Tuesday, October 11, 2022

You get a Sherpa release, you get a Sherpa release, and you get a Sherpa release

It turns out I haven't been keeping this blog up-to-date with Sherpa releases, so we now have had, since the last post, the following releases:

  1. Sherpa 4.12.1, July 14 2020, which was an "emergency" fix to Sherpa 4.12.0 as well as introducing Python 3.8 support;
  2. Sherpa 4.12.2, October 27 2020, which focused on plotting and modelling fixes, although there were a number of fixes and improvements related to flux calculations;
  3. Sherpa 4.13.0, January 8 2021, which is essentially the same as the Sherpa 4.12.2 release;
  4. Sherpa 4.13.1, May 18 2021, which mainly added support for newer versions of NumPy and matplotlib, but also had a few other build updates and bug fixes;
  5. Sherpa 4.14.0, October 7 2021, which introduced macOS ARM support (at least for those bits of the system which support the ARM architecture, which unfortunately means not all optional extras are supported at this time), Python 3.9, XSPEC 12.12.0, re-worked filtering and binning code to address a number of edge cases, particularly with PHA data, and improvements to the documentation;
  6. Sherpa 4.14.1, May 20 2022, which added support for Python 3.10 and XSPEC 12.12.1, reworks to the plotting and I/O backends (with more to come, but unfortunately not in Sherpa 4.15.0), and a number of internal changes;
  7. and Sherpa 4.15.0, October 11 2022, which updates the aging build infrastructure to better match the current Python ecosystem (but as this ecosystem hasn't quite settled down yet, particularly for Python packages like Sherpa which have a bunch of extension modules that require NumPy, means it is an ongoing process), updates to the behavior of the notice and ignore set of commands in the UI layer, changes to how data creation and validation happens to better catch corner cases, and preparation for plotting updates (which unfortunately did not make it this release).

Thanks to the Zenodo site which makes it easier to create a list like this! It also means that as I can link directly to the Sherpa 4.15.0 release notes I'm not going to bother going through the individual changes in this release.

As with previous releases, the Sherpa 4.15.0 release is available via our Conda channel [1], pip [2], Zenodo [3], and GitHub [4]. We thank everyone who contributed to this release with feature requests, bug reports, testing, code contributions, questions, cookies, and any other positive interaction. Please join in the fun on GitHub: https://github.com/sherpa/sherpa
The documentation for this release can be found at https://sherpa.readthedocs.io/en/4.15.0/