|
Bernard
Aboba [mailto:bernard_aboba@hotmail.com] writes: “I’m wondering
when it was decided (& who made the decision) that if an
implementation doesn’t follow the standard, it’s the standard
that’s broken & not the implementation…” There are several issues that
have been brought up: 1. What
does RFC 3162 say about the type of the Interface-ID attribute? The current
Design Guidelines document asserts it is an 64-bit unsigned integer.
As you pointed out, this does not appear to be supported by the
text of RFC 3162. 2. Should
the Design Guidelines document recommend that implementations support the
64-bit unsigned integer type? a.
Because it is currently supported? b.
For some other reason? The question you have raised
doesn’t seem relevant to either of the two above questions. It’s not.
However, neither is the question of how “existing implementations behave”,
unless we are now basing BCPs upon the vagaries of (possibly brain-dead)
developers rather than published standards. |