- Last week's presentation to LArPix was very well received, with lots of great discussion afterwards
- Talks of a test stand for ND-LAr DAQ purposes
- Currently awaiting response from Roland on the preferred way of building unit test to include them for 2.10
- Upon 2.8 release can start working with LArPix and their new student on further development of PACMAN readout and updating the PACMAN frame
- All ready to contribute to the readout reorganization (also waiting for 2.8 release)
— A problem related to the ordering of PACMAN data in the latency buffer was identified. This tracked back to a one line fix in NDReadoutTypes which has now been merged into develop.
— Following this fix, ND data selection is behaving in a comprehendible way. TriggerRecords written to HDF5 are empty or non-empty when they are expected to be based on the timestamps of information sent to the buffer by the PACMAN packet generator.
— Non-trivial to meaningfully verify that the packets in the output hdf5 are exactly what was sent by the PACMAN generator.
— Based on the rhetoric in last week's meeting, it is not necessary to push this ND DS stuff into 2.8 without a formal test to show that it is properly working in place.
— Given it is quite an important step, better to not rush and to aim for 2.10 instead.
— I am going to get the ND DS work into a state where it is fit for another human (Krzysztof) to consume and branch off from the 2.8 release when it is cut. Krzysztof and I will then work together on formulating a definite input with the intention of showing a definite output can be produced. Currently the input and output data formats can’t really be compared, may be necessary to write some code to make this possible.