Summary of meeting on 4/19/02

[Tim Mooney <mooney@aps.anl.gov>]

Summary of the beamline controls collaboration meeting of 4/19/02

1) Cat reports:  19 of 27 existing and prospective CATs reported
briefly on the software and techniques in use or planned.  Of those
reporting, the typical CAT runs EPICS and spec.  The predominant
computing platforms are Solaris, Linux, and Windows.  A variety of
custom EPICS-client software is in use, including clients in C, IDL,
tcl/tk, LabView, Igor, Visual Basic, and Python.  Several CATs run MX,
one exclusively.  Two CATs run spec exclusively.  Many CATs run
specific instruments (typically CCDs) with vendor supplied software.

2) Mark Rivers described the software he's developed over the past year
or so, including support for Canberra and XIA MCA hardware,
multi-element x-ray detectors; the Struck MCS; the EPID record and
support for fast (kHz) and slow (Hz) closed-loop control; support for
McClennan, ACS, and Newport motor controllers; support for IndustryPack
modules, the Bruker Smart detector, and trajectory scanning.  Most of
this software is described in the GSECARS web pages:

        http://cars9.uchicago.edu/gsecars/index.html

3) Bill Lavender described the beamline-control system, MX.  A
presentation very similar to the one he gave is available online at

       http://www.imca.aps.anl.gov/mx/meetings/ssrl_feb01/slide01.html

4) Peter Siddons described a prototype 400-element x-ray detector with
integrated signal-processing electronics, including shaping amplifiers,
discriminators, two single-channel analyzers, and three scaler channels
per detector element.

5) Eric Boucher described SciDXP (Scientific data exchange and
processing), a toolkit under development by APS and ESRF staff, which
is intended to support custom visualization and analysis of data.

6) APS developers describe their recent work.  (This topic postponed
until next meeting)

7) Discussion centering on how the APS beamline-controls community, and
the BCDA group in particular, can most usefully respond to recent
restructuring by DOE and APS.  Among the suggestions:

- produce more and better documentation, and improve access to existing
  documentation, of APS-developed software, and of EPICS development and
  use.

- provide access to a reasonably comprehensive EPICS-development system,
  so that anyone who develops software for APS users can test that
software.
  The system should also be useful for training new developers and
users.

- hold beamline-controls meetings more frequently, perhaps monthly.

- in addition to the existing software-distribution model (one big tar
  file with everything including source and compiled binaries for
Solaris,
  distributed after it's been tested on SRI-CAT beamlines) BCDA should
  separately distribute versions of top-level applications, so that
  developers can get the most recent software.  Alternatively, granting
  access to BCDA's CVS repository might serve this purpose better than
  explicit distribution.

- provide organizational support for distributing CAT developed
software,
  and leverage existing expertise among users and CAT developers in
spec,
  tcl/tk, python, etc.  (Resurrect the 'Software Exchange', maybe?)

- provide built, directly runnable copies of EPICS base and extensions,
  synApps, and other common software via existing workstations currently
  used to serve PV's to CAT beamlines.
  

That's what I was able to recollect from notes taken during the meeting.
Let me know if I missed or misrepresented anything.

-- 
Tim Mooney (mooney@aps.anl.gov) (630)252-5417
Beamline Controls & Data Acquisition Group
Advanced Photon Source, Argonne National Lab