 22:31:51 bonechewer: speaking of "illegal bitcoin", ru seems to now want to stamp it out in favor of a digital ruble |
14:41 |
watchglass |
143.202.160.10:8333 : Busy? (No answer in 100 sec.) |
14:41 |
asciilifeform |
( 'pop where no one has pushed' ? ) |
14:41 |
watchglass |
75.106.222.93:8333 : Alive: (0.508s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=719768 |
14:42 |
* |
asciilifeform distinctly recalls there already being no Official goxes in ru, for yrs at this pt |
14:48 |
* |
asciilifeform doesn't view 'bans of goxes' as anti-btc politic; rather opposite. |
14:48 |
dulapbot |
Logged on 2021-12-29 22:21:01 asciilifeform: counts the usg-tolerated goxes , which buy & sell largely imaginary coins, dispense leverage, etc. under institutions |
14:48 |
dulapbot |
Logged on 2022-01-19 13:03:00 asciilifeform: the reason why various 2010s 'they'll ban goxes!' prognoses (incl. asciilifeform's) came to nuffin, is that turns out goxes are a very effective instrument of exchrate control, and therefore quite valuable to the reich. |
14:49 |
asciilifeform |
sadly not likely in the reich. |
14:49 |
dulapbot |
Logged on 2021-08-11 10:43:18 asciilifeform: doesn't expect it'll happen tho. short of a gift from the gods in the form of, say, a hamhanded usg ban which kills all goxes or similar |
| |
~ 2 hours 3 minutes ~ |
16:52 |
bonechewer |
Scott Locklin has some good snark relevant to asciilifeform's interests: "there is no reason to make computers the way they are presently made other than habit and lack of imagination" |
16:52 |
* |
asciilifeform read |
16:53 |
* |
bonechewer should have known |
16:57 |
asciilifeform |
bonechewer: asciilifeform back'n'forthed w/ fella for many yrs in his www comments |
17:12 |
whaack |
!e view-height |
17:12 |
trbexplorer |
block_height: 719787 |
17:13 |
asciilifeform |
hm whaack , is there an oddball utfism b/w ':' and '7' there ? |
17:13 |
whaack |
asciilifeform: yes there is i have it on my long list of notes to get rid of |
17:13 |
whaack |
it's a tab character |
17:13 |
asciilifeform |
a ok |
17:14 |
whaack |
the explorer now has automatic reorgs, so it will stay up to date with trb tip, it currently queries every 45 seconds to see if there's a new block |
17:14 |
asciilifeform |
neato! |
17:15 |
asciilifeform |
may be worth to show 'time since last block' |
17:17 |
whaack |
asciilifeform: okay i can add that to the view_height command |
17:18 |
asciilifeform |
a++ |
17:21 |
whaack |
orphaned blocks happen about once a day, correct? |
17:21 |
asciilifeform |
varies |
17:21 |
whaack |
i'm wondering how long it will be before i get to see my reorg code in action |
17:21 |
* |
asciilifeform defo seen days w/ >1 reorg |
17:21 |
asciilifeform |
seen weeks w/out reorg to |
17:22 |
whaack |
i guess it is all dependent on the status of the network between miners |
17:22 |
asciilifeform |
whaack: oughta be able to test a block eater w/ reorgism by feeding it blox from e.g. 'blkcut', from a trb db |
| |
↖ |
17:22 |
whaack |
in addition to of course the random probability of near simultaneous block discovery |
17:23 |
asciilifeform |
iirc they're in chrono. order |
17:27 |
signpost |
this market rip is so sexy to watch. |
17:28 |
signpost |
mats: what do you say; are you buying the dip? |
17:29 |
thimbronion |
Acquaintance of mine has lost 40k so far on a dog coin |
17:30 |
signpost |
chortle |
17:35 |
asciilifeform |
thimbronion: lost-lost or 'ohnoez,exch fell'-lost? |
17:36 |
thimbronion |
asciilifeform: exchange rate. obv has no clue how to self custody. |
| |
↖ |
17:37 |
* |
asciilifeform thinks 'current - costbasis' when things 'lost' |
17:39 |
