Mark Koennecke (then ISIS, now PSI) made a proposal using netCDF^[18] for the European neutron scattering community while working at ISIS

Jonathan Tischler (ORNL) proposed an HDF-based format^[19] as a standard for data storage at APS

Ray Osborn convened a series of three workshops called SoftNeSS (http://www.neutron.anl.gov/softness). In the first meeting, Mark Koennecke and Jon Tischler were invited to meet with representatives from all the major U.S. neutron scattering laboratories at Argonne National Laboratory to discuss future software development for the analysis and visualization of neutron data. One of the main recommendations of SoftNeSS'94 was that a common data format should be developed.

At SoftNeSS 1995 (at NIST), three individual data format proposals by Przemek Klosowski (NIST), Mark Koennecke (then ISIS), and Jonathan Tischler (ORNL and APS/ANL) were joined to form the basis of the current NeXus format. At this workshop, the name NeXus was chosen.

The HDF-4 API is quite complex. Thus a NeXus Abstract Programmer Interface (NAPI) was released which simplified reading and writing NeXus files.

At SoftNeSS 1996 (at ANL), after reviewing the different scientific data formats discussed, it was decided to use HDF-4 as it provided the best grouping support. The basic structure of a NeXus file was agreed upon. the various data format proposals were combined into a single document by Przemek Klosowski (NIST), Mark Koennecke (then ISIS), Jonathan Tischler (ORNL and APS/ANL), and Ray Osborn (IPNS/ANL) coauthored the first proposal for the NeXus scientific data standard.^[20]

SINQ at PSI started writing NeXus files to store raw data.

MLNSC at LANL started writing NeXus files to store raw data

NeXus API version 2.0.0 is released. This version brought support for the new version of HDF, HDF-5, released by the HDF group. HDF-4 imposed limits on file sizes and the number of objects in a file. These issues were resolved with HDF-5. The NeXus API abstracted the difference between the two physical file formats away form the user.

Przemek Klosowski, Ray Osborn, and Richard Riedel received the only known grant explicitly for working on NeXus from the Systems Integration for Manufacturing Applications (SIMA) program of the National Institute of Standards and Technology (NIST). The grant funded a person for one year to work on community wide infrastructure in NeXus.

In 2003, NeXus had arrived at a stage where informal gatherings of a group of people were no longer good enough to oversee the development of NeXus. This lead to the formation of the NeXus International Advisory Committee (NIAC) which strives to include representatives of all major stake holders in NeXus. A first meeting was held at CalTech. Since 2003, the NIAC meets every year to discuss all matters NeXus.

The community asked the NeXus team to provide an ASCII based physical file format which allows them to edit their scientific results in emacs. This lead to the development of a XML NeXus physical format. This was released with NeXus API version 3.0.0.

NeXus API version 4.0.0 is released with broader support for scripting languages and the feauture to link with external files.

NeXus API version 4.1.0 is released with many bug-fixes.

NXDL: NeXus Definition Language is defined. Previously, NXDL NeXus used another XML format, meta-DTD, for defining base classes and application definitions. There were several problems with meta-DTD, the biggest one being that it was not easy to validate against it. NXDL was designed to circumvent these problems and all current base classes and application definitions were ported into the NXDL.

NeXus API version 4.2.0 is released with additional C++, IDL, and python/numpy interfaces.

NXDL and draft NXsas presented to canSAS at SAS2009 conference

NXDL presented to ESRF HDF5 workshop on hyperspectral data