Any and all feedback on the use strict 'sql'; presentation would be welcome. It still needs some work and naturally, since it's only slides, much of the context is absent. Plus, I biffed a few images and need to clean up the KPresenter HTML. Hopefully, I can get this done before Wednesday.
Miscellaneous notes for context:
The idea that I am working on is combining a very lightweight persistent object system with phrasebooks. This is hardly a novel solution, but what if the phrasebook SQL is built dynamically and cached to disk? Each object knows its table and column names and by running a registration system, each object can register its dynamic SQL via a registration class. However, in production, the registration class is not used and instead the cached copy of the SQL is used. If you need to change something in the database, you can update the information in one spot, reregister the changes and everything automatically picks it up. The SQL would be readily available to hand to a DBA and much less dynamically generated SQL would be required at runtime.
This would be very useful for building projects with much refactoring or redesigning an existing database, but might have less utility with a larger system that's already in place and fairly static. Also, trying to come up with a clean, simple API is going to be a challenge. Is it even worth it? Would anyone be interested in something like this?
Some of the POOP modules seem to have performance issues because they try to do so much. Currently, my core code is only about 400 lines, it's fast and wouldn't grow too much. I think it's a good idea, but it can easily be brought down by a poor implementation.