>>>>> On Fri, 22 Mar 2002 10:52:36 -0800 (PST),
>>>>> Tim Hartrick <[EMAIL PROTECTED]> said:
> How the unspecified address is treated by the forwarding code is orthogonal
> to how applications make use of it to specify which connections or datagrams
> are received on a socket.
Okay, and I think this also means we should separate architecture (or
protocol) issues and API issues more clearly. So, I'd propose:
in the architecture draft, clarify that the form of
<address>%<zone ID> is intended to be used for disambiguating scoped
addresses (that have ambiguity) and that the format should not be
used for global addresses or addresses that do not have scope (i.e.,
::).
(but this does not mean it prohibits an implementation -particularly
an API- from using the format for global addresses or for :: in an
implementation dependent manner.)
With this policy
> That is, I would like to be able bind a socket to any of:
> ::%<specific type>.<specific zone>
> ::%<specific type>.<any zone>
> ::%<any type>.<any zone>
all of the above are out of scope of the architecture draft (but are
not prohibited by the draft).
> <specific address>%<specific zone> (The address specifies the type)
The architecture draft only defines this format, and the usage in the
API will not contradict the semantics of the architecture.
> <specific address>%<any zone> (The address specifies the type)
This will be implemented by using 0 as <any zone>, but this is rather
an API issue and is out of scope of the architecture draft.
Does this make sense to you?
JINMEI, Tatuya
Communication Platform Lab.
Corporate R&D Center, Toshiba Corp.
[EMAIL PROTECTED]
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page: http://playground.sun.com/ipng
FTP archive: ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------
