OWG meeting 12-Nov-2008

Agenda

• Announcements - room uncertain for the next few meetings
• Review of minutes from 3-Oct-2008 meeting (29-Oct meeting was canceled)
• Report from EPICS collaboration meeting (Padua) and PCaPAC conference (Ljubljana) (see GlueX-doc-1159-v1) - Elliott
• Update on fiber test, electronics, trigger, DAQ, etc - Many people

Time/Location

Note different room

Wed 12-Nov-2008 1:30pm CC L207

Next Meeting

26-Nov-2008 room TBD

(unless everyone is gone for Thanksgiving!)

New Action Items from this Meeting

Query collaboration concerning short and long-term notebook/logbook needs and requirements - Elliott

Minutes

Attendees: Elliott W, Dave D, Chris C, Carl T, Dave A, Graham H, Dave L, Simon T, Fernando B, Alex S, Jim S.

Report from EPICS Collaboration Meeting and PCaPAC Conference

Elliott presented results from the two meetings plus pictures of Padua, Venice and Ljubljana. See slides for more details.

• The EPICS meeting included a two-day tutorial followed by a general EPICS collaboration meeting.
• Many packages that were in development a few years ago are now in production, and look very good. Some new packages have been developed. Interesting packages include archiver/display package and a new alarm system. Most come from SNS, BNL, and DESY.
• Driver support has completely been rewritten.
• The IRMIS package manages a database of channel names, devices, cable connections, power connections, crate and rack locations, etc. It may be just the thing we've been looking for, i.e. what Jim was asking about at a previous meeting. Once the tagger review is over Jim and I will have a good look at it.
  • Eclipse-based Control System Studio (CSS) is being widely used to develop gui's and as the framework for many other packages.
  • NAGIOS is being used in many places to supply web access to controls data, and even as an alarm system. NAGIOS is similar to MonAlisa, and the two appear to be converging.
  • The EPICS community now embraces modern computing techniques (Java, XML, SOA, Web, JDBC, JMS, Python, AJAX), and is replacing many old packages.
  • RTEMS is replacing VxWorks at many sites.
  • Some of what was reported at the EPICS meeting was repeated at the PCaPAC conference.
  • I presented a plenary talk on the cMsg IPC package.
  • Labview is being used more and more in control systems. In some cases, at smaller sites, the entire control system is developed using Labview.
  • National Instruments now produces a general-purpose FPGA board that can be programmed using Labview.
  • WebCA and the CAML language can be used to display EPICS channel access info on the web.
  • J5, a commercial electronic logbook package, is used extensively at ESRF and may be interesting to Hall D. In J5 you design a logbook and then it deploys it for you, unlike other systems, where the fields are fixed and you just use them.
  • We had a discussion about when we might need a centralized elog for detector development groups and test runs. Suggestion (Jim?) to collect information and requirements from detector groups as to what they might need, and when.
  • Next EPICS meeting is in Vancouver 2009
  • Next PCaPAC is in Saskatoon in 2010

Overall both conference were quite useful. I found many interesting packages and talked with many of the core developers. The latter are open to adding features that Hall D needs. Unlike the case a few years ago, the latest EPICS utilities are modern and up-to-date, and are as good as any I've seen.

Progress Reports

• Fiber test stand - Beni reported no DAQ problems since a typo was fixed in startup file a few months ago.
• Electronics (Fernando, Chris, Dave D)
  • The F1 TDC chip order is out, about $195k, 14 weeks delivery, 20% extras. This is the last time the chips can be ordered, as the production line is being shut down.
  • CTP boards are being fabricated, and bids for the SD boards are coming in.
  • The TI is running with multiple boards and in a multi-crate system (or will be soon).
  • GTP and SSP designs not started yet, waiting for CTP's. TS design still in the future.
  • PCI-based TI card ready, libraries written, being integrated into Coda 2 by Dave A.
  • With CLAS startup Ben is again working on Hall D stuff.
  • Many GUI's and test stands being developed.
  • CNU is looking for students to work on the trigger system.

• DAQ
  • Dave A reported on progress with the GE FANUC Core-Duo Intel-based VME single board computer. He recently received driver source from GE and can now customize the board to run any way he wants.
  • Graham reported EMU progress - EB now logs via cMsg, handles recovery better, and the ER is coming soon.
  • Graham also reported that Coda 3 development is in the 6 GeV hardening budget, not in the 12 GeV project, which means it should be easier to hire the two FTE's that have been approved for the DAQ group, although the hiring is still not approved.
  • Vardan is almost done modifying AFECS to no longer use the external package JADE, i.e. all agent management software is written at JLab.