Adam Roach wrote:
You'll note that this work is coming out of GEOPRIV -- where
the "PRIV" part of that name stands for "privacy."
That's not more obvious in the "registry" draft. Joe Abley
posted a pointer to draft-ietf-simple-rpid-08 for the missing
context, and there I found lists of locations and moods. The
latter appears to be a list of smileys (without the smileys),
and it's clear that the user picks what his or her mood is -
a device can't know this, and it better doesn't try to guess
the mood of its user.
For the locations it's not so clear. Why should a device tell
me that I'm in the "cafe" of a "jail" in an "airport", I would
know this. And why would I tell my device where it is, all it
needs to know is "be quiet" or "make as much noise as you can",
different ways to attract my attention depending on where I am.
There are many good use cases in real-time communications
that detail *why* conveying this kind of information is
potentially useful. Going into a tutorial on these basic
topics seems a bit inappropriate on the general list.
Maybe the "registry" draft could be integrated into a draft
where that's immediately clear, or bound to a list of special
purposes where devices need to know why "cycle" and "water"
are different - at first glance both could fall into a category
"now isn't an optimal time to talk, but it's not impossible".
Without context a "location dictionary" as Internet standard
appears to be an odd idea. And a list of smileys without the
smileys removes the fun - if the moods get their own registry.
Ietf mailing list