[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Regestry Identifiers
- Date: Tue, 19 Nov 1996 14:15:00 +0900
- From: "David R. Conrad" <davidc@apnic.net>
- Subject: Re: Regestry Identifiers
Michael,
>Domain names are not allocated. IP addresses consist of all possible
>combinations of four 8 bit integers. The addresses are allocated from that
>fixed pool in a roughly sequential manner. Domain names are not severely
>limited in length and are not allocated sequentially by some managing
>authority. Instead, the domain name is chosen by the registrant.
I do not believe it is either fruitful or interesting to get into
arguments over the definition of "allocation" (although I was not
aware quantity related to whether a resource was allocated or not).
Suffice it to say that there are organizations which are responsible
for changing the status of either IP addresses or domain names from
"free (as in 'not in use')" to "in use (as in 'not in the free pool')"
and that I feel that these organizations, while their functionalities
are distinct, can share a common database schema and data format in
order to facilitate interoperability.
>Thin edge of the wedge. Either they are related or they are not. If they
>are related then a whole host of issues that should be dealt with
>separately will end up being mixed together in both groups.
This is like saying that since your electric bill and your credit card
bill both show up on pieces of paper, the discussions of how
electricity will be metered and how your credit card company will
adjust interest rates will get mixed together.
Perhaps I'm being silly, but I think it would be nice for the
registration of Internet address objects, autonomous system number
objects, contact information objects, route objects, and yes, even
domain name objects to share a common format that allows them to be
transferable among registries if for no other reason that the people
who register domain names are often the same people who register other
objects.
The reason I have suggested IRE kick off a discussion on database
schemas and data formats is that historically, the people involved in
running address allocation (by whatever definition you choose)
services are the ones who wrote the database software to provide the
registration service. I would have preferred an independent BOF for
this, but apparently it is impossible to schedule for the San Jose
IETF. However, given the database stuff is unrelated to the evolution
of Internet address allocation policies (which is what IRE is
primarily discussing), it is unlikely future IRE (or whatever it is
called) working group meetings will pursue database issues.
Don't worry, should the IRE mailing list agree to discuss the database
stuff in San Jose, I'll be sure to make an announcement before we
start so people whose sensibilities might be offended can leave the
room.
Regards,
-drc