What's with the insistence of running it off of just MySQL? Textpattern being small and lean would IMHO be prime candidate for SQLite.
I used to use txp for my personal blog and couple of client sites back in 200x and have still soft spot for it. Every now and then I take a look but decide not to try it out because I do not have MySQL in my stack and I honestly can't be bothered to set it up just for txp. Any chance it'll going to change?
…which will open up a whole new world of possibilities.
The Textpattern dev team & user base is pretty small, and the user base is largely patient, so Textpattern can sometimes fall into a trap of being 'good enough for now' and go for extended periods of time with few commits. What tends to happen is a release is scheduled, takes place, and then the plans for the next release are more forefront in our minds. The most recent release was nearly two years ago, which is a long time in Textpattern terms, but I'm confident we can get Textpattern 4.9 into the release pipeline this winter. More on that here:
It's from before PDO was a thing, I believe. They have their own "abstraction" on top of mysqli - I'm not sure about the return values, but might be possible to switch in PDO?
Happy to answer any questions as best I can!
Further reading:
https://github.com/textpattern/
https://textpattern.com
https://docs.textpattern.com