> Historically, what I had (VCTCA and IUCV connections), was P2P. With P2P > you don't have a router address nor do you have a broadcast address. Just > wasn't needed.
Well, you do have a router address; it's just the other end of the link. The presence of broadcast depends on the type of media. Typically, P2P links have no "all stations" capability (in the comms world, P2P links are usually serial lines with only two ends), so there's no need for broadcast. > And I assume the reason why Linux shows me a netmask of > 255.255.255.255 for P2P connections is there is some code, No, there's only one host on the other end of the link, so you don't actually have a subnet on a P2P link. The key thing to remember about IP routing is that it's done on a hop-by-hop basis. The only thing an IP stack knows is that to get to a destination, you take x.y.z.q as the next hop along the way. Each node makes independent decisions, and subnet masks are just a way of saying that I can reach a group of a certain size via this next destination. The VM stack has always done subnet masks in a particularly arcane manner, but the basic rules are the same. > And then with z/VM 5.2, and the major update to TCP/IP (must be a major > update in order to change the syntax of all the control cards), the > "special" was of handling P2P connections was dropped (or got broken). > And the solution is to get on board the legit way of specifying a P2P > connection. Earlier versions of the VM stack allowed defining links with the same IP address. 5.2 is finally enforcing the spec (as it should). In all fairness, Chuckie gave fair warning long ago that the official answer was that each link had to have a unique address and the short cut would stop working at some point. Now it has.
