About the Project

The project started when several different library constituents wanted to have something more useful or topical to give people to understand what RSS was and why they should care about it. Instructional people wanted good examples for classes. Techies wanted a reusable code-base. Everyone wanted something relevant to our patrons, like journals they use and we subscribe to.

We started out to build a portal for all this, inspired by the wonderful work at the University of the Health Sciences - Terkko in Finland on their FeedNavigator product. We quickly decided that this might be overwhelming and we didn't want to force all our patrons into one application or one product. So we decided to create OPML bundles as a distribution method, divided up by subject areas of relevance to our primary patrons, and build in the RSS functionality into our existing journal web pages as well. This allows for a wide range of usages by different groups (affiliates, non-affliliates, novices, power users, etc).

We took our existing journal listings (some of which already had RSS info), divvied them up among a working group and found all the feeds we could, then added them to our database to significantly fill in the gaps. We added subject information based on a combination of SFX categories, Journal Impact Factor listings and internal decisions. And, voila! The RSS Project was born.

Possible next steps include some sort of centralized, cross-institutional repository of feeds for journals (since the vendors are taking so long to do it) and maybe a shopping cart application to allow patrons to "roll their own" OPML bundles.

Related Posters & Presentations

SlideShare | View | Upload your own


Categorization Process

The large number of feeds in the alphabetical list prohibited the categorization of each title individually. We did not have the required time nor did we have an established taxonomy for categorization. To resolve these problems, we leveraged the category assignments in our Ex Libris SFX knowledgebase. SFX, the library's OpenURL link resolver software, has an internal taxonomy for electronic journals. Although the taxonomy is not as refined as other health sciences taxonomies, particularly MeSH, we deemed it adequate for the task. We exported the SFX categories for the library's active subscriptions and then matched them against the library's electronic journals database. After the export and match, we divided up the categories and reviewed the assignments for accuracy. Following the review, the group had the content to create the categories lists and web pages.

Proxy Issues for EZProxy Users

At the Ebling Library, we put our dynamic TOCs behind our proxy server and EZProxy's host-name based rewriting maintains the proxy for the browser session of the off-campus user. No other scripting is necessary.

We do not proxy feeds on the alphabetical and categorical lists (and corresponding OPML files) because proxied URLs cannot be parsed properly by feed readers. Additionally, we want the feeds to be available to users everywhere. Our proxy prefix in the feed URL would be an impediment to users beyond our campus.

Can I Have Your Data?

You may be better served by grabbing individual feeds or some of the OPML bundles, but if you really want everything, here it is:

File > save page as > web page, xml only
You should be able to import this file into anything capable of reading OPML. Fair warning that it could take a LONG time to import this many feeds in to your reader.

Please contact EblingWeb if you would prefer a MySQL script, or other output format.

Any Plans for a Collaborative Maintenance/Sharing Space?

Please contact EblingWeb if you are interested in helping maintain this collection!

We are collecting names and contact info of folks that have indicated that they would be interested in some sort of shared environment to allow for collaborative editing and maintenance of this feed collection. When there are enough interested parties that it looks like we could get a project going, we'll look more closely at how we could set this up.

Displaying Feeds on Your Web Site

SimplePie (what we use now)

Currently, we use SimplePie to display feed content on our site. We transitioned from Feed2JS to SimplePie, as it gives us a lot more in the way of customization with the output. Non-PHP environments may want to give SimplePie Live! a whirl, but I haven't tried it, myself.

Feed2JS (what we used in the past)

Initially we used Feed2JS to display much of the syndicated content on our site. We installed and ran the code on our own server, but if you're a non-PHP environement, or this just sounds hard, they also offer a quick and dirty cut and paste feed builder option that makes displaying individual feeds on your web site ultra-simple.

I Just Want to Display a Few Feeds on My Site, Not a Few Thousand! Isn't There an Easier Way?!?

Oh, you bet! There are lots of services that you can use to do this. If you are just wanting to display the content of a few feeds on your site, the database-driven/programmatic route we took is absolutely not for you. That would be complete overkill! You would be better served by using any number of the fine cut-and-paste feed builder options that are out there now. They all work approximately the same. You supply a feed and answer a few questions about how you want it to look, numbers of items to display, etc. and it generates a couple lines of code you just paste into your web page. That's it!

Here's a few to consider:

What About FULL TEXT?!?

For resources we subscribe to, or are freely available, we always provide our patrons with access to the full text directly from the links within the tables of contents displayed on our site. This was accomplished with a small modification to the SimplePie code which appends our proxy server link to the article link in the feeds.

What if a Patron Subscribes to a Feed? Do They Get the Full Text?

On campus, the vendor supplied feeds generally work just fine and will get our patrons to the full text of resources that we subscribe to.

Off campus is a whole 'nother matter. Vendor supplied feeds typically WON'T take you to the full text (there are some exceptions, but that's the general state of affairs). There are ways to work around this problem, however. We create new feeds with proxied links using the fabulous Yahoo Pipes. We don't supply these feeds on our public pages as of yet, but are creating them and supplying them to patrons as needed.

Making Pretty OPML Styles

Our OPML is styled with just a slightly modified version of Christopher Finke's stylesheet for transforming OPML files.

I Still Have a Question

Please contact EblingWeb with any additional questions you have.