This also temporarily removes `coreLogger`'s config from the public API One reason is that we don't have many logs that could benefit from suppressing (see diff) so the experimental API would not be useful to the user yet.
Also, the default config was suppressing useful warnings. Those warnings _would_ have been dead-code-eliminated in production mode anyway, so having a separate config to suppress them in dev mode makes it confusing.
Fixes P-171
Addresses the lack of options we currently have for surfacing issues in
our application via debugging tools. Prioritizes performance and
usability (visually) over clarity in some places that could have been
object mapped.
A logger with three separate audiences:
* `internal`: Logs for developers maintaining Theatre.js
* `dev`: Logs for developers using Theatre.js
* `public`: Logs for everyone
This logger supports:
* multiple logging levels (error, warn, debug, trace),
* multiple audience levels (internal, dev, public),
* multiple categories (general, todo, troubleshooting)
* named and keyed loggers (e.g.
`rootLogger.named("Project", project.id)`)
* console styling with deterministic coloring
* console devtool maintains accurate sourcemap link to logging origin
(e.g. `coreExports.ts:71` as opposed to `logger.ts:45` or whatever)
* swappable logger
* customizable filtering
* Accepts lazy `args`: `args: () => object` via
`logger.lazy.<level>("message", () => <expensive computation>)` (e.g.
`logger.lazy.debugDev("Loaded project state", () => ({ save: bigProject.exportToSaveable() }))`)
* Support multiple/nested sheets and sheet instances
* Add playground for instances
* Fix playground and example
* Change r3f's objectKey to storeKey to avoid confusion
* Update all editable/uniqueName variables used as store key to storeKey too
* Fix lint warnings
* Make editable schema based and fix a couple UX issues
* Refactor the icons a bit
* Add support for points, lines, line segments, and line loops
* Adjust nudge multipliers
* Fix types
* Fix helpers not showing on hover in some cases
