Wednesday, 14 March 2012

Developing widgets for Pelagios

On Monday, we held a kick-off meeting here at The Open University for the development work on the Pelagios widgets. Liz has already written a post about the workshop part of this.

As part of the prep for this, we drafted a short overview document to make sure that we are all on the same wavelength. I sent a set of questions to Elton, Leif and Rainer which formed the initial basis of this document with input from me as well on the technical side. This was really interesting for me as I realised that I had some misconceptions about the goals of the work and it was good to get these cleared up right at the start.

Here is a draft of the document. Any comments or feedback would be very welcome.

Pelagios Work Package 3 Development Overview

The goal of the development in this work package is to build a suite of 3-4 widgets (i.e. small applications that can be embedded within a web page).

Target audience

There are two target audiences for this work:

  • Site owners who build, own or maintain place-related web pages, web sites or blogs in the ancient history domain. In particular the owners of the Pelagios partner sites fall in this category. These are the people in a position to embed the widgets on their web pages.
  • End users with expertise in the ancient history domain e.g. ancient history researchers, teachers and students. These are people who are likely to use ancient history place-related web pages and will actually view and use the widgets once they are embedded on web pages.

Both groups are critical as if the site owners do not embed the widgets then the end users can't use them, but if the functionality doesn't help end users then there is no reason for the site owners to embed them.


The core focus of these widgets is to allow site owners to easily add some 'Pelagios context' to their own place-related web pages.

This would allow end users to both discover the content of the Pelagios partners and also to become aware of the existence of Pelagios. Ideally the widgets should have some sort of 'recognition value' to help with this awareness.

The following have been suggested as inspiration for the widgets:

  • The 'Find more from Pelagios' links on Open Context and GapVis
  • The 'Related Flickr images' on Pleiades.
  • Product recommendations on Amazon


The widgets should be stable and reliable rather than just a proof-of-concept. They need to be sufficiently bug-free and usable that this is not a barrier to their adoption. This is more important than providing extra functionality.

Overall, we will try and keep the first versions of the widgets simple and then improve or extend them based on how they are used, the feedback we receive and the results of the user testing at the end of each iteration.

We would be delighted if just one of these widgets becomes widely adopted by appropriate sites. The purpose of building a suite of widgets rather than a single widget is to provide a range of options so that site owners can pick the widget best suited to their context.

Technical and Legal

We assume that the widgets will use either the Pelagios Graph Explorer API or a SPARQL endpoint for the Pelagios partners' data. We will therefore be restricted by the functionality provided by these, although we will not rule out using other available data (e.g. other data from the Pelagios partners or data provided by the Flickr API on Pleiades machine tags) in addition to this.

There are no current plans to build on the Pelagios Graph Explorer itself (as opposed to the API), although the code may be useful for reference.

We are currently using the term 'widget' in the broadest sense. So for example, 'widget' does not necessarily mean a 'W3C widget', 'Google gadget' or widget conforming to any other specification. It is more likely that instead an approach similar to this will be taken:

The source code will be made available under the GNU Public License version 3.0 and be hosted on GitHub from the alpha version onwards. We intend for development to be in the spirit of open-source. We need to be aware of legal issues related to any images that we display in the widgets.


There is no specific resource to make the widgets accessible for people with disabilities. As this is a short project, we will attempt to adhere to accessibility best practices while developing widgets where this will not increase development time too significantly, but there will be no specific work on accessibility.

Browser support

We will test the widgets in the latest versions of Firefox and Chrome and in Internet Explorer 8. We will endeavour to also support the latest version of Safari and Internet Explorer 9 but will not explicitly test in these. We will not be supporting Internet Explorer 7.

No comments:

Post a Comment