Re: [ovs-dev] Windows datapath interface

2014-07-22 Thread Alessandro Pilotti
Hi Ben, On 22 Jul 2014, at 20:10, Ben Pfaff wrote: > After some pondering, I've concluded that the initial Windows datapath > should use Netlink for userspace-kernel datapath communication, > primarily because Netlink is extensible in a forward- and > backward-compatible fashion. That is, as we

[ovs-dev] Windows datapath interface

2014-07-22 Thread Ben Pfaff
After some pondering, I've concluded that the initial Windows datapath should use Netlink for userspace-kernel datapath communication, primarily because Netlink is extensible in a forward- and backward-compatible fashion. That is, as we add new features to the kernel module, existing userspace con