[NeXus-tech] Summary of last telco

[Mark Koennecke]

NeXus Telco Week 43 Summary
=============================


NXpinhole, NXslit and axes
---------------------------

The first topic were Eugens suggestions for NXpinhole and NXslit at:

http://wiki.nexusformat.org/NXaperture_and_Slits

In general this was well received. Eugen was encouraged to put this to use
at a DESY beamline and make a suggestion to the NIAC for adoption when
this is known to work.

Some confusion occurred about the positional fields in those base classes:
x_translation, height, distance etc.

The first clarifcation necessary was that each of these fields is an 
axes and
thus will have the required attributes: transformation_type, offset, 
vector, etc.
And that there may be more axes when required.

The second clarification necessary was the way how the axis depedency chain
is decoded. The rule is that the depends_on field in a base class names the
axis at the top of the dependeny chain. Then you follow the depends_on
attributes of the axes until the end.

The example in: http://download.nexusformat.org/doc/html/design.html, 
section
1.3.2.3 describes a situation where phi is sitting on top of chi which 
in turn
sits on top of rotation_angle. And rotation_angle sits at 0,0,0 and is 
the end
of the chain.

The discussion raised again the topic of assigning a class to a field. Such
that we can verify the existence of the axes attributes. This discussion 
was
deferred until our validation experts become available. This is initmately
related with NXgoniometer as NXgoniometer is just a collection of axes.


Variants and ESD
-------------------

Two open issues from both the CIF-NeXus integration and the MX requirements
discussion are variants and standard deviations. We are circling these 
issues
for several Telcos now. The decision now is that we write up the current
state of the discussion and seek comments from the wider NIAC. Under
consideration of the comments obtained, the issue will be put to a vote.


MX Requirements
------------------

We checked the list of MX requirements again. In order to move this on
we decided on a two pronged approach:
- We continue to address the issues raised.
- Tobias collaborator will try to construct a MX application definition 
with what
   is available now


Multi Module Detectors
--------------------------

In the context of FEL's Herbert brought up the issue of multi module
detectors. This is a detector which consists of several separate detector
modules. It is both necessary to know about the detector as an assembly
as well as to know about each separate module. It really is a hierarchy.
In principle, the problem can be solved by having a NXdetector group for
each module. But this looses the hierarchy.

The issue turned up in the context of the collaboration with DECTRIS too.
Thus it was decided that we need to find a solution. But with -1 minutes
left in the teleconference we decided to defer the topic until the next
Telco.  And think about possible solutions in the meantime. Herbert
suggested to invite FEL people to the next teleconference in order to
help us get this right.




Issues pending but not discussed for lack of time
======================================================

Polarisation
-----------------

Current polarisation fields in NeXus reflect neutron use cases.
Needs enhancement for X-rays. Waiting for an expert to make
suggestions.