Data Structures Task Force, Minutes, Jan. 28, 2002, 08:00 PST, VRVS, Saturn

Present: Heather, Toby, Seth, Michael K, Tracy, Hiro T,  Leon

Leon presented a summary of a discussion between CAL and TKR about some common issues.

• Toby suggested putting the GlastAxis class into the idents package.
• On the issue of the numbering of the layers, Toby pointed out that using iterators should avoid the problem. I'm not sure if the layer numbers can be reduced to mere identifiers everywhere, but it's worth looking into.

Toby showed us the current structure of McVertex and McParticle. (For a good intro to UML, and a very useful book in general see: Design Patterns, Erich Gamma, et al., Addison Wesley, $49.95.) He argues that it seems unnecessarily complex.  An entire class (MCParticle) exists only to provide the particle Id.  He suggests:

• keep only one class (probably called McParticle), containing all the information
• implement some methods to set things up, for example, to make daughter particles.

This is similar to what is done in Gismo in the GParticle class. The current scheme was implemented by Richard.  Toby needs to check with him about this.

There was some discussion between Toby and Heather about inline methods in classes derived from data objects.  I didn't quite get the thrust of it, but I'll try to get better informed.

There was more discussion about our meeting time.  Leon is proposing alternating with the soon-to-be-resumed TKR meetings on Tuesday mornings, which would conflict with Heather's Doc TF.  We'll try to gather more information on this.

Action Items:

• Toby will make his proposed changes to McVertex/McParticle, after checking with Richard.
• Leon will add GlastAxis to idents.
• Leon/Heather will work on the meeting time.

Next meeting: Tuesday, February 5, 2002