thimbronion |
asciilifeform: that would be a negative number in her case |
17:41 |
asciilifeform |
presumably didn't wander into shitcoin-casino by happenstance, but knew what signed up for. |
17:42 |
asciilifeform |
casino aint built for you to win, lol, rather opposite. |
17:42 |
dulapbot |
Logged on 2022-01-19 13:56:02 asciilifeform: the 1 aspect common to all casinos is 'house wins' |
17:44 |
thimbronion |
asciilifeform: if she didn't know she does now |
17:45 |
asciilifeform |
if only erryone learned from 1st trip to casino. |
17:45 |
asciilifeform |
many folx -- do not |
17:46 |
asciilifeform |
(instead go to 'make system for beating the house' etc) |
17:58 |
whaack |
so last night i was thinking about the schema of my sql db that essentially mirrors the trb db in an organized, indexed fashion that i like and i was pondering about a better name for a table i have that is currently named 'output_input', which stores UTXOs when they are created and then marks the place where they are spent when they are consumed. And the better name I came up with was 'ticket' |
17:58 |
whaack |
(also considered 'voucher') and I thought that perhaps the concept and term 'ticket' is a better way to create an intuition for newbs on how bitcoin works under the hood |
17:59 |
whaack |
this is because tickets can have some nominal value, and they are issued, spent once, and then destroyed |
18:06 |
asciilifeform |
imho notbad term |
18:07 |
asciilifeform |
'ticket' is 'punched', could say |
18:07 |
whaack |
mhm ^ |
18:08 |
whaack |
there are various reasons why output, input, and utxos are confusing terms - first off is utxo has no tangible image that someone can put in their head (like ticket) and 'unspent' is a misleading term because when you spend something you have changed its location / owner, but that's not what happens in bitcoin, when you spend a utxo it is destroyed |
18:09 |
whaack |
or even better as you say, punched, since it still exists (in history) but it is marked as used and not to be used again |
18:10 |
asciilifeform |
well not destroyed, gets to live on disk 4evah if yer running a troo noad |
18:10 |
asciilifeform |
aha |
18:11 |
asciilifeform |
(not to mention that you'll need it again if the blox on top of it reorged away) |
18:12 |
asciilifeform |
( and see also... ) |
18:12 |
dulapbot |
Logged on 2021-11-12 12:15:51 asciilifeform: trb already has a max reorg length, we simply dun know what it is, lol |
18:12 |
whaack |
asciilifeform: is that because it will OOM or something ? |
18:12 |
whaack |
or politically collapse if it is too long? |
18:13 |
whaack |
likely both lol |
18:13 |
asciilifeform |
whaack: more immediately, cuz we haven't any notion what the longest possible reorg is |
| |
↖ |
18:13 |
asciilifeform |
(but more unpleasantly, also dun know what longest reorg trb will actually perform w/out choking) |
18:14 |
asciilifeform |
iirc longest historical one was... 8? |
18:14 |
asciilifeform |
^ may be worth documenting if doing a walk |
18:14 |
dulapbot |
Logged on 2022-01-21 12:23:13 asciilifeform: whaack: oughta be able to test a block eater w/ reorgism by feeding it blox from e.g. 'blkcut', from a trb db |
18:14 |
whaack |
wasn't there one that was like 87 blocks when someone found out how to give themselves like 2 ** 80 btc or something |
18:15 |
asciilifeform |
possibly, in the shitoshi days |
18:15 |
whaack |
yes it was back then iirc |
18:15 |
asciilifeform |
again imho worth to document |
18:15 |
whaack |
my reorg is on top of trb reorg, it assumes that trb is able to do the reorg |
18:16 |
* |
asciilifeform to date not encountered a trb in the field which perma-choked on a reorg |
