[Nexus-developers] Some more definitions

[Mark Koennecke]

Hi,

I added application definitions for NXtofraw, NXarchive and NXtas. In 
this I
encountered a couple of issues:
- Many of the fields of NXarchive are optional. Can I signal this with 
the minoccurs/maxoccurs  attribute pair?
  Will this be honoured by nxvalidate?
- I have taken massive liberties on NXtofraw:
-- I removed tons of metadata which did not seem relevant for DA. May be 
to much, please check.
-- I am missing something important: IMHO each TOF instrument has a 
pulse defining chopper. This is the
   chopper, which sends the signal to start counting time-of-flight. 
There may be tons more choppers but they
   are only there to have the neutron packet  win the Miss Neutron Pulse 
Beauty competition.  From what I know, 
   the distance of the pulse defining chopper to the sample is very 
important in order to calculate the total flight
   length of the neutron for DA.  Now, no chopper was in the tofraw on 
the wiki.  I think I have to add this.  Any 
  opinions?
-- With NXdetector I have devised yet another scheme: The data comes as 
detector number versus TOF. In addition
   there is a square array, arrangement, which maps detector number to 
their spatial arrangement in an area detector.
   Where a bank of single detectors is an area detector with a y size of 
1. Thus an interested program can reconstruct
   the 3D data volume when necessary. The advantage is that I can always 
do a an easy plot of the data for inspection.
   IMHO, visualizing 3D data from a TOF-area detector is not trivial; no 
easy standard plot here. What do you think?
   I am quite willing to remove this and use the tofraw as a base class 
and derive some more detector geometry
   specific tofraws from it.

Best Regards,

   Mark