In your previous mail you wrote:
> we really need to convince random book authors to update their
> programming examples, from gethostbyname() to getnameinfo()...
how can you expect them do so, when there's so much difference between
the behaviour of the various TCP/IPv6 stacks?
=> this discussion is about to find a way to solve difference between
compliant stacks...
HoraPe has explained us that on *BSD systems, you can have two different
sockets (one for AF_INET and one for AF_INET6) bound on the same port.
in this way you can use an af-indipendent programming style.
=> you can use it with the V6ONLY stuff.
this is not true with linux. you have to use a single AF_INET6 socket
and listen on both incoming ipv4 and ipv6 connections.
=> this is the standard way but with the V6ONLY way you can use the
single socket (my favorite) or the socket per IP version (Itojun's
favorite). The user shall choice his own favorite or the one which
matches the application constraints.
draft-ietf-ipngwg-rfc2553bis-03.txt does not specify a specific behaviour
for bind. it's left to the implementors. i'd really like to know the
reason for doing so. IMVHO, this is a problem, because it will lead
developers to write platform-specific code and technical writers to
produce misinformation.
=> oh no! Read the mailing list archive and don't reopen this discussion!
(the conclusion of the discussion is against your proposal which is not
as new as you seem to believe)
however, as you have already said, the real problem is that there is
not a "Unix Network Programming"-like reference book covering ipv6
socket programming.
=> there is one but it is too old and the author is dead...
Regards
[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]
--------------------------------------------------------------------
