Experimental Physics and
Industrial Control System

"Chestnut, Ronald P." <[email protected]> · Mon, 8 Mar 2004 10:42:08 -0800

Ned,

I'm glad you're doing this!

At SLAC we have implemented many Active-X Labview interfaces to EPICS. One of the operators took over support for this, for which we are very happy. We are very much looking forward to moving to the 3.14 shared memory interface. The drawbacks to the Active-X implementation which we have seen include:

1) Adding more PV support seems to fail poorly, i.e. N records work OK and N+1 are really flaky.
2) The PVs are fairly brain dead, i.e. you just have a VAL without more work.
3) As a result of (2), we often "mirror" the Active-X PVs through a real IOC, so that display limits, state severities, etc. can be used.
4) Having only one task able to serve CA makes complex labview apps difficult to write.

In looking forward to the 3.14 shared memory interface I do realize that one interesting problem is the agreement between Labview variables and EPICS PVs with respect to address mapping, type, poll rates, etc. Being able to look at the labview "program" and extract some basic PV information would be a good thing. 

/Ron Chestnut

-----Original Message-----
From: Ned Arnold [mailto:[email protected]] 
Sent: Monday, March 08, 2004 8:11 AM
To: APS tech-talk
Subject: [Fwd: LabView/EPICS]

I recently had a discussion with the Product Manager of LabView Realtime/Embedded about possibly implementing a tightly-coupled EPICS interface to LabView. They have heard alot about EPICS from their customers and would like to investigate what would be involved in 
providing this capability in an integrated way.

The first question for us to answer is "what do we want?". So, I would like to hear people's experience with the EPICS/LabView interfaces that are currently available. I know of two major ones:

   - Active-X interfaces (for clients and server)
   - Shared-memory interface to standard EPICS records (3.14)

Please contribute your experiences, comments, ideas, wishes, etc with these solutions or let us know of any others that have been done. For 
those of you familiar with LabView, what other 
capabilities/features/performance requirements would you like to see?

The goal would be to develop a list of requirements for the "out-of-the-box" LabView/EPICS solution.

Please reply by March 12.

      Ned

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System