RE: TISPAN proposal for PPP IPCP extension to handleP-CSCFaddress???

Jerome Moisand writes:
> All right, DHCPINFORM would work, sounds a bit contorted to mix PPP and
> DHCP, but why not.

It doesn't really sound contorted to me.  In fact, it has a history.
Before PPP was deployed, people used BOOTP over SLIP to handle these
sorts of application configuration issues.  The RFC 1877 feature is
ahistorical because it ignores that practice.

The implication of the bolt-it-onto-IPCP approach is that we must have
an IPCP extension for each defined DHCP option.  Otherwise, we've got
a poor step-child on our hands, as DHCP will always be more capable.
Such an effort seems to me to be quite wasteful, and likely confusing
as well.  (On a multihomed machine, for instance, how do you set
policy for DHCP-like information that may arrive via different
protocols?)

It's not really a question of whether it "would work."  It's a
question of whether the protocols are being used as they're designed
to be used (DHCP is the standard for distributing network-related
configuration options of this sort), and whether the proposed new
feature (IPCP extension) is needed.

In this case, the existing protocols solve the job nicely, and the
newly-proposed feature has no notable benefits over the existing
protocols.

(The sole exception is apparently among the folks who think that it's
easier for them to add a brand-new PPP extension that all other
vendors will eventually be forced to implement, rather than send and
receive a single UDP datagram in an already-defined format that needs
no additional standardization work.  It's that false economy that
causes this sort of proposal to come up time and again.)

(Why not bolt SIP addresses onto ARP?  After all, on non-PPP media,
you'll need an equivalent solution to avoid using DHCP.)

-- 
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

_______________________________________________
Pppext mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/pppext