On Mar 17, 2016 12:09, "Tassos Chatzithomaoglou" <ach...@forthnet.gr> wrote: > I consider DHCPv4 one of the most-used protocols that will stay for a long time with us, because it's not directly affected by what happens on the global internet.
And, as with IPv4 itself, declaring v4 historic does nothing to prevent people from using dhcpv4 as long as they require it. It simply means the IETF will no longer work on the protocol. DHCP already includes a way to define custom options if users of the historic IPv4 protocol decide to define and use a new option. I'm not sure I see a need for the standards body to spend time on it. No strong objection either if it's something enough people want to spend time and energy doing, but since DHCPv4 is tightly coupled with IPv4, it seems reasonable that it be considered historic as well. > I can even imagine whole "islands" staying on IPv4 and using it for many decades. >From an enterprise perspective that seems incredibly foolish. It's actually at the edge that v4 is required for the long haul. Having begun the transition, the sooner most of our enterprise is on one protocol again the better. > Assuming IPv4 is declared as historic, does this mean that no other options are expected to be defined for DHCP? For DHCPv4, that seems reasonable. New work should focus on DHCPv6, which is a pretty different protocol. But again, I don't feel particularly strongly either way. > Do we assume that any new applications won't need those? Part of making v4 historic means that the IETF is only committed to ensuring new protocols work with v6, at least if I've understood the definition correctly. New stuff might still work over v4, but it becomes a case of caveat emptor. > Maybe a note should be added about the exclusion of creating new options for such cases. So the actual protocol standard won't see any development, but protocol "extensions" through standard-defined options may happen. It is an interesting question how a protocol that's as tightly coupled with IPv4 as DHCPv4 is impacted by making IPv4 historic. I don't recall seeing anything that directly addressed that question. Scott
_______________________________________________ sunset4 mailing list sunset4@ietf.org https://www.ietf.org/mailman/listinfo/sunset4
