Experimental Physics and
Industrial Control System

Gerrit Kühn via Tech-talk <tech-talk at aps.anl.gov> · Fri, 3 Nov 2023 16:37:29 +0100

Am Fri, 3 Nov 2023 12:37:35 +0000
schrieb Simon Rose <Simon.Rose at ess.eu>:

> The INP link is meant to be the record that the stringin record fetches
> data from, which it would then  write into the VAL field. You can see
> this by creating a second (e.g. stringout) record with the name "test",
> and then processing your stringin record at which point it will fetch
> the value from the stringout record.

Ah, great, there was my misconception, thanks!. Although I still don't
quite understand what a "constant" in "a database or channel access link,
or a constant" would mean in that context then. In your scenario above this
would not be a constant, right? Somehow this still feels like it does not
work as advertised in the manual.

> That said, if this is meant to be a constant string used by other
> resources, it would likely be better-suited as a stringout record;
> stringin records are for reading input from another source.

Yes, I was already thinking about that, too (but it looked like I had some
other misunderstanding that I wanted to sort out first ;-).
The string would be constant under most circumstances, but it should be
possible to change it via caput in certain (rare) cases.

Another issue I have been running into with this is the order of record
initialisation. The string in this record is needed by other records and
thus should be initialised first. But from what I have read on the web so
far I understand there is no mechanism to guarantee that within one
database. However, I probably could put records like this in a separate
database and load that one in st.cmd first?

cu
  Gerrit

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System