Thanks for the comment.

If you visit http://www.ukoln.ac.uk/web-focus/papers/papers.rss you’ll find a manually maintained list of my papers which I use to feed into various services, as I’ve described in my post. I’m aware that that’s not a very elegant way of managing the data and several years ago was informed that our repository would manage the information for me and provide the RSS feed for me to use, either for individual papers, departmental papers as well as for arbitrary search queries.

Whilst I appreciate dynamic generation of RSS feeds may be computationallly intensive for the first two use cases the feeds (which will not be very volatile) could be generated overnight.

Note that the small limit on the number of RSS items was defined in RSS 0.9. However RSS has developed since then – and it should be noted that it is no longer regarded as just a news alerting mechanism; rather it is a syndication technology with syndication of changed items just being one aspect of what it can be used for. RSS has grown up!

Note that if there was a button which said “Export current view into foo format” and foo could be transformed into RSS, that would seem to be a solution which could minimise load on ePrints servers.

Thanks

Brian

I’m a bit confused by why you think EPrints has “very limited” support for RSS. Every live RSS feed I’ve come across limits the number of records returned.

RSS isn’t intended to be used for harvesting the historical record of events. It’s supposed to be a very lightweight (and small) set of data that can be regularly polled with minimal overhead on the server + client, which is what EPrints does.

You need to either retrieve records in a different format or arrange a separate access mechanism that will generate the complete record set in RSS format. That’s pretty easy to do but must be be done separately to avoid clobbering normal RSS consumers.

Sincerely,
Tim.

It’s OK… none of the rest of us in UKOLN knew what on earth you were doing either! :-)