Visiting Údine, Fall 2003

• Wonderful food and drink
• Wonderful people; see this partial sample
• Many thanks to Riccardo, Marco and Alessandro for a very pleasant and productive visit

Highly recommended!

Diagnostic detector display, with Riccardo

Motivation: Correct geometry description is a requirement for accurate GLAST simulation, reconstruction and event analysis. The instrument, and hence its description, is sufficiently complex that reliable, easy to use tools are necessary to verify its correctness.

Goal: Design and implement a stand-alone program which generates a complete (full depth) heprep description of the detector, suitable as input for the pre-existing interactive display program Fred. Additionally,

• The description will include "extra" attributes not needed for the standard event display, but invaluable for detecting errors in the geometry description.
• The per-object information will be readily available via the gui.
• Users will be able to choose interactively by means of a "show/hide list" which objects shall be visible.

Status: Some parts of the show/hide functionality have not yet been implemented; all other features exist and function as spec'ed. The stand-alone program is in the geometry diagnostic package detCheck. It depends only on detModel

Here the SiLayerBox instance belonging to the bottom tray has been selected.

More screen shots:

• Using the file browser
• The Type Tree

Gui for calibration metadata database, with Marco

Motivation: Currently all additions to the database are done in SQL (by hand, so to speak) by me. The procedure is too awkward and error-prone to allow direct access to the larger class of users (e.g., subsystem experts) who should legitimately have it. The only tool currently available for browsing the database is SQL. A higher-level tool, incorporating information about application-specific aspects of the database, would save a lot of time and effort.

Goal: Design and implement a program or programs to reliably

• create a database with specified characteristics
• allow a user to add a row to the database, providing help to increase the chances that the row is sensible and consistent w.r.t. application requirements as well as the syntactical requirements which follow from the formal SQL description of the table.
• prompt for and display the results of an SQL query (or multiple queries, displayable simultaneously in separate windows)
• verify that inter-row conditions (e.g., disallow multiple production calibrations of the same type covering the same time period) are met

Status: Major design choices have made which will support all the requirements.

• Single XML description of database will be input to programs implementing all of the above. Most, but not all, of the details of the XML language to be used have been worked out. The description supports specification of constraints beyond those expressible in SQL, so that the programs using the description need not contain any GLAST-specific hard-coded information.
• FOX will be used for graphics. It already contains most of the widgets needed, has excellent performance, and easily supports the requirement for multiple windows.

Implementation is just starting. It will be staged in such a way that sophisticated functions will be deferred so that more elementary, but more immediately desirable, requirements may be met sooner. We believe we can do this without too much disruption as additional functionality is added. See this draft of and XML description of our current calibration database structure (you might have to view source to see anything meaningful).

J. Bogart
Last modified: