RUN 2B L1CAL MEETING MINUTES
                COLLABORATION MEETING
                 29 September, 2005

Present 
o Fermilab      M.Abolins, T.Adams, L.Bagby, E.Barberis, J.Blazey, 
		N.Buchanan, M.Cwiok, D.Edmunds, H.Evans, G.Ginther, 
		C.Johnson, S.Lammers, M.Mulhearn, Md.Naimuddin, 
		V.O'Dell, J.Parsons, A.Roe, N.Varelas, D.Wood
o MSU           C.Brock, P.Laurens



Discussion of Splitters and Other Ways to Deal with the Shutdown
----------------------------------------------------------------

o Apology
  I will not attempt to reproduce the detailed discussion concerning
  the splitter option but will simply summarize the main arguments and
  the conclusions of the meeting.

o Three possible options were discussed
  0) Install splitters for only the "barrel" region of the
     calorimeter.
     - this was discarded almost immediately because no one could come
       up with any real benefits of it.
  1) Install splitters for the entire complement of TTs
     - either delay the shutdown until the splitters are ready and the
       system tested if they are late, or
     - install the splitters whenever we are ready regardless of the
       shutdown timing
  2) Define a "stable operation state", which can be achieved on the
     timescale of Oct 31 - mid-Nov. Operate the system as part of
     standard D0 running, with monitoring, triggers to the TFW, etc.,
     in this Stable Operation State until it is time to install. Then
     follow the standard (MCH1) installation.

1) Splitter Option
   - pros
     . may allow running under more realistic conditions than with test
       data and a few real TTs before installation
     . allows a fallback scenario of switching back to the IIa system
       if the IIb trigger fails (this is NOT seen as a strong
       argument) 
   - cons
     . The schedule for producing ~100 splitters by January is *very*
       tight. No one had confidence that we could achieve it, leaving
       open the very real possiblity that the shutdown could happen
       before the splitters are ready.
     . Because of the tight schedule, full-time work (by Dan and
       others) would need to start on this immediately. Substantial
       orders would have to go out very soon.

2) Stable Operation State
   - This is essentially the same plan that was in place when the
     shutdown was scheduled for Oct. 31 with the addition of more
     emphasis on stable control room operation.
   - issues
     . need to define precisely what is meant by the "Stable
       Operational State"
     . this will require more emphasis on monitoring and interactions
       with COOR than is foreseen in the current (Oct 31) plan

o DECISION
  - The large uncertainty in the schedule for producing a working
    splitter system, with infrastructure, that could be used by D0 for
    the rest of the experiment's life made everyone present very
    nervous about endorsing this.
  - Those present therefore decided NOT to continue to pursue this
    option. 
  - If you do not agree with this decision, you need to prepare a case
    for the Splitter Option, which will convince your colleagues, by
    next Tuesday's meeting. This will be very difficult.

o ACTION ITEMS (Homework for everyone)
  - Following the Stable Operations State option requires that we
    define exactly what this state means as soon as possible.
  - Everyone is asked to prepare a list of things that they consider
    to be crucial to a stably operating system for the L1Cal meeting
    on Tuesday. Concentrate particularly on the area in which you are
    working. 
  - Areas to consider in the definition are
    . Hardware operation
    . System initialization
    . Interactions with COOR
    . Monitoring: examines, online monitoring of data and errors,
      pulser, monitoring of crate power, cooling, etc.
    . Trigger terms: define which ones we need to send to the TFW
    . others that I forgot