Open maflcko opened 2 weeks ago
An alternative code fix would be to ignore
ACK
after anACK hash
. However, this may interfere with the stale-ACK detection, so it should only kick in whenACK hash
is not stale.
Fixed it in https://github.com/maflcko/DrahtBot/commit/cf46398f045566b28748adac92369b5d0383eb86.
This also means that a ACK hash
, followed by NACK
may or may not be misdetected as ACK
.
Thanks for fixing!
This also means that a
ACK hash
, followed byNACK
may or may not be misdetected asACK
.
This doesn't necessarily sound too bad if the PR changed and it shows up a stale ACK. Maybe worse in some other situations, though, but it is hard to be perfect.
Parsing will never get 100% accuracy, unless everyone is forced to use an exact format.
Here is a collection of parsing issues.
Misdetection
HACK
https://github.com/maflcko/DrahtBot/issues/26#issuecomment-1370710327not a NACK
,retract my NACK
https://github.com/bitcoin/bitcoin/pull/26076#issuecomment-1294093144, https://github.com/bitcoin/bitcoin/pull/26525#issuecomment-1324400724They should be fixed by providing a fresh review, or by downvoting the DrahtBot comment. ("If your review is incorrectly listed, please react with π to this comment and the bot will ignore it on the next update.")
Review going stale
ACK hash
, thenACK
https://github.com/bitcoin/bitcoin/pull/30141#issuecomment-2206072773Workarounds:
ACK
, orAn alternative code fix would be to ignore
ACK
after anACK hash
. However, this may interfere with the stale-ACK detection, so it should only kick in whenACK hash
is not stale.