On 7 August 2012 09:01, Pierre Ynard <[email protected]> wrote: > For information, my personal transfer rate should be minimal enough, so > I'm inclined to spare myself extra TCP or IPC sockets for the loopback > case; that wouldn't fix the multiple sender issue anyway. >
I've been thinking of implementing one of the new atomic reliable UDP protocols that Oracle have been using, it should work great for this. > > > It has been this way a long time and can be easily reproduced with > > multicast testing tools. Windows fairs better as sockets are handled > > outside of the kernel but performance is significantly worse. > > > On Solaris and other UNIX flavours this is true, however as noted > > before on Linux it is broken. > > That's interesting. Do you have any pointers to discussions about the > roots of this? All I could find was stuff related to the basics I > mentioned earlier. > > It's on LKML somewhere, expect some minimal conversation along the lines of: you shouldn't be doing that. -- Steve-o
_______________________________________________ zeromq-dev mailing list [email protected] http://lists.zeromq.org/mailman/listinfo/zeromq-dev