18:23 |
signpost |
http://logs.nosuchlabs.com/log/asciilifeform/2022-01-21#1075380 << I'm probably approaching a wash on bitcoin if we end up ~20k, but who cares. |
| |
↖ |
18:23 |
dulapbot |
Logged on 2022-01-21 12:37:11 thimbronion: asciilifeform: exchange rate. obv has no clue how to self custody. |
18:23 |
signpost |
this is why one keeps cash on the side to eat when cheap again. |
18:24 |
asciilifeform |
imho is a matter b/w you & odin, 'when to hold / when to fold' etc |
18:24 |
* |
signpost ain't trying to maximize USD; if bitcoin fails I'll go maximize USD in one of the soviet design bureaus |
18:24 |
signpost |
sure |
18:25 |
asciilifeform |
'i'ma sell at 20k so i can buymoar at 100' is perfectly valid personal algo. |
18:25 |
asciilifeform |
( for that matter, asciilifeform's pov re subj ) |
18:25 |
dulapbot |
Logged on 2021-07-11 15:07:56 asciilifeform: tangentially, imho absolute best thing that could happen to btc from 'actual commerce' pov -- is that it were to fall to e.g. 100$ and stay there for next decade. |
18:26 |
thimbronion |
I personally prefer the price to stay high so I can work on meaningful things, as opposed to almost anything you can get fiat for doing. |
| |
↖ |
18:27 |
whaack |
yeah, i'm in strong disagreement with asciilifeform on this point lol |
18:27 |
asciilifeform |
right, yer choice of constants will depend on where you stand re hodl |
| |
↖ |
18:27 |
signpost |
I would rather every weak hand be bankrupted first. |
18:27 |
signpost |
then moon etc |
18:27 |
signpost |
so it stays. |
18:27 |
whaack |
right, little shakes like this are fantastic |
18:28 |
asciilifeform |
they're fascinating to watch, simply of 0 practical effect from asciilifeform's pov thus far |
18:28 |
signpost |
interesting that there was lots of institutional chatter before this. bet they're eating. |
18:30 |
whaack |
do these institutions ever touch btc or do they just have the coinbase web interface tell them they have some? |
18:32 |
signpost |
but it's coinbase *prime* whaack; they're fancy. |
18:39 |
asciilifeform |
lol |
18:40 |
* |
asciilifeform suspects is reason wai 'institutional' 'mega-buys' dun have much naked eye effect on gox exchrate |
18:40 |
asciilifeform |
it dun cost gox anyffin to 'print' fractional-reserve promisecoinz |
18:44 |
whaack |
well as much as i'd like to believe that coinbase is on thin ice with fractional reserve coins, if the % in segwit address meter is any indictator, they have a lot of fucking reserves. |
18:45 |
whaack |
although, of course, maybe coinbase doesn't even keep their own coldwallets with segwit addresses. |
18:45 |
asciilifeform |
whaack: nobody outside their bunker knows what the mass of obligations is tho |
18:46 |
whaack |
troo, and erry1 nowadays 'owns bitcoin' |
18:46 |
asciilifeform |
( and that's just 1 gox, there are others where nobody even pretends to publish a 'here's our stash' addr ) |
18:47 |
whaack |
does coinbase publish such an address? |
18:47 |
* |
asciilifeform regards all goxes as 'guilty until proven innocent' of being, well, goxes, i.e. fractional-reserve scams |
18:48 |
asciilifeform |
whaack: asciilifeform assumed from your observation above that it did ? ( nfi , not looked ) |
18:49 |
whaack |
no, i'm just saying that 25% or so of coins are in segwit addresses, which is a (perhaps poor) guesstimator of how many coins are controlled by enemy |
18:49 |
asciilifeform |
a |
18:49 |
asciilifeform |
this then doesn't tell us anyffin re solvency of coinbase |
18:50 |
signpost |
never hurts to treat them like they are, and exfiltrate bought coin immediately. |
18:50 |
asciilifeform |
