RE: Communication means in draft-ietf-impp-cpim-pidf-05

> The example it meant to illustrate that pidf, as it stands, does
> not carry any useful information to the watcher if only the
> minimum mandatory elements as they are now are sent in the pidf document.
>
> If I, as a watcher, receive a pidf document that looks like the
> example I included in earlier emails, I have no means of
> understanding or using this information in a useful way. What
> does it mean to have 2 tuples, one with ID 342jrmeklrm and status
> open, and the other with ID jklejr983 and status closed.

In this particular example, it's not hard to believe that there could be
some external context allowing the watcher to determine a suitable meaning.
This could be the presence state of the doors on a 2-car garage, where the
left-hand door always has a lower tuple ID. Once you know that context, this
doesn't seem useless at all.

Is this nonsensical for many other uses of presence? Yes. But in general,
we'd rather lean toward enabling expression rather than outlawing nonsense.
It's often hard to outlaw nonsense without also eliminating useful but
non-obvious expressions.

I would also observe that the optional vs. mandatory aspects of 2778 were
argued at great length, and it has been a long time since I encountered a
novel argument about the choices made. Broadly speaking, we as a working
group decided to follow 2778/2779 very closely so as to avoid rehashing all
of those discussions again.

--Mark




  [reminder: [email protected] for non-technical discussions, please]