Since I do the vast majority of this development on a platform that doesn't have a database, testing was a problem. I've been spending time getting DBD::AnyData working and there were several interesting challenges but I finally got it running just a couple of days ago.

Big thanks to jZed for writing a mountain of code that makes it possible for me to implement a limited-functionality SQL database without using a C compiler, and for being very helpful with the snags I ran into.

Now I'm writing unit tests and have found a few tiny (in code size) bugs that were introduced in the group-handling rewriting and preventing the rather tricky boot-strapping of the node cache from working.

Given how successful the near-live testing previously went and how small the bugs I've found were, I once again feel I'm very close. Of course, this project often progresses very slowly (obviously), so calendar-wise it is hard to predict when the next near-live testing will begin.

Thanks for y'all's patience...