signpost: aa but if 'immediately' then can't specu-trade over9000x/day, eh |
18:50 |
signpost |
*insolvent |
18:51 |
asciilifeform |
or 'leverage' |
18:51 |
thimbronion |
asciilifeform: if tru someone should tell the insurers: https://www.coindesk.com/markets/2019/04/02/255-million-coinbase-confirms-extent-of-crypto-insurance-coverage/ |
18:51 |
signpost |
nope, added benefit |
18:53 |
asciilifeform |
thimbronion: let's say coinbase has 8.5k coins and they walk away and insurance pays. at current exchrate insurance will 'pay for' all of'em. but where would coinbase find 8.5k coins to buy at that rate then. |
18:53 |
asciilifeform |
moar likely the goxers will be stuck w/ fixed usd per. if that |
18:54 |
whaack |
ofc if/when coinbase explodes the gox'd will get USD at best |
18:54 |
* |
signpost figures their control over keys is actually pretty good. |
18:54 |
thimbronion |
asciilifeform: wouldn't the insurer first verify the presence of the coins before insuring? This is wrt the premise that they are fractional reserve. |
18:55 |
* |
asciilifeform admits that he finds unclear how even possible to meaningfully insure coinz, given 'moral hazard' where they can be 'walked' and it made to look like anyffin at all. |
18:55 |
thimbronion |
don't get me wrong I hate coinbase with a passion |
18:56 |
thimbronion |
otoh unlike the orig gox it is a public company with various obligations |
18:56 |
signpost |
asciilifeform: this is the kind of bureaucracy usg.corps love and excel at. |
18:56 |
asciilifeform |
imho primary hazard to goxcoins aint 'walking' but simple bank run. |
18:57 |
signpost |
hack, sure. doubt they are running fractional though. |
18:57 |
asciilifeform |
what'd be their incentive not to, tho ? |
18:57 |
whaack |
coinbase is a much more professional gox, no doubt |
18:58 |
signpost |
asciilifeform: goes back to that thread about prosecutors and "cost-of-acquistion" vs roi |
18:58 |
asciilifeform |
aha |
18:58 |
asciilifeform |
and not many things w/ moar roi than to 'print' coinz neh |
18:58 |
asciilifeform |
esp. given the impossibility of proving that you aint doing so |
18:59 |
asciilifeform |
goxes are in that sense 'lemon market' |
18:59 |
asciilifeform |
how woudlja go about proving that your gox aint fractional? |
18:59 |
signpost |
local american tech already has infinite money |
18:59 |
thimbronion |
asciilifeform: someone proved it to the insurance company's satisfaction |
18:59 |
signpost |
*loyal |
18:59 |
signpost |
they don't need more money. |
19:00 |
* |
signpost operating on laggy link atm |
19:02 |
signpost |
anyway, can't be proved. and most likely time to lie is after a hack. |
19:02 |
* |
asciilifeform doesn't deny the abstract physical possibility of a solvent gox. but finds difficult to imagine. |
19:02 |
signpost |
"all goxes are insolvent in the limit" is a reasonable view. |
19:03 |
asciilifeform |
'solvent until you need to withdraw' |
19:03 |
signpost |
just risk management there. never wire more than you can lose, and always transfer out to 1-address |
19:18 |
asciilifeform |
indeed , if one uses goxes the way asciilifeform used on 1 occasion mtgox, unlikely to lose yer coin |
19:18 |
* |
asciilifeform in fact recently used a gox to recycle some shitcoins, in exactly this 'in - out' style |
19:18 |
dulapbot |
Logged on 2021-12-02 14:33:51 asciilifeform: recently recycled some phorklulzcoinz |
19:22 |
asciilifeform |
re: hypothetical mega-sales of coinz -- i'd offer to my l1 1st ( if only we still had the auctionbot... ) |
19:22 |
dulapbot |
Logged on 2022-01-21 13:23:33 signpost: http://logs.nosuchlabs.com/log/asciilifeform/2022-01-21#1075380 << I'm probably approaching a wash on bitcoin if we end up ~20k, but who cares. |
