Whilst XCRI's baselined project plan may have been a little over-ambitious in expecting XSD and WSDL specifications for Q4 2005, there has been no deviation from the orginal plan of shifting into a build phase for Q1 2006. Against that target, the level of interest and activity has exceeded expectations, with development activity stretching from Glasgow (N) to Exeter (S) and Liverpool (W) to Hertford (E):

• Phosphorix have been piloting with early drafts of the XCRI standard for aggregating learning opportunities in the Liverpool Learning Matrix, East of England Lifelong-Learning Support and Icebox projects. [Find out more...]
• Alan Paull of APS Ltd is working with Plymouth University using the XCRI standard to upload their entire course catalogue to the UCAS database [Find out more...]
• Michael Aherne has developed a prototype, XCRI-compliant catalogue for Reid Kerr FE College in Renfrewshire, Scotland [Find out more...]
• Paul Walk of London Metropolitan University is developing a web-service interface for searching and managing a repository of valid XCRI documents
• Ben Ryan of Kainao.com is working with the XCRI schema to develop prototype repository and authoring features tailored towards the FE sector
• Oxford University are using XCRI XML to transfer course information from their central courses database to the careers service

Details of the first three initiatives appear below. Full details of the others will appear in subsequent postings.

Phosphorix' XCRI deployments

Phosphorix is an Exeter-based software development company with extensive experience in web-based software. Keen advocates of Open Standards, Rapid Application Development and eXtreme Programming, Phosphorix have made significant contributions to educational and 'not for profit' organisations with the launch and release of a range of interoperability middleware solutions known as ioNode servers and systems.

XCRI's presentation at the CETIS Enterprise SIG meeting at Oxford in March 2005 alerted Phosphorix' Founding Engineer, Selwyn Lloyd, to the possibilities for co-evolution of the XCRI standard with his work for two JISC Regional Pilots aggregating information about learning opportunities. Close dialogue, including meetings with Roger Clark and Anthony Beal of the Liverpool Learning Matrix project, maximised the benefit of this co-evolution. Although timings inevitably meant that versions of XCRI XML used in Phosphorix's deployments were pre release 1.0, summaries of the architecture and lessons learned should be nonetheless interesting.

Links below provide summaries of Phosphorix' XCRI deployments. Documents reveal how the eXtreme Programming philosophy of getting something going with "noddy" XML, knowing that it would be replaced later, was well-suited to the challenge of working with the emerging XCRI schema.

• XCRI implementation in the Liverpool Learning Matrix project (Download PDF)
• Wider XCRI deployment in Phosphorix-supported projects (Download PDF)
• IoAtom: a secure, reliable messaging infrastructure used to carry XCRI messages (Download PDF)

UCAS-Plymouth upload

UCAS and the University of Plymouth wanted to facilitate the database-to-database transfer of courses data for course entry requirements, in preference to re-keying the data using the Web-link facility (formerly Netupdate). UCAS wished to develop a long-term method for exchanging courses data with member institutions, alongside its existing methods for exchanging data about applicants. In the document below, experienced consultant, Alan Paull of APS, describes how he used the XCRI schema to facilitate this data transfer.

• Case study: uploading Plymouth's course information to UCAS using XCRI
• XSD used to transform Plymouth's couse information to XCRI
• UCAS pipe delimited format: a transform output from XCRI for data upload to the UCAS database

Overall, Alan felt the exercise demonstrated the utility of the XCRI Schema: "APS was able to map from the Plymouth format to the XCRI and, more importantly, back from the XCRI to the required UCAS format successfully, with no critical anomalies."

Lessons learned on this project will feed into subsequent development of the Release 1.1 schema.

Reid Kerr

Reid Kerr College is a Further Education College in Paisley, Scotland, about 6 miles outside Glasgow. It offers a range of academic and vocational qualifications, including Higher National Diplomas (HNDs); Higher National Certificates (HNCs); National Certificate (NCs); Professional Development Awards (PDAs); General Scottish Vocational Qualifications (GSVQs); Scottish Vocational Certificates (SVQs); Scottish Group Awards (SGAs) and National Qualifications at Access, Intermediate 1, Higher and Advanced Higher. These 400+ courses are advertised online and via a printed prospectus. The website allows students to:

• request copies of the printed prospectus
• search for courses according to subject area, mode of study and start date
• display further details on a course, including:
  • Owning Department
  • Course Level
  • Subject Area
  • Course Leader
  • Course Summary
  • Course Content
  • Career Paths
  • Entry Requirements
  • Articulation Details (highlighted follow-on opportunities)
  • Study Modes
  • Assessment Strategy
  • Course Comments
• apply online

The website is managed by a Web Developer, Michael Aherne. Course data is held in an Access database, which is used to create pages with course information dynamically. The courses database can be searched from the website using Active Server Page server-side scripting.

Michael is an active member of the CETIS Enterprise SIG and began the group's interest in standards for course information with a presentation at the Group's 6th meeting on Prospectus Information Standards. As a member of the XCRI partnership, Michael's work fed into the emerging XCRI schema, and following release of the 1.0 version, Michael was keen to see how much work would be involved in making Reid Kerr's prospectus catalogue XCRI-compliant.

In many ways, Michael had set himself an acid test for XCRI adoption: could a lone developer in an FE college generate XCRI-compliant XML for potential use by an aggregator?

Michael now explains how he approached the problem, and how long it took him:

"The process began with mapping our database fields to XCRI elements, which was relatively simple. Most of our textual data has a direct equivalent in the schema (eg "Course Content" becomes "Syllabus", "Course Leader" becomes "Contact" etc), and there's a sensible way to express each piece of reference data (eg course identifier, SCQF level etc).

The only problem element was "Articulation". The schema has a good mechanism for specifying particular courses for further study, but our data is too vague to fit in with this. To get around this I added an “articulation” element as an extension.

Our course codes are only unique within our own systems so it was necessary to create global identifiers by prepending "reidkerr.ac.uk/courses/" to the code. I also made up a simple scheme for SQA (Scottish Qualifications Authority) unit identifiers although ideally this is something that would be controlled by the SQA themselves.

The actual XML creation was very low-tech, and was done by adapting the page which creates our HTML output (course_summary.asp). This basically involved setting the content type to XML, adding a function to insert the correct escape codes into the text where necessary, changing the boilerplate HTML to XCRI tags, and shuffling the data into the correct order. I then used URL rewriting to map this to http://www.reidkerr.ac.uk/web/courses/<course-code>.xml

The whole process (apart from getting to grips with the schema!) took about two hours or so, with time split about evenly between data mapping and coding."