Argonne : R. Blair, J. Dawson
CERN : P. Golonka, K. Korcyl, S. Wheeler
Michigan : B. Pope
NIKHEF : J. Vermeulen
UCL : R. Cranfield
1. Paper modelling: Draft version 0.9 of the document on paper modelling results was finalised the day before the meeting by M. Abolins and J. Vermeulen. This version is the final version, unless comments are received requiring updating of the document. J. Vermeulen showed some of the results.
2. ROS modelling: The UML description is nearing completion, feedback from D. Francis was incorporated but was lost and has to be incorporated again. Reverse engineering the existing code and also learning how to use Together has been found to take a long time. The diagrams and a document containing an overview of a the diagrams (at the moment of the meeting not completely up-to-date) are available. R. Cranfield is still looking for a good way to automatically include text stored with Together in a Framemaker document. J. Vermeulen reported that M. Mueller had found the diagrams useful when studying the ROS software.
3. Data Collection modelling: K. Korcyl and P. Golonka have made progress with switch models and the model for the TCP/IP stack, see the slides presented by P. Golonka. A paper has been submitted for the proceedings of the RT2001 conference to be published in IEEE Trans. of Nucl. Sci., see this web page.
4. HLT modelling: S. Wheeler is waiting for UML diagrams describing the foreseen message flow and actions of the DFM. The new PC has not yet arrived.
5. Reference testbed model: The document describing the results was almost finished: K. Korcyl and J. Vermeulen had agreed on a number of small modifications to be made the morning before the meeting. It was concluded that the reference testbed model no longer needs to be discussed in future meetings.
6. Ptolemy: K. Korcyl is finalising the implementation of sequential processing in at2sim.
7. Simdaq: There was nothing to report. It is planned to run the program with updated configuration and parameters, as used for the paper model, to cross-check the paper model results.
8. "Chiba City": R. Blair reported that a number of succesful runs has been made, some runs did not produce results (slides). The cause of the unstable operation is not understood. K. Korcyl reported that a configuration for at2sim describing the Chiba City system has been produced by M. Brezuleanu. The model can be run, studies with it can be started.
9. Workplan and status in relation to workplan: The workplan was discussed. The deliverable of the most important milestone, the updated paper model, was available about in time. Another deliverable, the document comparing modelling and measurement results for the reference testbed is now also available, although later than originally planned. Work on Chiba City, the production of UML diagrams and the implementation of a full system model with rough component models are late. It is felt that this not problematic: phase 2A being later than foreseen gives us more time for these tasks. One item, "reconfiguration and/or review of models to conform to Phase 1" will be deleted from the task-list, as its usefulness is small because no detailed results are available. R. Blair will update the workplan taking into account the delays.
10. Next meeting: The next video meeting is foreseen for Monday 10 September, 15.30 h CERN time.
11. AOB: J. Vermeulen has attended the "Connect" (QA group) meeting held just before this meeting. A proposal for a software development process was discussed and would be presented by D. Burckhart - Chromek in the plenary meeting of the following day .
J. Vermeulen, 9 July 2001