Provide text support in dplug:canvas and/or canvasity

An awful lot of the needed parsing of fonts is done in `printed:font`, the only missing things is rasterizing.
Rasterizing does exist in `dg2d`.

`printed:font` is two-pass, first it parses only necessary to do font selection.

`dg2d` demonstrate amazing text performance in its demo.

Create a future with `fillText`, that will go fast and avoid the infamous glyph cache red-black tree.

Possibility to fetch a `ttf` file data.
Option to parse system fonts?

Start by reading the existing font parsing in `printed:font` and text rasterization in `dg2d`, then compare how `dplug:canvas` and canvasity currently handle drawing. Done should provide a defined `fillText` path with fast rasterization, including a decision on loading TTF data and parsing system fonts.

Written by the indexing model from the issue text.