WAYFless than perfect 2: gonna hack me up some links

See the first part of this blog post here.

.@benelwell says a WAYFless URL is one that takes you to an error page lol #RAPTOR — Laura J. Wilkinson (@laurajwilkinson) March 19, 2013

As a priority, we need to try and create WAYFless URLs for the following e-resources:

1. Adept Scientific – Adept4Education
2. Anatomy.tv (Primal Pictures)
3. Arts on Film Archive
4. BoB (Box of Broadcasts) National
5. British Library Sounds
6. Carl Giles Digital Archive
7. Copac
8. Creative Club
9. Digimap – Ordnance Survey Collection
10. FAME (Financial Analysis Made Easy)
11. Foreign Broadcast Information Service (FBIS) Daily Reports 1974 – 1996
12. HeinOnline English Reports
13. HeinOnline Law Journal Library
14. The John Johnson Collection
15. LexisNexis LexisLibrary
16. Mintel Oxygen Academic
17. Radio at BUFVC
18. Reading lists (i.e. Talis Aspire)
19. RefWorks
20. Television and Radio Index for Learning and Teaching (TRILT)
21. TVTiP : TVTimes Project 1955-1985
22. UK Data Service

In a 2010 blog post, Matt Durrant of Bath Spa University suggested four methods for tracking down WAYFless URLs:

1. Asking the service providers for WAYFLESS url’s

2. Visiting the Federation site to see if they have instructions for particular service providers

3. Visiting the websites of Shibboleth run institutions, before copying and altering their links to fit

4. Using a Firefox add-on called ‘HTTP Headers’, allowing me to trace WAYFLESS URL’s during the authentication process.  URL’s generated using this method however have to be constantly monitored as they are not stable or supported by the service providers.

I think we can now add a fifth:

5. Using WUGen, the WAYFless URL Generator.

When I first looked at WUGen it was a bit ‘rough and ready’ – it seems to have been developed more recently into quite a useful tool. The UK Access Management Federation’s best practice document on WAYFless access divides WAYFless URLs into two types:

• The more reliable Service Provider (SP) side, or “controlled” WAYFless URLs, built on the service provider’s web domain and including our EntityID as a URL variable to identify the University of Lincoln;
• Identity Provider (IdP) side, or “uncontrolled” WAYFless URLs, built on the institution’s web domain and single-sign on (SSO) endpoint – in our case, this will be https://idp.lincoln.ac.uk/oala/sso

It recommends that service providers implement a recommended naming convention for SP-side WAYFless URLs, and that libraries favour SP-side URLs over IdP-side URLs, which are apparently less stable but which can be used “as a temporary measure until the service supports the SP-side form”.

I’d add a further important requirement for WAYFless URLs: they must permit deep linking to the component parts of an e-resource (e.g. to individual journal articles), not just to the resource home page or designated “login point”.

Twenty minutes spent hacking URLs, and we get the table below, including WAYFless URLs for all 22 resources.

(N.B. a lot of these URLs won’t work yet! Don’t get carried away. We have not registered our new EntityID with many of these service providers, so actual authentication won’t work.)

• 14 out of the 22 WAYFless URLs are IdP-side, which is the less-desirable format according to the UK Federation.
• None of the SP-side URLs follow the recommended format, as far as I can see.
• 8 out of the 22 look as if they will not be amenable to deep-linking. This may or may not be a problem, depending on the nature of the resource – it may make access via our link resolver problematic.

Can anyone improve on these?