| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
Network.Lossless
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This isn't really completely correct; there should be some possibility for
denial of service. The call to setContactAddr is destructive of whatever
information was there previously, but we haven't verified that we're talking to
the real contact at the time it is called.
This problem exists even in the ordinary path where the nodeinfocallback isn't
really requiring proof of anything before throwing out data.
|
|
|
|
|
|
|
|
|
|
The handshake code would flood the network with interminable handshakes.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Also, don't send anohter Online, just send duplicate packet.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* In the case that the session seems to lock up,
the packet requests could be (re-)handled without
sorting. Just in case both sides are waiting
on a missing packet.
|
|
|
|
tput - like dput, but works in any Applicative
dtrace - like trace, but takes DebugTag
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
