Complex Problems

[Harald Tveit Alvestrand]

--On fredag, januar 10, 2003 09:20:24 +0200 Jari Arkko 
<jari.arkko at piuha.net> wrote:

> If we want to emphasize the need for early architectural advice we may
> want to think about splitting our documents into "Architecture for X"
> and "The X Protocol" documents, and get them approved separately. The
> architecture document would be done early of course, and would contain
> the major design decisions such as what the protocol entities are, on
> top of which protocol are we running, do we use hbh or e2e security,
> etc. But all the protocol details such as message formats and behaviour
> rules would be in another document.

Quite reasonable, and has been tried.

For one example of what happens if you try that, consider 
draft-ietf-opes-architecture-04.txt.

It describes the problem space architecture reasonably well, but does not 
even specify the protocols to be used to the level of saying whether they 
are message-based or RPC-based, far less stuff like the naming of endpoints.

The OPES folks seem to think that this particular level of architecture 
description is useful to them (they asked the IESG to publish the document).
But there's still plenty of room for "architectural gotchas" in the next 
steps.

For other examples, search the I-D directories for "-arch"; there are a few.
This one came to mind only because I read it in December...

                  Harald