HPX documentation now uses Sphinx

With the release of HPX 1.2.0 we moved from a Boostbook-based documentation system to Sphinx. The latest documentation is now hosted on GitHub pages and can be found here.

There were multiple reasons for moving to a Sphinx-based documentation setup:

  • Modern look: the Boostbook-based documentation used awkward, non-responsive styling
  • Search: Sphinx natively supports search which makes it much faster to find what you’re looking for in the documentation
  • Table of contents in sidebar: this also helps navigating the documentation more easily
  • More familiar markup with reStructuredText

While the above are mostly visual reasons for moving to Sphinx, we used the opportunity to restructure the documentation at the same time. The first page of the documentation now tries to guide new and old users to sections that they might be interested in. For new users there is a quick start guide since we know that starting out with HPX can often be a big hurdle (to help with that there are also HPX packages available on Fedora). Our examples have been labeled with what you will learn by reading through the examples. For more advanced users the comprehensive manual has been restructured in a more logical way to start from getting and building HPX, to writing single-node HPX applications, multi-node HPX applications, and finally to optimizing and debugging applications. We’ve also added a new section for developers getting started with HPX. A less noticeable feature is that we link key concepts to our terminology page. Next time you see a term you haven’t seen before you’ll hopefully be able to just click the word to get an explanation.

We hope you enjoy the new documentation and would love to hear your feedback on it!