Re: 3. WoT Thing needs to have meta Band

Sure.

(1)

I'm not proposing that computers declare themselves as heaters, just that
it's possible for you to do so.

The _real_ reason you need "facet flexibility" outlined with the WeMo
example, that is e.g., that you might want to declare your WeMo controls a
light - or a space heater.

(2 - aside)

Even you're not convinced that things like facets should be dynamically
editable, at least I hope I've convinced you that you might want to have
more than one facet (or "type" in the original WoT Thing definition)
associated with a Thing.

(3)

I guess the other principle I should be outlining is "no miracles". The
client has to make a decision about how to use it as a heater.

Going with our joke example of the computer heater, you can use facets in
the metadata to find that this device declares it's part of the heating
system (e.g. it is "iot-facet:climate.heating").

Now that you know that, the client still has to make a decision about how
to use it. And let's say that's out of scope, but I hope I've convinced you
that this is flexible enough to encompass it if "turning it on" is an
insufficient answer. And you've pretty well outlined the logic, e.g. does
it have a temperature setting, etc.

D.

On Wed, Jan 27, 2016 at 11:29 AM, Lynn, James (Application Defender /
Fortify on Demand) <james.lynn@hpe.com> wrote:

> Can you clarify something regarding this example?
>
>
>
> A ‘real’ heater typically has a thermostat which can be set to some value.
> A computer, etc. does not. So they are different ‘types’ of heaters. One
> proposal seems to be saying it is useful to know that a computer can be
> used as a heater but not necessary. Another is that ‘things do what they
> say they do’. Are we proposing that a computer may declare that it is a
> heater? Or would your approach suggest that some other property lets us
> know that it generates heat, in which case I’m guessing that the lack of a
> thermostat property and/or the absence of a setTemperature property allows
> us to differentiate between ‘real’ heaters and devices that may be used as
> a heater?
>
>
>
> Or perhaps the view is that it really doesn’t matter, it’s outside the
> spec?
>
>
>
> Regards,
>
> J Lynn
>
>
>
> *From:* davidjanes@gmail.com [mailto:davidjanes@gmail.com] *On Behalf Of *David
> Janes
> *Sent:* Wednesday, January 27, 2016 5:18 AM
> *To:* Carsten Bormann
> *Cc:* t2trg@irtf.org; public-wot-ig@w3.org
> *Subject:* Re: 3. WoT Thing needs to have meta Band
>
>
>
> So this gets into one of the principals I'd like to promote, that:
>
>    - things do what they say they do
>    - things are what they say they are
>
> So if you have an old Pentium in the basement that you'd like to heat the
> room (or the student's computer pool room at Karlsruhe), you can add
> "iot-facet:climate.heating" to the metadata facets. When some client wants
> to heat up the room, they introspect the environment, find that there's a
> device that (says its) a heater and turn it on.
>
>
>
> D.
>
>
>
> On Wed, Jan 27, 2016 at 5:08 AM, Carsten Bormann <cabo@tzi.org> wrote:
>
> > "iot-facet:climate.heating",
>
> But maybe I don't need to know that something was sold as a heating?
> Switching on the computer may heat my room fine.
>
> Most people who build computers won't think of the fact that it also can
> serve as a 200 W heater.  Or TVs, lamps; ...
>
> Knowing the purpose of a thing certainly is useful, but knowing the
> actual effect may be more useful.
>
> (But maybe you can make those facets fine-granular enough that they are
> useful for searching for a particular effect.)
>
> Grüße, Carsten
>
>
>