Re: Netiom-xAP

[Kevin Hawkins <lists@xxxxxxxxxxxxxxxxx>]

David Pullen wrote:

>Rob
>I understand your problem and I think it is partly a result of the fact
>that we are dealing with stuff that is still developing. I take a
simple
>(primitive!) approach. I try a BSC message (and play around with the
>formatting just in case that is an issue). If that doesn't work then I
>try a device specific command. Actually there is a bit more
intelligence
>at work in that I do think about whether is it is  more likely to be
>suited to a BSC message or not. In this case I did start with a BSC
>message but couldn't find one that worked and so swapped over to the
>netiom class.
>
>
Interested to know if you've found a BSC message that  should work that
seems not to David, I think I have most working here...

Rob, really BSC is the 'realtime' interface as it issues events when
things change. The Netiom schema is more a query/response schema. One or
other is usually more approriate to the way you want to interact with
Netiom.  Also the Netiom schema offers a nice way of getting all the
inputs/outputs etc grouped as one block/category  whereas BSC always
sends each node as an individual message. So if you were to ask for the
state of all I/O via BSC you will get 102  messages cf the categorised
half a dozen or so via the Netiom schema.  If you want just one node BSC
will respond just with one (short) message, or you can select groups
based on wildcards of the node addresses  eg   Phaedrus.Netiom.machine:
inputs.* or  Phaedrus.Netiom.machine: lights.> or similar.  BSC also
interfaces in a plug and play way with other BSC applications so they
always are in synch with current status and know the capabilities and
how to control the I/O on xAP Netiom.


>There are still a few teething problems with the Netiom and message
>formats. For example, it is case sensitive with alphabetic hex
>characters in the ID.
>
:-(  oops that may be my fault - I notice that I haven't stipulated in
the BSC specification that ID should be uppercase only :-(  - I had
assumed because it matches against the UID field which is defined in the
xAP main spec as uppercase only that they would always be presented as
uppercase.  I guess an ideal implementation should ucase() the ID
field.      It's always tempting to use ID=* too but that causes Netiom
to have to check far more permutations so it is slower and to be avoided
if possible.

> I know that Roy is going to be fixing this.
>Unfortunately it causes a problem with Homeseer in that outputs 10 to
15
>(ID 1A to 1F) cannot be controlled from simple device control in HS as
>the BSC messages that are generated are not recognised.
>The good news is that once you have worked out a message format that
>works you don't have to worry about it again!
>
>
Does Netiom expect lowercase then in ID and HS issues uppercase  ?  -
unfortunate.....   missed that one in the beta testing.

Kevin