19:22 |
asciilifeform |
before goin' to a gox, that is |
19:25 |
* |
signpost only buying, had some cash arrive after btc was already in vertical moon-mode. |
19:31 |
asciilifeform |
right, this in re hypothetical btcocalypses (or simply folx needin' a parcel of fiatola in a hurry) |
19:37 |
* |
thimbronion sells on a roughly monthly bases, but all sales will be reported to the IRS, etc. |
19:37 |
thimbronion |
*basis |
| |
~ 29 minutes ~ |
20:07 |
PeterL |
http://logs.nosuchlabs.com/log/asciilifeform/2022-01-09#1071643 << what's the incantation in sqlite to manually delete a key? |
20:07 |
dulapbot |
Logged on 2022-01-09 15:05:39 billymg: you don't have to nuke the whole db, just go into sqlite and delete the key for the user you just unpeered |
20:09 |
thimbronion |
PeterL delete from keys where key='FbJF...' |
20:10 |
whaack |
man i must say i think this sharing of keys even while testing is terrible, it's like building dangerous muscle memory |
| |
↖ |
20:10 |
thimbronion |
whaack: that's a fake key |
20:10 |
whaack |
ah |
20:17 |
mats |
signpost: yes |
20:17 |
mats |
ABB always b buyin |
20:31 |
thimbronion |
$ticker btc usd |
20:31 |
busybot |
Current BTC price in USD: $37848.52 |
20:32 |
thimbronion |
wen 19k? |
20:39 |
mats |
https://news.ycombinator.com/item?id=30025298 |
| |
↖ |
20:44 |
thimbronion |
https://twitter.com/DeItaone/status/1484620305020694533 |
20:47 |
signpost |
thimbronion: I was just digging for that one myself |
20:48 |
mats |
where is the actual content? |
20:51 |
signpost |
https://archive.fo/HpS0E |
20:52 |
signpost |
this sounds like fud |
20:52 |
signpost |
"shares fell nearly 18%" << all of tech is falling |
20:53 |
signpost |
that said, one of my favorite conspiracy theories is that saylor's a fed tasked with blowing up bitcoin will the collapse of his gigantic leveraged hodl. |
| |
↖ |
20:54 |
signpost |
*with |
20:54 |
thimbronion |
when it rains fud it pours fud |
20:54 |
mats |
nice, mstr discount day |
20:54 |
mats |
gbtc is also at 26.7% discount to nav |
20:56 |
thimbronion |
signpost: interesting never heard that one. of all the public figure wallstreet bitcon face fags he seems to actually get it the most though. |
20:57 |
signpost |
that's why he's the perfect fed! |
20:57 |
* |
signpost wiggles eyebrows |
20:57 |
signpost |
bbl |
20:57 |
thimbronion |
lol! |
21:04 |
* |
whaack is sus of saylor |
21:04 |
whaack |
seems elon musk-like |
21:16 |
thimbronion |
whaack: his strategy seems sound to me. as a public corp he can take on massive amounts of low interest debt and buy btc with it. |
21:27 |
asciilifeform |
http://logs.nosuchlabs.com/log/asciilifeform/2022-01-21#1075499 << >> http://logs.nosuchlabs.com/log/asciilifeform/2022-01-12#1072642 |
21:27 |
dulapbot |
Logged on 2022-01-21 15:10:38 whaack: man i must say i think this sharing of keys even while testing is terrible, it's like building dangerous muscle memory |
21:27 |
dulapbot |
Logged on 2022-01-12 19:26:33 bitbot: (pest) 2022-01-13 asciilifeform: http://logs.bitdash.io/pest/2022-01-12#1002644 << we're gonna need some means of identifying keys for purposes like this w/out leaking bits. possibly e.g. crc32(sha512(key)). will amend spec after folx comment. |
21:28 |
asciilifeform |
http://logs.nosuchlabs.com/log/asciilifeform/2022-01-21#1075514 << >> see also |
21:28 |
dulapbot |
Logged on 2022-01-21 15:54:09 signpost: that said, one of my favorite conspiracy theories is that saylor's a fed tasked with blowing up bitcoin will the collapse of his gigantic leveraged hodl. |
21:28 |
dulapbot |
