Re: X10 Schema and Preset Dim

[Gerry Duprey <gerry@xxxxxxxxxxx>]

Howdy,

> That seems like an odd abstraction of the X10 protocol. If the
> underlying hardware is suppose to pick the correct then what command
> would you put in the xPL message, just a dim? If that is the case why

Well, not so much the hardware (i.e. an X10 switch) as the xPL gateway to
the hardware.  Some hardware does this already (HomeVision) and others
don't
and require assistance by the X10->xPL gateway.

Thing is, some X10 devices have 16 dim steps, some have 32 dim steps and
some have 63 dim steps (levitons). Some can go directly to a level and
others can only do relative dimming. So it has always seemed logical that
someone controlling an X10 device via xPL (where you are removed and not
directly in contact with the hardware) would specify a level 0-100% and the
xPL gateway and/or hardware would be able to pick the closest possible dim
level for the target device.

> does the X10 schema document not describe that and why are the preset
> dim commands in the schema. I think for an X10 schema it should
reflect
> the actual X10 protocol, if it was going to be abstracted then (in my
> opinion) the schema shouldn't have been called X10.

Well, I can see two camps here -- the "just make xPL an virtual
wire" where
you are expect to handled the various X10 vagaries on each xPL
"client" and
the "let the gateway handle the fiddly low-level stuff" camp.
Short of
coming up with some sort of extensive and (I believe) confusing way for a
remote xPL client to "know" what and how to dim each type of X10
device (and
which X10 device address is what type of device), the former isn't
workable.

We're not talking about a lot of abstraction here -- just enough to unify
the areas that have different implementations on different devices (which,
believe is pretty much dimming and scene control -- everything else is
pretty well understood and consistent).

In fact, I'd go so far as to say that direct predim1 and predim2 aren't
standard X10 because not all X10 devices support them, so exposing is again
an issue of device compatibility (I'd vote for them being removed from the
X10.basic schema is I were a deity :-).

> Some X10 devices use
> the preset dim commands to invoke scenes, where would I put the preset
> dim level for those instances? If the level= value must be 0-100% then
> it would be a violation of the schema to use that name/value pair,
> wouldn't it?

Yep and that probably argues for an X10 extension of X10.basic that knows
about scenes and such and again, handles figuring out how to invoke a given
scene on a given device (again, scene activation differs from X10 device to
X10 device) and keeping that "mapping" on the adapter side so a
client can
just say "Activate scene 1" and not worry about the 3 different
ways it
would have to send that (depending on how varied the hardware on an X10
network is).

> I know nobody likes to document (me included) but with all this
> knowledge of the intricacies of the schema just in peoples heads it
> makes it very hard to get started using xPL.

I agree -- it can be very frustrating.  On the upside, there is a list
where
questions can be asked :-)  Looking at the schema, it does say levels are
0-100, though other issues that you mention are not addressed.

One complaint about many volunteer project is the quality of the
documentation (or lack there of).  Very few people like to write
documentation at all.  Not that explains things, but the schema documents
to
date are more on the "minimal" descriptive side and could stand
some
clarification.  A good way would be to grab the doc and update it the way
you think it should and send it to the list for proofing and once it gets a
few nods, updating of the web page with that schema.

Another thing, which I know you're aware of because of our off-list email
is
that there is an effort underway to come up with a machine readable xPL
schema.  That would not only make schemas more parsable, but would help to
more explicitly codify the schema and related documentation (the XML files
will be translatable into HTML, human readable docs so everything
originates
from an authoritative source.

Gerry
--
Gerry Duprey
Ann Arbor, MI 48103
http://www.cdp1802.org