VuFind 2.0 Conference - Was a Multi-LMS implementation of interest?

My primary aims in attending the VuFind 2.0 conference were 1) to learn from other implementations of VuFind and apply best practice to iFind Discover 2) to take advantage of the assembled programming talent to solve problems and improve performance 3) to learn more about where VuFind was headed 4) to gauge the potential 'market' for multi-lms implementations of VuFind and 5) to try and get my VuFind developments into the trunk so that they could be improved and supported by the community, thus increasing the resilience and sustainability of the iFind Discover project. Here are my thoughts on how each aim eventually panned out:

To learn from other implementations of VuFind and apply best practice to iFind Discover
One of the great things about VuFind is that it is incredibly customisable - this allows each institution to really create a discovery platform which matches its customer profile. Of particular interest to me was the consortia code developed by CJ O'Hara, Keith Dedman, Mark Noble at opac.marmot.org, the Bookbag functionality developed by the team at York University in Canada and a similar implementation by West Michigan University. Beyond VuFind but intimately related to the discovery process, I was also amazed by the work on concordances developed by Eric Lease-Morgan and the Fascinator project demonstrated by Greg Pendlebury. Unfortunately, this customisable strength is also one of VuFind's greatest weaknesses - in essence, every installation of VuFind has the possibility of being a standalone product which makes upgrading to the latest trunk extremely problematic.

To take advantage of the assembled programming talent to solve problems and improve performance
Thankfully, everyone on the conference was very sociable and keen to exchange ideas. We were able to discuss everything from user interface design to useful programming tools and improving system performance. From this viewpoint alone, the conference was a complete success.

To learn more about where VuFind was headed
It was great to learn that Joe Lucia and the VuFind team are determined to manage the project in a professional manner and that they have begun to develop processes which will make VuFind upgrades and developments less 'ad-hoc' and model focused. There was an incredible session on the last afternoon where all the desirable additions and developments were thrown into the ring, debated and discussed and assigned to specific strands and workflows. The VuFind 2.0 roadmap is the outcome of this session.

To gauge the potential 'market' for multi-lms implementations of VuFind
After talking to delegates, it seems that the major focus (at least in the US market) is not for a multi-lms approach to the discovery layer. This obviously has massive implications for the work that I have been doing and for the possibility of getting my code into the trunk. From what I could ascertain, there is a far stronger momentum in both the academic and public sector for implementations of a single joint open source consortia lms. Evergreen is the buzz word and it seems that VuFind hopes to present itself as its OPAC of choice. The code developed by Marmot may therefore become crucial to the success of VuFind.

To try and get my VuFind developments into the trunk
Given the market focus on consortia, this seems unlikely though my code is being looked at by some interested parties. Having met such talented people at the conference, I was also made more aware of the fact that as a "Shambrarian" and a "masher" rather than a skilled librarian or developer, that my code would need a lot of TLC by some professionals in order to get it up to scratch.

Luke