< pedro>
Is there a way to validate a random transaction via transaction ID on the debug window? Imagine I want to validate a transaction between 2 wallets I don't control. I believe its possible to check if such a transaction exists on the mempool "getmempoolentry" but can I also check the confirmations without having to decode a whole block like with "getrawtransaction"?
< Dean_Guss>
people can and do control the centralized dns - mostly core devs running their own services on their own domains for your convenience. Go ahead and start up bitcoind and you can see for yourself
< echeveria>
Dean_Guss: please don't talk here if you're not going to tell the truth.
< echeveria>
qpeng: this would be a better question for #bitcoin.
< Dean_Guss>
k. I will shut up
< qpeng>
Dean Guss thx,
< luke-jr>
qpeng: the DNS seeds are only used in an untrusted manner for the first run; after that, the node keeps track of its own peers its seen
< stevenroose>
In rust-bitcoin this somehow fails because of the segwit code that gets activated when the inputs are empty.
< stevenroose>
And then the outputs are skipped and some bytes are left over.
< stevenroose>
Does Core have something special to differentiate "no inputs -> segwit" and "no inputs -> incomplete tx"?
< stevenroose>
The SERIALIZE_TRANSACTION_NO_WITNESS is only used internally for serializing without the witness, right?
< stevenroose>
Ah, seems that decoderawtransaction first tries without witness, then tries with witness
< sipa>
stevenroose: it has heuristics where it tries to decode both; if only one succeeds while consuming all input bytes completely, that's the solution. if both succeed some sanity tests are used
< sipa>
stevenroose: decoderawtransaction has a parameter to tell it whether it's possibly a witness tx or not for this reason
< sipa>
stevenroose: and psbt also solves this
< promag>
sounds good in having a RPC endpoint with no wallet support?
< bitcoin-git>
[bitcoin] laanwj closed pull request #15209: zmq: log outbound message high water mark when reusing socket (master...zmq-always-log-high-water-mark) https://github.com/bitcoin/bitcoin/pull/15209
< gkrizek>
Would be great to get a few more reviews on #15196. Master CI is failing until this gets merged
< gribble>
https://github.com/bitcoin/bitcoin/issues/15196 | [test]: Update all subprocess.check_output functions to be Python 3.4 compatible by gkrizek · Pull Request #15196 · bitcoin/bitcoin · GitHub
< bitcoin-git>
bitcoin/master 85f0ca9 Gregory Sanders: Remove errant past from walletcreatefundedpsbt for nLocktime replaceability
< bitcoin-git>
bitcoin/master 6be52d6 Wladimir J. van der Laan: Merge #15213: doc: Remove errant paste from walletcreatefundedpsbt for nLocktime replaceable...
< bitcoin-git>
[bitcoin] laanwj closed pull request #15213: doc: Remove errant paste from walletcreatefundedpsbt for nLocktime replaceable (master...wcfpsbt_paste_err) https://github.com/bitcoin/bitcoin/pull/15213
< bitcoin-git>
bitcoin/master f0c9e1c Wladimir J. van der Laan: Merge #14906: refactor: Make explicit CMutableTransaction -> CTransaction conversion....
< bitcoin-git>
[bitcoin] harding opened pull request #15223: Doc: add information about security to the JSON-RPC doc (master...2019-01-rpc-security) https://github.com/bitcoin/bitcoin/pull/15223
< bitcoin-git>
[bitcoin] sipa opened pull request #15224: Add RNG strengthening (10ms once every minute) (master...201901_rand_strengthen) https://github.com/bitcoin/bitcoin/pull/15224
< midnightmagic>
says they removed it to fix crashes when using the routines under the vs debugger.
< midnightmagic>
Interesting. There's a reference to a closed tracker in their pull 512: "See also https://rt.openssl.org/Ticket/Display.html?id=2485 Probably are some other ones related." But it requires a log in, and I see no way to register.
< arubi>
guest\guest
< midnightmagic>
arubi: thanks
< arubi>
np :)
< gmaxwell>
sipa: We've called randscreen explicitly, did they gimp that too?
< sipa>
RAND_screen still exists, but it's a compatibility API that just calls RAND_poll now (which uses the standard entropy sources, and feeds them into the RNG state)
< sipa>
the actual screen reading code was removed
< gmaxwell>
Lameo.
< sipa>
nothing prevents us from copying their old screen reading code... but there are probably good reasons why it was removed
< sipa>
(causing debugger crashes sounds like one)
< gmaxwell>
I wish the *BSDs had /proc/ like linux, the proc files with interupt/network/disk stats are very high entropy.
< echeveria>
realistically the network messages of bitcoind are probably fairly high entropy as well.
< echeveria>
not that you’re ever using entropy at any time other than wallet init I guess, which is before networking would be up.
< sipa>
echeveria: creating a new seed is really the only time you need high quality entropy
< sipa>
hd wallet master, i mean
< sipa>
though we have a bunch of thing with various degree of impact if seeded insecurely (like the addrmam randomizer, which if under control of an attacker could probably simplify eclipse attacks)
< bitcoin-git>
[bitcoin] achow101 opened pull request #15225: GUI: Change the receive button to respond to keypool state changing (master...receive-button) https://github.com/bitcoin/bitcoin/pull/15225