Logged on 2021-12-28 23:31:44 asciilifeform: vex: imho the sooner *all* the slimy chrematist scum maggoting on top of bitcoin go to gasenwagen, off cliffs, to feed the fishes, etc. the better. |
21:28 |
dulapbot |
Logged on 2021-08-11 10:42:12 asciilifeform: http://logs.nosuchlabs.com/log/asciilifeform/2021-08-11#1051957 << whole space is imho sorely in need of a great flood to wash away all the scum. |
| |
~ 1 hours 1 minutes ~ |
22:29 |
thimbronion |
Would anyone here object if at some future point the blatta db sat inside postgreql instead of sqlite? |
| |
↖ |
22:30 |
thimbronion |
Learning today the extent of sqlite3's limits such as no way to drop or add columns without creating new tables and renaming them. |
22:32 |
asciilifeform |
ugh |
22:32 |
asciilifeform |
thimbronion: asciilifeform very much likes to be able to shuffle blatta.db around on filesystem for testing |
22:33 |
asciilifeform |
having a pgism would muchly get in the way of this |
22:33 |
asciilifeform |
(not to mention make the operation of multiple blattas on 1 box rather gnarly) |
22:33 |
thimbronion |
asciilifeform: from my perspective it would be a pain for folks without sysadmin experience to run postgresql also |
22:34 |
asciilifeform |
it's an ugh, asciilifeform in e.g. phuctor did not even consider moving from sqlite to postgres until had multi-GB dataset |
22:34 |
asciilifeform |
cuz it's a whole other world of pain |
22:42 |
asciilifeform |
in an adult pestron, the only things in the station db oughta be wot/keys/at/knobs. (and defo not messages, for the obv. reason ) |
22:42 |
dulapbot |
Logged on 2022-01-20 13:59:15 asciilifeform: if an arbitrary packet can force you to access disk, yer sunk |
22:43 |
asciilifeform |
i.e. a coupla kB at most. |
22:44 |
asciilifeform |
( ~possibly~ some msg logs, but strictly for 'getdata' on best-effort basis ) |
22:44 |
thimbronion |
asciilifeform: oh hm didn't fully comprehend that. So pestrons should not store all logs, but only n log entries, and in memory only? |
22:45 |
asciilifeform |
thimbronion: e.g. longbuffer gotta live in memory (and be quite compact ) |
22:45 |
dulapbot |
Logged on 2022-01-20 13:58:44 asciilifeform: indeed oughta be queryable in o(n log n) or around |
22:45 |
asciilifeform |
otherwise yer trivially hosable |
22:45 |
dulapbot |
Logged on 2022-01-20 12:01:00 asciilifeform: under the current algo, a flood of captured packets turned into a dupe flood which includes e.g. a prod will end up querying the longbuffer errytime and exceed the avail. cycles to maintain linerate rejection. |
22:47 |
asciilifeform |
pestron oughta be able to store arbitrary mass of logs, but in such a way that only getdata (which only can come from a peer) can cause you to look on disk |
22:47 |
asciilifeform |
^ major subject on asciilifeform's to-do for 0xfa |
22:47 |
thimbronion |
asciilifeform: ok haven't gotten to longbuffer yet in my re-read. |
22:48 |
asciilifeform |
a ok |
22:49 |
asciilifeform |
( to rephrase above -- no hearsay packet should ever result in disk access ) |
22:52 |
asciilifeform |
^ nor, naturally, any packet from a stranger, incl. a replay of genuine packet ) |
22:54 |
thimbronion |
asciilifeform: in that case it doesn't make a ton of sense for me to go to much effort to preserve existing logs in the migration process for the next release, yes? |
22:54 |
thimbronion |
we have billymg's pest logger anyway |
22:54 |
asciilifeform |
thimbronion: given as blatta doesn't look at chain hashes yet, i'd suspect not |
22:54 |
asciilifeform |
(once it does, will need to keep'em somewhere) |
22:54 |
thimbronion |
asciilifeform: unreleased version does |
