Pixel ROD software meeting
CERN, 9/12/2002
 
Presents:

Roberto Beccherle
Giovanni Darbo
Kevin Einsweiler
Tobias Flick
Guido Gagliardi
Peter Gerlach
Martin Imhaeuser
Susanne Kersten
German Martinez
Paolo Morettini
Aldo Saavedra
Joachim Schultes
Lukas Tomasek
Norbert Wermes


Comments, corrections, additions to Paolo Morettini


The goal of this first meeting was to try do define the
major steps towards the construction of the software support
for the Pixel ROD. This software will be used in several
environments:

- Test of stave and multi-stave systems
- Test Beam
- DAQ and Trigger performance studies
- DAQ in the experiment

The main task we have identified so far are:

1) Internal ROD software (DSP programming)
2) ROD driver (a C++ class giving access to the basic hardware
   functions of the ROD)
3) TIM driver
4) BOC driver
5) DCS interface (a C++ class giving access to DCS)
6) Configuration/Calibration DB (and the correspnding interface) 
7) Pixel Library (a set of C++ classes representing the primary
   interface between the Pixel modules and the different applications;
   this library will implement all the functions needed to operate a
   Pixel system, and will make use of the drivers and interfaces 
   mentioned above to accomplish its task).
8) GUI for the Pixel Library (an interface between the TurboDAQ
   user interface and the Pixel Library allowing interactive module
   calibration in a ROD environment).
9) Simple Pixel Library based run control (for the use of the ROD in
   the DAQ-1 environment).
10)... (more advanced applications).

For many items we were able to identify one person/group already
working or willing to contribute in the immediate future. 
In particular:

1) -> LBL
2) -> Tom Meyer
3) -> Wuppertal + SCT
4) -> Wuppertal
5) -> Wuppertal
6) -> ???
7) -> Genova
8) -> LBL
9) -> Genova

It's obvious however that this is just an initial arrangement
which will change as soon as we will have a more detailed idea
on how to implement the different parts. In particular Bonn has
no task assigned for the moment and one important task as the
configuration database is still uncovered.

Another important comment is that most of the people involved
never saw a ROD, so it's important for most of us to start
playing with it as soon as possible.

This is the end of the minutes of the meeting. Trying to summarize
what has been discussed and to convert it to an action list I 
would propose this plan:

- LBL prepares a very minimal implementation of the ROD internal
  software (DSP and FPGA). [Now->Jan-15]
- Genova prepares a proposal for the structure of the Pixel
  Library. [Now->Jan-1]
- The available RODs are distributed to the groups interested
  in the development of the software for a first look. The labs
  will need: a 9U VME crate; a Concurrent Technologies Single
  Board Computer running Linux. The examples of use of the
  RodModule class already being used in the SCT group will be
  used for the startup. [Jan-15->Feb-15]
- LBL continues the preparation of the ROD internal software.
  [Jan-15->Feb-15]
- Genova prepares the skeleton and the infrastructure for
  the Pixel Library [Jan-1->Feb-1] and the entire group
  dicsuss on the proposed structure.
- After having solved the most trivial problems and having
  acquired a minimal set of programming skills, the sotware
  developers go to LBL for a "full immersion" training session 
  with the ROD designers. [Feb-1->Mar-1]
- After having discussed and revised the structure of the Pixel
  Library a minimal version is implemented and interfaced
  with TurboDAQ [Feb-1->Mar-1]

Let me know if the proposal looks reasonable and feasible.
We will discuss this using the mailing list atlas-pixel-daq@cern.ch.