I know, I know -- not fair of me at all to just vanish like that for months on end. I had a lot of real-life stuff, and that's made progress difficult on this project. Here's the biggest difficulties I'm facing now:
The installer is a mess, and has not been widely tested. It needs a rewrite to be cleaner internally.
Deployment is still "by hand". I'd like to put together a Docker container and keep that updated, so that it can be deployed to any Docker-friendly environment.
Template management is not what it could be:
Default templates are still hard-wired at installation.
It's not really possible to package assets with a template yet (e.g., images).
Exporting and importing templates is still very clumsy.
Pure Python code in templates (modules, basically) are not handled well by the editor or template manager.
Various bugs in the front-end, mainly around when the queue runner activates on a save.
The routine that performs page saves is a Big Ball Of Mud that cries out for a rewrite.
No unit tests.
It's hard not to look at all that and feel intimidated. I'm probably going to solve the easiest and most tractable problems first -- the front-end bugs, the save routine.
What I most need to do in the long run is create a new environment I can deploy to and test in, so as to work out the other problems in a way that's not constantly at odds with my actual workflow. My original test and deployment environment is a shared host that runs FastCGI -- roughly the same kind of environment most people would deploy something like WordPress in. I did that deliberately so I could see what would be feasible in such an environment and what would not be. But there's no question it's also making development harder than it has to be.
I'll try to post a little more regularly from now on.