RFC Compliance of Click Router Elements?

[George V. Neville-Neil]

> It is usually difficult to talk about the RFC compliance of a specific
> element. For example, is DecIPTTL "RFC-compliant"? I suppose it is RFC
> 791-compliant, because it produces a valid IP packet. HOwever, it
> complies with RFC 1812 only in the context of a larger configuration.
> 
> So no, we haven't documented this compliance. I don't tend to think it
> would be useful, but maybe you can change my/our mind.
> 

Indulge me in presenting my dilemma.

I am trying to convince the local powers that be to use Click as the basis
of a project to build a router toolkit.  This is, essentially, what Click is.
Until now we have always just ported the BSD TCP/IP stack to our RTOS
(VxWorks see www.windriver.com) and then tried to keep current
with that code base.  This is plainly, to me, insufficient.  

If we are to satisfy our customers needs to be able to move pieces of code around
(such as putting bits of functionality into hardware or just adding and removing
special bits of code) then we need a framework within which to do this.  Until
I came upon Click I was going to use an internal system developed for PPP
(much like NetGraph in FreeBSD).  Click IS the right answer but companies
find it hard to change their ways.  The current objection is that the Kame
folks (who we're tracking for IPv6 code) have already validated their
stack with respect to the RFCs for IPv6.  This is a huge issue for us in dealing with 
European companies who come along with lists of RFCs before buying 
your software.

I know that it's kind of a dumb way to build/buy software unless you have 
a validation suite but that's the way it is.

So, in order to convince the powers that be I have to find a way to validate
the IPv6 elements (or a collection of them) against the IPv6 RFCs so we can
see what effort it will take to fulfill the RFC requirements.

I have now found/downloaded the TAHI suites (www.tahi.org) and will try to
run those against Click.

If someone could send along what they think is a good, representative, IPv6 
router configuration (I've seen the examples but I'd like some advice
from folks) that would also be a great help.

I think this should be part of the Click project eventually (perhaps I can
even contribute this if I get the time to do it) so that it can be part of
the selling point of Click to other companies and projects.  The more
people we get on board with Click the more elements we get, etc.  The
usual Open Source model.

If someone has done this already it would be awesome (hence my plea).

Thanks,
George