Re: Summary: What to do with expired signatures

[Paul Vixie <[e-mail]>]

> ... Or are you saying that "servers should not edit zones" is undebateable?

Yes.

> ... That's why I said that dynamic update is related to zone loading.
> As far as being a "complication," dynamic update is a new process -
> meaning that there are now more ways to get data into a zone.

Some folks here have suggested, both in this thread and in threads past,
that a zone server ought to be free to elide some nodes or RRs during its
"loading" process, and still serve the resulting (implicitly) edited data.

I'm now coming to the conclusion that "zone loading" should be deposited in
the same dustbin as "zone file format."  I was shocked to see $TTL codified
since there were, even at that time, totally compliant (on the wire) DNS
implementations which had no concept of a "zone file" or "loading."  The
standard ought to talk about what's on the wire and what to do with stuff
you got off the wire and what to put on the wire (and when and why), not
implementation details like "zone file format" or "zone loading."

If we re-turn RFC 1034/5 I will strongly recommend that all descriptions
of "zone file format" be given BCP status.  A correct and complete DNS
implementation ought not have to deal with these historical card images
unless its implementor wants to offer a backward-compatibility crutch.

to unsubscribe send a message to [e-mail] with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://ops.ietf.org/lists/namedroppers/>