optimize numInbound count #960
Conversation
| @@ -0,0 +1,35 @@ | |||
| package identify | |||
There was a problem hiding this comment.
For posterity. why'd we move this to a new file/what is a reader supposed to think when distinguishing between obsaddr_test.go and observed_addr_test.go. It seems like these names are too similar.
There was a problem hiding this comment.
This needed access to private types so I moved it from the identify_test package to the identify package.
There was a problem hiding this comment.
seems reasonable. nbd, but maybe we should name the files differently?
| // Don't trump an outbound observation with an inbound | ||
| // one. |
There was a problem hiding this comment.
Do you mean the opposite? You're saying to mark as inbound if it was, or is currently inbound.
There was a problem hiding this comment.
I think the wording here is just throwing me off (maybe just tired 😃). Maybe something like "Observation is considered inbound if the peer ever gave it to us as an inbound observation"
|
Seems reasonable, although we need to figure out why CI is failing here |
We don't use and/or expose these anyways. Making them private makes it easier to reason about their state.
We prefer addresses from inbound connections. We don't want outbound connections to hide these perfectly good addresses.
We call this _very_ frequently when computing our local addresses.
|
Well, the comment above the failure is:
Ant it appears to be very rare. |
We call this very frequently when computing our local addresses.
This PR also:
Review commit by commit.