Michael Graff <explorer@flame.org> writes:

> 
> A while back there was talk about SACK making it into the kernel "this
> friday" which has come and gone.

I said there was supposed to be an implementation available.  I didn't
say that it would go in by then.

> Is there a status update?

Anyway, Jamshid Mahdavi has announced his implementation for NetBSD.
It needs to be ported from 1.1 to 1.2, it's not particularly
efficient, and there's a licensing problem to work out before we could
distribute it.  There's also another implementation, which I think is
in slightly better shape at the moment, for BSD/OS.

Unfortunately, both of these implementations are substantially more
complicated and less efficient than they could be, because they try to
hack around limitations in the TCP reassembly code.  My intent is to
fix this and do a better implementation (unless Jamshid is interested
in doing it).