RE: OMS for motor record

[Mark Rivers <RIVERS@BNLX26.NSLS.BNL.GOV>]

Jim writes:
>This is just an observation concerning Joe's and Mark's mail about the
>OMS motor driver for Tim's motor record.  I find it interesting that
>both of you have been studying the code and doing timing measurements without
>contacting me at all.  I must have something of value to add, it could
>at least same a few hours - especially since the send/receive times have
>already been measured.  I would have at least expected a question such
>as 'why in the heck did you design this thing to work like this?'

When I started this discussion, I honestly did not know who wrote it - there
are NO comments which document the author, revision dates, etc., etc. in the
source code for drvOMS.c (at least in the version which Tim Mooney distributes
to CATs). recMotor.c lists you as the original author and Tim Mooney as the
current author.

As I understand it, Joe Sullivan will be taking over maintaining this piece of
code. I hope you can fill him in on the timing measurements you have done and
the background behind the design. In particular, the rational for abandoning
the Los Alamos interrupt driven design. Is this documented anywhere?

I think the motor record is probably going to be the single most important
record for CATs, so it is worth doing it right. This effort can only be aided
by having input from as many sources as possible. We'd appreciated anything
you can offer Jim. Sorry for the oversight.

Mark Rivers