With all the commerical and open source content management systems on the market, why would a library still choose to build their own? In 2006, the University of Houston Libraries did just that.Â Rachel Vacek discussed their rationale and effort inÂ Putting the Library Website in Their Hands: TheÂ Advantages and Challenges of a Homegrown Content Management System.
(Note: Rachel indicated her slides will be available on rachelvacek.com, but were not there as of this writing, or I didn’t go deep enough into her site.)
UHL chose to develop their own CMS primarily because they wanted a system based on their vision of what a CMS is and should do, rather than modifying someone else’s. UHL feels that the CMS should be a growing and changing system. They felt that by building their own system their staff would be able toÂ fix problems and incorporate customer feedback more quickly.Â They felt that by building their own they could build more custom modules. They felt that by building their own they could make greater use of microformats and metadata then many of the existing solutions offered. Their current versionÂ took about aÂ year to build.
Much of the presentation focused on the demo of the adminstrative backend. Â (Note: theÂ backend currently only works with Firefox) Rachel created a virtual library (subject guide) using an interface that hadÂ iGoogle / gadget functionality, including the ability to move content modules between the predefined three column format.Â One feature I liked was the ability to incorporate incomingÂ RSS feeds into an aggregated subject/news feed.Â There was also in indication that there is the ability to Â share the content out with other system through syndication, although specific examples were not demonstrated.
The system also allows content to be imported and integrated. Content currently and soon to be integrated include LibrayFind, Archon, Serials Solutions, Wordcat, Flickr, WordPress MU. and LibraryThing comments and reviews. Work is also being done to pull in Delicious. They also hope to use theÂ Worldcat API, a metadata search engine, the ability to add media, making it easier to move modules, and creating a platform for customer contributed content.
Rachel indicated that efforts are under way to release their code as open source sometime in 2009. Having gone through the process of releasing an open source application at a University myself, I appreciate the challeneges in doing so. However, the concern I expressed to Rachel was the fact that they were using ColdFusion and not a more ‘open’ codebase such as PHP. Â (Will not open the ColdFusion debate here…)
Lastly, Since Rachel came onboard in Houston after the project completed, she referred many questions back to Karen Coombs.