# Difference between revisions of "August 26, 2009 Software"

### Location

The meeting will be at 2:00 pm in CEBAF Center F326/327.

### Telephone

To connect by telephone:

1. dial:
• 800-377-8846 : US
• 302-709-8424 : International
2. enter participant code: 77438230# (remember the "#")

### Video Conferencing

A) ESNet: 8542553

# Agenda

1. Review minutes from August 12, 2009 Software meeting
2. Announcements
3. Collaboration Meeting Talk: Mark
4. Action item review

# Minutes

Present: Pawel Ambrozewicz, Eugene Chudakov, Mark Ito (chair), David Lawrence, Simon Taylor, Elliott Wolin, Beni Zihlmann

## Review of Last Meetings Minutes

We went over items from the last meeting. Most of the issues left over from last time were addressed in the review of action items below.

## Announcements

• David: A bug in Jana was been found, having to do with multiple threads setting configuration parameters at initialization. This was only discovered when running many tens of threads under a single process. It has been fixed.
• Elliott: The [??? halld group changeover is done.]
• Elliott: Next week's online meeting conflicts with the first staff scientist interview. Look for changes in the schedule for that meeting.

## Talk at Collaboration Meeting

Mark solicited input for the offline talk at the upcoming collaboration meeting. David suggested that he target individuals and ask for slides.

## Action Item Review

We went over recently completed and outstanding items on the list

Drawing Storage: We need to ask Tim Whitlatch about how we can access Document Control at JLab and further, how to navigate the terrain once we get access. We might want to extract certain key drawings and make them available on the wiki. JT2GO???accessible drawings might be nice as well.

Group Disk Usage: CLAS is using about 150 GB of space compared to our 35 GB. This ratio does not seem unreasonable at this point. Beni pointed out that 9 TB of RAID 6 storage would run about \$1,200 these days.

FDC Half-Cell Stagger Studies: Simon gave a [??? presentation] at last Monday's tracking meeting. His preliminary conclusion is that we do not need the half-cell stagger. The $\chi ^{2}$ distributions do not look much different between the two designs he studied. Eugene wondered if the results depended on the starting parameter of the fit; if the start is far away from the true parameters, then it may be more likely that incorrect left-right choices are made. Simon sketched the procedure used in track finding and fitting. He has a few more things to look at before he is done, including a geometry where a packages has six layers with all wires in all layers parallel to each other, with an alternate half-cell stagger.

???do the other layers in the default geometry help resolve l-r in a given layer???MMI answer is yes

???kalman filter reference???

poisson: some maps generated, 4x # points tried successfully along beam line discreprency with ansys compare field directly run more simulations compare resolutions