In <7978105(_at_)pamho(_dot_)net> "Roger Moser"
<Roger(_dot_)Moser(_at_)pamho(_dot_)net> writes:
Colons are not allowed in hostnames, but they are allowed in domain
names. Try "dig FFFF::1.midwestcs.com txt +short".
Yes, it works. But still inet_ntop() or similar should not be used for
creating a domain name because its result can have several different forms
(e.g. "0:0:0:0:0:0:FFFF:102:304", "0:0:0:0:0:0:FFFF:1.2.3.4",
"::FFFF:102:304", or "::FFFF:1.2.3.4").
Good point. I guess the format needs to be nailed down in the spec.
Of course, there are issues with DNS lookups for the email addresses
also, because each component of a DNS address (the parts between dots)
can't be longer than 63 characters, and the local part of an email
address can be at least 64 characters. *sigh*.
Also, is foo=%{i} a domain name, or a string?
If "foo=%{i}" is the whole directive, then it is an unknown modifier.
It is an unknown modifier, which is perfectly ok. I can see times
when modifiers should be domain names and times when they should be
arbitrary strings.
What would "exists:%{c}.foo.example.com" do when you are using IPv4?
The same as "exists:%{i}.foo.example.com".
Then you are saying that for the most part we should replace all uses
of %{i} with %{c}?
-wayne