RE: Renaming of instances and SubInstances?

[Kevin Hawkins]

Hi Ian - It's always a sign of a good holiday when you come back for
a rest.
Would the below (wildcarding) not mean though that you had source
addresses that now became identical and you would therefore be unable to
target correctly ?? I am not sure if the '*' construct is valid here - I
can see that you would have different UID's still but you would be changing
one without the other (albeit in setup only).
I think in my implementation I will have default names 'Ouput1
Output2 ..' with UID's 'FFxxxx01 FFxxxx02..' and I will either rename based
on targeting or based on addressing the xAPp without a sub address at all
and with a UID of FFxxxx00 and then listing the sub instance name changes
in
one body - but I am not sure of the validity or usefulness of wildcarding
here. I may even check that a duplicate isn't created and send an error if
it is.
I had always envisaged that the last two digits of the ID would be
hardcoded by the developer. This would help in any hardware central profile
that was held for a device perhaps used in 'discovery' - a central
(internet) server could provide a profile for say HomeVison showing that
ACME.HomeVison.Node0:HeatReq with a UID of FF123405 meant 'output 5' on
'parallel port A' for example. Simialr to the schema server idea. The 05 on
the end of the UID would be developer fixed and mapped in a device
'profile'
and the documentation - your relays and inputs could be the same - then you
always know which terminal you are talking about - although you can choose
to name it whatever you want.
One thing I am using in my C-Bus implementation is an 'error' or
'debug' schema - these are messages that are sent out when something is
wrong, or may need examining - and could potentially be logged by a xAP
network monitor application. I personally use a severity level of 0-9 with
0
being a fatal type. Mostly I use these for debug at levels 7, 8 and 9 - and
they can be turned off with a xAP message - perhaps this (or a variant)
would be a sensible thing to standardise. Probably DEBUG and ERROR should
be
two different body types allowable with one schema. (NETWORK, HEALTH,
STATUS, MONITOR, INFO, or something).

K



> Anyway, if it helps I use targeting for all my renaming. I can rename
> multiple SubInstances at once to the same name but only one back to
> another
> name i.e. if my SubInstances were all named relay something then
rename
> 'relay*' to 'output' at once no limits. This gives me say 8
> SubInstances all
> with the name output. If I now say rename 'output' to 'watering' the
> program
> will only rename the first SubInstance and not all of them otherwise I
> would
> get into a loop I could not recover from.