Every file starts by saying what it is

A small metadata block arrives at the top of each .phlo file:

@ version:     1.0
@ creator:     q-ai.nl
@ description: Serverside WebSocket handler
@ extends:     model
@ type:        abstract class

Until now there was a single directive for one purpose, naming a parent class. A file can now choose its class name, be abstract, sit in a namespace, and carry a version, an author and a description.

The description is the part that changes the platform rather than the file. Once every library states in one line what it is, a browsable catalog becomes possible without anyone writing one: the list of available libraries is generated from the files themselves, so it cannot fall out of date, and turning one on is a click rather than a lookup.

That opens a direction worth following. A file that can describe what it is can also describe what it needs, which would make dependencies something the build resolves instead of something you remember. And a description that lives in the source is one that stays correct when the source moves, which a separate documentation file never manages for long.

The general principle is worth stating plainly, because it keeps recurring: the code should be the description. Anything written in two places will eventually be wrong in one of them.