It used to be said all applications evolve until they can handle email, but maybe that should be CMS, or CMS and email. I am fairly certain that 'bespoke CMS' is consuming enough developer-hours to largely account for the 'developer shortage' we have. If you supply an extra dozen developers, eight of them will end up working on another bespoke CMS.

Off the shelf CMS ends up being a mix of feature factories and lock-in traps, though, which probably explains why NIH is so particularly pernicious in this case.

We need a SQLite for CMS.