Comments from the 06/03/05 and 07/01/05 LCLS SLC-Aware Requirements Review: 
	Magnet-like Device Monitor and Control

Presenter:  Kristi
Attendees:  Patrick, Doug, Stephanie, Diane, RonC
	    
Specification:

http://www.slac.stanford.edu/grp/lcls/controls/global/sw/slc_ioc/requirements/mgnt_funcReqts.html

Note that the spec only has the SLC requirements but you'll be adding the 
EPICS requirements in the near future.

Comments from the July 1 EPICS-related Reqts Review:

(1) For magnets that do not have fast control, records only need a rough 
timestamp.  IOCs that don't have fast magnets will not have an EVR.  For 
fast magnets, the command values must use the EVR timestamp so they can be 
correlated with BPM data, etc.

(2) On day 1, the main SCP apps using magnets are steering and lattice 
diagnostics.

(3) There will be no special SCP code added to control the kickers.  The 
apps for the LINAC and LTU kickers will be EPICS.  No support of any 
LCLS pulsed-amplitude-units (PAU) from the SCP.

(4) Unlike the SLC micro job, when an action is requested for a group of 
magnets on an IOC (ie, a trim), all magnets will perform the action in 
parallel.  If anything on the SCP breaks because of this, a fix external to 
EPICS will be done.  If it is important that actions be done in sequence, 
the higher level app (ie, a script) will perform the sequence.  

(5) Do we really need the ability to set the I (current) setpoint to the 
I readback value?  If yes, does the B setpoint need to be backed out from 
the I setpoint to be consistent?  Perhaps this function should be done later 
or never.

(6) The function to force a readback is not needed.  Readbacks should be 
done often enough periodically.

(7) The only knob box support on day 1 may be SLC-only.  RonC's group plans 
to upgrade the software used by the existing PEPII control room knob boxes.  
Patrick and RonC and others to discuss further.

(8) Kristi says all individual binary input and output states will be 
provided to the users (no knowledge of bit masks required).  Kristi says 
alarm handler will be used to find bad units, out-of-tolerance units, etc 
without the need for specialized displays.

(9) EPICS waveforms needed to create B vs I and B vs Error EDM displays are 
not needed.  These are "static" plots and some other tool like Matlab might 
be better to plot the data using the polynomials.

Comments from the June 3 SLC-related Reqts Review:

(1) Degaussing required on day 1.  Best to do it in EPICS so no multiknob 
setup needed in VMS.

(2) Micro intermixing in LINAC - further discussion needed.

(3) Undulator magnets controlled by EPICS host apps only.

(4) Remove chromaticity from apps.

(5) No SLC support for XMOV, YMOV, probably STEP, SEPT and SPTS (septum), LENS.

(6) All CA scripts will be done on EPICS hosts, not MCC.

(7) SLC multiknobbing and button macros will be used.

(8) Monitoring of static data (supertype 1) that is duplicated on SLC and 
EPICS and reporting differences will be done on an EPICS host instead of IOC.

(9) The SLC magnet CSTA (CAMAC status register) will not be supported.  
Interlock status for LGPS will be supported on EPICS only.