22:55 |
asciilifeform |
well if unreleased, not makes a diff there, does it |
22:55 |
asciilifeform |
logging end is where imho a pg glue would be useful |
22:56 |
asciilifeform |
(for folx who want to stand up existing logotron, for instance) |
22:56 |
asciilifeform |
but (as described upstack) gotta be 'write only' |
22:57 |
thimbronion |
asciilifeform: it might? If next release keeps the logs it should prevent spurious "forked!" warnings. |
22:58 |
asciilifeform |
as i understand erryone will see these when 1st runs the new one |
22:58 |
thimbronion |
or spurious "met <speaker>" warnings |
22:58 |
asciilifeform |
( well not 'forked', but 'met' yea) |
22:59 |
asciilifeform |
i suspect eryone's gonna rekey anyways, and will be reasonably certain who he's talking to |
22:59 |
thimbronion |
alright I'll just blow away the logs then |
22:59 |
asciilifeform |
( also recall that forkage msgs only appear from hearsays ) |
22:59 |
asciilifeform |
per spec |
22:59 |
* |
asciilifeform tries to remember whether specified this detail |
23:00 |
asciilifeform |
possibly not, lol |
23:00 |
thimbronion |
asciilifeform: this is new to my understanding - could have missed it in whatever version I read |
23:00 |
asciilifeform |
thimbronion: a msg from a direct peer is prima facie authentic |
| |
↖ |
23:00 |
asciilifeform |
i.e. if he tells you his selfchain is nao zero, then it is troo |
23:01 |
asciilifeform |
( who has key -- is your peer , definitionally ) |
23:01 |
thimbronion |
yeah I get it now. didn't realize before that forking was only for hearsy. thought perhaps it was for letting you know you missed a message from a peer. |
23:02 |
asciilifeform |
is also for closing chain gaps, yes |
23:02 |
asciilifeform |
the chain hashes, that is |
23:02 |
thimbronion |
but with getdata, this should not happen |
23:02 |
asciilifeform |
well the hash is how you know when you gotta getdata |
23:02 |
thimbronion |
although theoretically it *could* |
23:03 |
asciilifeform |
( the concrete purpose of fork alarm is this ) |
23:03 |
dulapbot |
Logged on 2022-01-19 22:49:54 asciilifeform: they're a pill against ~indistinguishability~ of colliding speakers. |
23:03 |
thimbronion |
i.e. if somehow the log was missing from all peers, getdata could fail |
23:03 |
asciilifeform |
indeed would |
23:06 |
asciilifeform |
thimbronion: btw in your reread note the new mechanism for explicitly showing your current chain tips to a peer. |
23:07 |
thimbronion |
asciilifeform: ah so that's what that's for |
23:07 |
asciilifeform |
it's also for nat drilling via addr cast, but yea |
23:08 |
* |
asciilifeform realizes the spec will need a good deal of massage still |
23:08 |
asciilifeform |
ideally when 'fully baked', the purpose of ea. part will be quite obvious |
23:09 |
asciilifeform |
i.e. no reader will think to say 'so that's what that's for', lol |
23:09 |
thimbronion |
asciilifeform: reread the section a few times and couldn't figure it out |
23:09 |
* |
asciilifeform believes |
23:10 |
thimbronion |
figured it would be made clearer after reading the rest |
23:10 |
asciilifeform |
plz lemme know whether in fact is, when finish rereading |
23:10 |
* |
asciilifeform must bbl |
23:24 |
shinohai |
http://logs.nosuchlabs.com/log/asciilifeform/2022-01-21#1075534 <<< what version of sqlite you using? iirc newer sqlite lets ya `ALTER TABLE <name> DROP COLUMN`. |
23:24 |
dulapbot |
Logged on 2022-01-21 17:30:12 thimbronion: Would anyone here object if at some future point the blatta db sat inside postgreql instead of sqlite? |
23:25 |
shinohai |
( 3.35.5 here if curious) |
23:36 |
thimbronion |
I have 3.34 hrm |