[NeXus-committee] Revisions to NeXus paper

[Osborn, Raymond]

Dear colleagues,
I have just uploaded a separate branch to the ‘communications’ repository that contains a revised version of the paper. I must apologize for being so late in the game. I know that there has been a lot of edits made over the past month, and most of the paper I thought was perfectly fine. However, I did feel that Section II, the Design Principles, didn’t provide the kind of overview that someone completely new to NeXus would need, so I have largely rewritten it.

A few other changes:

1) I removed “This is a very important use case” from the abstract. I think that’s redundant - it’s obviously important or we wouldn’t describe it as the first purpose of the format. I also switched “experiment” for “beamline” - non-facility users might be vague about what a beamline is.
2) I added the old Physica B reference from 1997. I think it shows how consistent we’ve been in our design principles, but it could be removed if people don’t want it.
3) In the introduction, I changed “A link structure” to “Features to enable rapid data visualization.” The original implies that data in the NXdata is nearly always linked to an NXdetector group, but we later talk about processed data not requiring links so it’s not mandatory (in any case, I never use such links in my own files). The important design feature is that HDF5 attributes define signals and axes, not that the signal and axes may be linked from another group. 
4) For similar reasons, in Section III.A, I rewrote some of the text about default visualization, changing “typically” to “often” concerning links.
5) Also in Section III.A, I rewrote the paragraph on the NXmonitor group. Obviously, some people think that decision was a mistake. I don’t agree - monitor data is almost as important as the other data before data reduction, particularly at pulsed neutron sources - but, in any case, I don’t think we should be apologizing for our past decisions in a paper designed to promote the format. That’s for our internal discussions. 

For the same reason, I would argue that we should remove the section on how long NeXus has taken to become established. We’ve made tremendous progress since 2006, so I’m not sure we should be defensive about it. However, I haven’t deleted it in my version because I didn’t know what to replace it with.

I haven’t issued a Pull request yet. I thought people might want to check the differences before anyone did a merge.

With best regards,
Ray
-- 
Ray Osborn, Senior Scientist
Materials Science Division
Argonne National Laboratory
Argonne, IL 60439, USA
Phone: +1 (630) 252-9011
Email: ROsborn at anl.gov