Kacheong Poon writes:
> Just wondering, does it make sense to have a "send" function
> which takes both source and destination addresses as arguments
> instead of using ancillary data?

Exploding the list of send*() variants didn't make sense to me when we
have something easily extensible like ancillary data, and when
effective extension of the function call interface would likely
include running the standards gauntlet.  (E.g., 3XNET would give you
ancillary data, but not "pollution" from any non-standard interfaces,
and 3SOCKET would give you the new feature but not ancillary data.)

Plus, it's already implemented as ancillary data on IPv6, so it seems
odd to do it so very much differently for IPv4.

>  To be symmetric, does it
> make sense to have a "receive" function to return both source
> and destination addresses?  I guess an app can have two wrapper
> functions which make use of ancillary data to do that.  But
> if we think that they are useful, we may as well provide them.

We already have receive functions that provide source address.

I suppose we could invent convenience wrappers for ancillary data, but
I'm not sure that's a necessary part of the project.

-- 
James Carlson, KISS Network                    <[EMAIL PROTECTED]>
Sun Microsystems / 1 Network Drive         71.232W   Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757   42.496N   Fax +1 781 442 1677
_______________________________________________
networking-discuss mailing list
[email protected]

Reply via email to