to be selected by a taker's sweeping coinjoin.
This is done by approximating the cjamount ceiling and filtering to one offer per
maker before the actual selection algorithm runs.
If a user doing sendpayment decided to reject the offered fees,
before this commit this resulted in a crash because the call
to Taker.on_finished_callback was passed with fromtx=True for
a single line schedule, which attempts to restart. fromtx should
have been set to False when it is the last item in the schedule
(in this case, both first and last). This incorrect argument
existed before restart-with-subset but was then harmless, now
it attempts to recover with no Taker.nonrespondants and crashes.
Fix is simply to provide the right second argument (False) to
the function call.
b741b24 Allow sendpayment to restart with honest makers If N makers are chosen and M fail to respond with sig, this alteration to taker code allows restart of entire transaction with that specific subset of makers (N-M) that originally responded honestly. (Adam Gibson)
If N makers are chosen and M fail to respond with sig,
this alteration to taker code allows restart of
entire transaction with that specific subset of makers (N-M)
that originally responded honestly.
Previously, if a bot reconnected and encountered a nick
collision, it would append '_' and connect, but counterparties
would ignore appended characters after NICK_MAX_ENCODED+2, and
so would send to the other nick. This happens in network
connection failure scenarios.
Strategy here is to simply insist on regaining the nick on that
message channel where it has been lost, retrying every 10s.
There is also a loud warning message printed.
This fixes the non-deterministic failures of the test_schedule tests.
It does so by addressing the underlying issue, which is described in
be made configurable in future, but assessed as not worth it here.
It will end up showing very small and actually wrong numbers like
5.38939837311e-11 % or even -1.48880395563e-11 % if it's newly funded
wallet with no coinjoins happened yet.
Prior to this commit, if a user chose to reject offers, or the offers
choosing process failed for another reason, the on_finished_callback
of the Taker was called, as well as the return value passed to the
client protocol being "False", and both execution paths called
reactor.stop(), resulting in an Exception in twisted (a reactor can
only be stopped once). This commit fixes that by only stopping the
reactor in the client protocol code (and only in cases of 1
sendpayment; for multiple tx runs, it waits and restarts).
4e6bee3 choose correct offer types for non-sw taker (AdamISZ)
619b4a1 Add support for taker using a non-segwit wallet As for maker, to use this feature set segwit = false in the POLICY section of the config file. This commit does *not* include support in the GUI. Additionally, minor fixes to utility scripts for non-sw wallets, and updated add-utxo tool to support non-sw wallets. Bugfixes to wallet_utils to support non-sw wallets and signmessage. (AdamISZ)
As for maker, to use this feature set segwit = false
in the POLICY section of the config file.
This commit does *not* include support in the GUI.
Additionally, minor fixes to utility scripts for non-sw wallets,
and updated add-utxo tool to support non-sw wallets.
Bugfixes to wallet_utils to support non-sw wallets
and signmessage.
605eebe add test cases for maker::verify_unsigned_tx (undeath)
d2196ad fix maker crash when verifying transaction with p2sh output in non-sw mode (undeath)
715985d add support for non-segwit maker (undeath)
This comment was formerly housed at another repo, which has now
been made inactive and the comment was removed. Note that this
manual process will be superceded soon by a change to the
installation script.
Alex Cato
fivepiece
AdamISZ
undeath
Kristaps Kaupe
Adlai Chandrasekhar
James Hilliard