357800+ entries in 0.234s

jurov: mircea_popescu:
to get rid of blockchain and address problems, and have richcoin, just use csv file with noodles of gpg signatures
mircea_popescu: if i were anyone else other
than me i'd have a head
the size of kamcheatka by now.
mircea_popescu: you gotta choose. either for
the poor and
then shit, or else gold but not for
the poor.
mircea_popescu: fundamental problem here is
that bitcoin started life
trying
to be "gold for
the poor".
mircea_popescu: asciilifeform
the correct solution is
to keep a
table.
mircea_popescu: ;;later
tell pete_dushenski re
the missing reports : i do intend
to publish
them asap. been delayed by an assortment of issues such as argentinian power, bitbet adventures etc.
they'll get
there.
mircea_popescu: protocol suffers from
the same pass-by-reference issues
the codebase suffers from.
mircea_popescu: kakobrekla no,
the only
thing an address can do is send its whole contents.
assbot: Logged on 07-01-2016 00:15:36; mircea_popescu: gotta get a COMPETENT accountant
to sit down and make a full list on fucking papert and
then croiss
them out and etc.
assbot: Logged on 07-01-2016 00:16:10; mircea_popescu: we're
talking ~400k fiats here or some shit.
assbot: Logged on 07-01-2016 00:10:34; mircea_popescu: yeah, right, spending
the 10's of k's
to fix
this is not good enough, ima commit
to endless millions
to have electrum audited
to production environment level.
assbot: Logged on 07-01-2016 00:09:03; kakobrekla: phucked
txes did not get removed from wallet file with rescan
assbot: Logged on 07-01-2016 00:06:18; mircea_popescu: here asciilifeform, even more stuff
to fix off your father's bitcoin.
assbot: Logged on 07-01-2016 00:06:03; mircea_popescu: i never afore
today realised
this braindamage is actually in
there ;/
assbot: Logged on 06-01-2016 23:57:31; mircea_popescu: a
transaction
that spends bitcoin from an address,
that is signed and
that has
the bitcoin may never be rejected.
assbot: Logged on 06-01-2016 23:42:35; mircea_popescu: i am not in
the business of debugging
the blockchain, really. i am in
the business of making payments. i can in fact prove
that i did sign a
transaction, from an address
that had
the coin,
to
the addresses
that should get it.
kakobrekla: "send
their
total contents as follows" now i noted
the 'total' -
then you must never reuse an address.
☟︎ kakobrekla: also, if you do not change something in
the process of paying out bbet winning
this
thing is bound
to repeat sooner or later.
☟︎ kakobrekla: sign
to move 1 btc from
that address
to another
mircea_popescu: this is
the correct format for a bitcoin
transaction : "address i[1..n] send
their
total contents as follows : B1
to address j1, B2
to address j2, Bm
to address jm.
therefore signed."
mircea_popescu: <kakobrekla> no, because <mircea_popescu> a
transaction
that spends bitcoin from an address,
that is signed and
that has
the bitcoin may never be rejected. << once one's accepted [some of]
the others no longer are discussing an address
that has
the btc.
mircea_popescu: well... so bitbet update : we went from "bitcoin is broken, will be fixed at an unspecified future date no earlier
than 2020"
to "bitcoin is borken, we might have a circumvention in place before
the year is out".
mircea_popescu: gotta get a COMPETENT accountant
to sit down and make a full list on fucking papert and
then croiss
them out and etc.
☟︎ kakobrekla: mircea_popescu as long as
the backup contains all
the used change addresses you are ok
kakobrekla: > unconfirmed
transactions relating
to your own wallet addresses are stored not only in
the memory pool, but in wallet.dat, so a restart won't clear
them.
To accomplish
that, you can start bitcoind with
the recently added -zapwallettxes option.
This will cause bitcoind
to forget all
transactions associated
to your addresses, and rescan
the block chain
to reconstruct
them. In particular, any unconfirmed
transactions will be forgotten and
kakobrekla: if you have
the addresses you will get
the
txes
mircea_popescu: kakobrekla fancy, "use a backup". yes, i have backups. and how
the fuck do i identify which payments of
the HUGE batch of bitbet payments for end of year were made and which weren't, so i don't end up doublepaying ?
kakobrekla: i have not
tested it yet but
this is suppose
to fix
that
kakobrekla: btw at some point 'zapwallettxes' was added
to bitcoind
pete_dushenski feels fortunate
to not be in physical proximity
to mp rage atm.
mircea_popescu: yeah, right, spending
the 10's of k's
to fix
this is not good enough, ima commit
to endless millions
to have electrum audited
to production environment level.
