>>>>> On Sun, 3 Jun 2001 19:38:46 -0700,
>>>>> Steve Deering <[EMAIL PROTECTED]> said:
>> And I know, that since an anycast address looks syntactically just like
>> a unicast address, this can be tricky at times...
> Part of the job of assigning an anycast address to a node is informing
> that node that that address is an anycast address, i.e., an address
> that is potentially shared with other nodes. That ensures that the
> node responsible for choosing the source address has the info needed
> to distinguish its own unicast addresses from its own anycast addresses,
> despite the lack of any syntactic indication in the addresses themselves.
That's correct, but from an implementor's point of view, anycast
addresses are more problematic than multicast, because there is no
standard API to see if a given address is an anycast address on the
node nor to see if a received packet is anycasted. We may have to
define a standard (portable) way as a part of the advanced API (bis).
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]
--------------------------------------------------------------------
