Re: Invariants draft

[Willy Tarreau <[e-mail]>]

On Thu, Nov 30, 2017 at 08:46:49PM -0800, Christian Huitema wrote:
> But at the same time, I am concerned with the tension
> between "focusing on extensibility" and "ensuring interoperability". At
> one end of the spectrum, we could define nothing but extensibility
> mechanisms, and let application developers define an application
> specific transport protocol. At an other end of the spectrum, we could
> rigidly specify a protocol, its formats, and its state machine. I do
> think that the working group has to strike somewhere in the middle, a
> well specified protocol that can be easily extended, but maybe not
> infinitely extensible.

I think this draft clearly leaves a *lot* of room for extensibility,
and defining invariants helps us (and future generations) decide when
a new protocol needs to be specified because it conflicts with the
definition of what was guaranteed to be invariant. It's normal to
reach an end with a protocol when new concerns need to be addressed
that were out of the initial scope. Here in short, as long as we can
use UDP with a few fixed bits and 64-bit connection IDs, we don't
need to define a new protocol.

I personally find it well balanced.

Regards,
Willy