We have hired Neophytos for our EU-funded iCamp project,
so Neophytos will be able to spend some resources for completing and
polishing it.
That's excellent, Gustaf.
I see nothing I disagree with in principle in your post.
Maintenance and development for application code is already much easier due to having implemented more and more stuff in the Tcl API, making it less necessary to write queries.
The big issue I see is that if we do decide to rationalize the datamodel doing so for both Oracle and PG will be a very large task because of upgrade scripts, mostly.
They're painful to write, painful to test, and testing's useless without meaningful real-world data to test with.
Coming up with a better application code environment is important, yes, but rationalizing the datamodel is equally important.