☟︎ kakobrekla: yes you need
to remove bad
txes from wallet.dat by removing
them with some wallet inspector
tool or use a backup from before bad
txes came in
pete_dushenski: reliant robin is an excellent example of bitcoin as it works
today : looks
to have
the right parts but has
them in
the wrong order and as a result it doesn't quite work as intended. compare
this with
the morgan 3-wheeler, which is perhaps where
trb is headed, which has
the same parts, but not welded
together by apes.
kakobrekla: phucked
txes did not get removed from wallet file with rescan
☟︎ mircea_popescu: ima have
the
thing rescanned and resent. but god damned
this has
to be fixed omfg.
mircea_popescu: nah, he drove a sort of british
trabant. kept causing someone else's robin
to overturrn
kakobrekla: no, because <mircea_popescu> a
transaction
that spends bitcoin from an address,
that is signed and
that has
the bitcoin may never be rejected.
mircea_popescu: here asciilifeform, even more stuff
to fix off your father's bitcoin.
☟︎ kakobrekla: if it werent so you could just phuck a
txid and respend
the address multiple
time given it has balance
mircea_popescu: i never afore
today realised
this braindamage is actually in
there ;/
☟︎ mircea_popescu: what
the fuck sort of braindamaged zombie idiot designs
things like
this ;/
jurov: i messed
that up as usual... so adgain: input is: a hash of a previous
transaction + Index of
the specific output in
the referenced
transaction.
pete_dushenski: also, "mircea_popescu: Fucking horrid
terminology, we really shouldn’t be in a situation where input and output means
the same
thing."
jurov: input is a reference
to an output from a previous
transaction. output is: a hash of a previous
transaction + Index of
the specific output in
the referenced
transaction.
☟︎ danielpbarron: a
transaction must refer
to specific other
transactions, or is a miner supposed
to decide which coin days
to destroy?
mircea_popescu: this is irrelevant.
txids are not part of
transactions.
punkman: inputs reference
txids not adresses, do
they not?
kakobrekla: yes but perhaps
the client selected
T2 in
the next
transaction you made and
T1 in
the
transaction after
that.
mircea_popescu: a
transaction
that spends bitcoin from an address,
that is signed and
that has
the bitcoin may never be rejected.
☟︎ mircea_popescu: then you should say "i see a doublespend on your input for
tx d8e60a0979024cce9159bd7b59594eae7dec627db5ac6bd088866afea533b0e9"
kakobrekla: T is relevant because unconfirmed coins on
the change address are ok
to be sent.
mircea_popescu: there is not an acceptable explanation for why
this should happen. in fact, it may not happen. period.
mircea_popescu: T is irrelevant for
this discussion. i send btc
to address x.
this confirms. what
t it has does not matter. i send btc from address x.
this fails
to be included.
kakobrekla: you make a
txid
T1 which gets phucked and your wallet ends up with a confirmed
T2 and unconfirmed
T1.
this
T1 stays in
the wallet forever if you dont do anything
to it.
the client may select
that never-to-be-confirmed
T1 change address at any
time in
the future if you dont not-use it specifically.
mircea_popescu: no dude, it's not orphaned or anything. it's in
the chain. again, 400+ fucking confirmations should mean something.
punkman: so you used a 6+ conf output
that got orphaned or what?
mircea_popescu: in any case : at
the
time d8e60a0979024cce9159bd7b59594eae7dec627db5ac6bd088866afea533b0e9 was broadcast, one of
the numerous
txids corresponding
to its input had in fact been confirmed, for 6+ blocks.
mircea_popescu: in fact, if anything can change output address, or if a
third party can invalidate a valid
txn i signed, bitcoin is dead.
kakobrekla: yes it will
try
to use
the outputs of
that broken
tx
mircea_popescu: as long as
the fucking address has coins and was signed,
that is ALL.
mircea_popescu: kakobrekla
transactions consist of a signed bitcoin address. not of a signed bitcoin
txnid.
kakobrekla: the wallet keeps broken
tx (a double spend attempt) and can use it for
the next
tx which wont go
through, ever.
mircea_popescu: but
that is ALSO irrelevant : if
the address had only cpoins now,
the
tx should go out... noe.
jurov: when you issued
the
tx?