We started publishing web feeds some three and a half years ago and were initially focused on machine-readable descriptions. Then (as now) this meant RSS 1.0 which as an RDF profile allowed for properties from multiple vocabularies to be added into the feed and (crucially) for the feed as a whole to be consumed by any RDF-aware application and for the properties to be added wholesale to any RDF model. We described the reasons behind this choice of RSS format in a paper published by D-Lib Magazine: The Role of RSS in Science Publishing. See the Web Feeds page on Nature.com for a listing of our current RSS offerings.

Since then many other scholarly publishers have followed suit and published web feeds using this strain of RSS. Some publishers, however, have preferred to go with the more widely deployed RSS 2.0 format and by making use of its looser approach to marking up content have embedded rich descriptions in human-readable form. (During this period another format – Atom – has been published as Internet standard RFC 4287 which allows for a more rigorous means of marking up rich content sections although it still falls somewhat short of being a fully RDF-compliant format.)

Back then there was already an RSS 1.0 module available for encoding content (‘mod_content’) although it was not widely supported by RSS clients. Following a major simplification introduced into that module, which allows for rich content to be encoded as a single property, and with the passage of time, many RSS readers can now support markup descriptions published using the RSS 1.0 format.

So now, to improve the user experience for our subscribers, we finally took the plunge and upgraded our web feeds to provide the best of both worlds: human-readable and machine-readable metadata. Currently this is implemented for our Nature-branded titles, but is being rolled out across all titles. Note also that we will soon be publishing our feeds in both RSS 1.0 format (as current practice) and in a companion Atom format – both formats featuring rich content descriptions for humans as well as for machines. Users can then choose the format that suits their needs best.

See below for a screenshot from the free Awasu RSS client: