network-transport-tcp
network-transport-tcp copied to clipboard
Add TCP_QUICKACK setting
See the commit messages for details.
I could run the normal tests, but not the mock tests since they don't compile on #master (see #65).
Related downstream PR: https://github.com/serokell/network-transport-tcp/pull/1
Also note I have not myself compiled the non-Linux side of the #ifdef
(but that will likely be done as part of testing of the downstream PR https://github.com/serokell/network-transport-tcp/pull/1).
BTW, regarding
This change breaks the external API, because
socketToEndPoint
(for reasons not obvious to me) takes all socket options as individual arguments, instead of as a settings object.
If you like, I can make PR to change this, but first want to know if people would have anything against it.
This PR LGTM.
Probably there is not a good reason for socketToEndPoint
to get every option as a separate parameter. It is fine with me if you want to change it.
Could we check isSupportedSocketOption
when the transport is created to ensure that if TCP_QUICKACK is used, it is supported?