Re: RAWL/RAWH

[Marty Kraimer <mrk@aps.anl.gov>]

It has been several weeks since a message has been sent about
adding RAWL, RAWH to the ai and ao records. Thus I am assuming
that everyone who had a strong opinion has responded.
This message attempts to summarize the responses.

The majority of the responses did not think it is a good
idea to add these fields. Perhaps the best reason is that
given by an Application Developer. The response was:

    Q. What am I supposed to put in the RAWH and RAWL fields ?
    A. You might not have to put anything in there, but then again, you might.

    Q. How do I know ?
    A. It depends on the device type you are using.

    Q. My device type is thisDeviceHere. Do I need to worry about it ?
    A. You have to ask the author of the device support.

    Q. Who's that ?
    A. < no answer >

Several other negative responses were given by device/driver developer's.
The comments indicated that RAW/RAWH are only sometimes a problem and,
in any case, a small subset of the problems. Thus adding these fields
will not be much help. A couple of responses indicated that for particular
cases they are a big help.


Several device/driver developers stated that they used the parm field
(often with INST_IO which has only a parm field) to solve their problem.
The responses also showed that the parm field needed a lot of syntax
associated with it in order to solve the problem.

Several responses stated that we need a hardware configuration tool.
This is something we have realized for a long time but have never been
able to get a consensus on what it should be or do.

As a result of the RAWL/RAWH discussion I also got a message from Ralph Lange
that gave some ideas he and several other people at BESSY came up with when
they discussed the hardware configuration problem.
This lead to several E-mail messages between them, a few LANL developers,
and a few APS developers.  Maybe this may lead to something.
I will send the last E-mail I sent on to them. No one responded
to it so that means that either everyone was getting tired or else
there was no basic disagreement.

Marty Kraimer