As Dave said, it is not a difficult thing to do I have worked on different solutions from using a/pspell to webservices or even the Voyager keyword index …

The thing that frustrates me the most, and perhaps is one of the reasons why we've done so many of these things ourselves at Huddersfield, is that it often takes too long for vendors to add new features.

It's not as if some of these are technically difficult to implement — once we'd decided to use Aspell for our "did you mean", it only took 3 or 4 hours to complete the development.

In fact, I probably spent twice as long in meetings saying why we should do it :-)

I'm a strong advocate of libraries having their own developer (even if it's just a part-time position).

In an ideal world, OPACs would have been developed to be easily extensible. That way, just one customer would need to do the development, and the code could be easily and safely shared amongst everyone else.

I'm always more than happy to share our code, but it often requires setting up a separate web server and MySQL database, as well as making alterations to the OPAC pages (which may then cause issues with getting technical support from your vendor).

The fact that we haven't implemented one of these features doesn't necessarily mean we don't want it. Your theory that the vendors help push it across the adoption chasm is right, but in many cases it's because we're so dependent on them to supply the functionality. If the "did you mean" functionality had been there years ago, when libraries were already asking for it, it would have been widely adopted by now.