Experimental Physics and
Industrial Control System

Benjamin Franksen <[email protected]> · Mon, 07 Feb 2000 12:40:21 +0100

Leo R. Dalesio wrote:
> 
> I would think that a soft bo with an undefined DOL that has
> been processed at initialization should be 0 and defined. If someone
> wanted the value to be something other than 0 - they could have put
> the constant 1 in the DOL field. As some of the configuration tools
> do not define a field if it is the same value as the default, an undefined
> field should be construed as a 0. Just my opinion....

The question is, of course, relevant to output records in general.

The Release Notes to R3.13 say: "Another feature is is that it is now
possible to distinguish link constants with a value of 0 from links that
have never been given a value".

I believe that the distinction between zero ("0") and undefined ("") has
been made in order to allow undefined links to be completely ignored by
the record support in contrast to constant-zero links. Link behavior
should be consistent throughout the system and IMHO undefined is not the
same as zero.

If people want to have a zero value at init, they can either set it in
the database, or, if they insist on a different *default* behavior for
the DOL links, we (or they) can set 'initial("0")' in the record's
definition of the DOL field and everything works as expected without
changing the record support.

I'd like to know if it is the general opinion that an output record that
has been processed is defined (i.e. UDF==FALSE), even when 
- no value has been put into VAL,
- DOL is undefined, and
- no readback was done by the device support.

Ben
-- 
The Notorious Neb Nesknarf

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System