Re: [RAM] Re: Ramblings about "locator"

    > From: Dino Farinacci <dino@cisco.com>

    > a MAC address is certainly used to find (that means "where", and
    > "where" means location) an ethernet attached station in a L2 switched
    > network.
    > ...
    > remember OSI and what an L1 area in IS-IS was used for? To route
    > system-ids which were based on MAC addresses. So in this case, the MAC
    > was a locator.

I have a hard time figuring out what "locator" means to you all if you think
a MAC address is a locator.

To me, a name cannot be a locator (or have 'location semantics') unless it
has the property that when the thing being named moves to a new location, it
needs a new locator (since the locator says 'where' it is). MAC addresses
explicitly do not have this property; they *always* stay they same *no matter
where* you plug them in and/or move them.

Ergo, they are not locators, and have no location semantics.

As I said in my previous message:

  .. just because a name identifies a thing which is at a particular location
  (e.g. an interface), that does not mean that the name has 'location
  semantics'. To me, there has to be *information* in the name (so that e.g.
  it can pass the 'are [the things named by] these two [names] close [to each
  other]' test).

To put it another way, *everything* has a location (or a set of them, e.g.
for replicated/distributed objects), and if "locator" simply means 'name of a
thing which has a location', that's a non-useful tautology, kind of like
saying 'name of a thing which has a name'.

	Noel

_______________________________________________
RAM mailing list
RAM@iab.org
https://www1.ietf.org/mailman/listinfo/ram