feat: Graceful frontend extension initialization failure

[SychO9]

Part of flarum/issue-archive#85

Changes proposed in this pull request:
Instead of allowing the application's frontend to crash, we can catch the frontend errors when initializing extensions and gracefully fail to minimize the impact and improve user experience.

Reviewers should focus on:
The approach, error message formulation, general code quality.

Screenshot

Before	After

Necessity

• Has the problem that is being solved here been clearly explained?
• If applicable, have various options for solving this problem been considered?
• For core PRs, does this need to be in core, or could it be in an extension?
• Are we willing to maintain this for years / potentially forever?

Confirmed

• Frontend changes: tested on a local Flarum installation.
• Backend changes: tests are green (run composer test).
• Core developer confirmed locally this works as intended.
• Tests have been added, or are not appropriate here.

[askvortsov1]

Shouldn't we rethrow e so the original error is accessible?

[SychO9]

The error is still accessible since we display it with console.error in the introduced helper. (see screenshot above)
If we throw the execution halts and the app will display the nojs content:

framework/framework/core/views/frontend/app.blade.php

Lines 27 to 36 in b64003c

	try {
	flarum.core.app.load(@json($payload));
	flarum.core.app.bootExtensions(flarum.extensions);
	flarum.core.app.boot();
	} catch (e) {
	var error = document.getElementById('flarum-loading-error');
	error.innerHTML += document.getElementById('flarum-content').textContent;
	error.style.display = 'block';
	throw e;
	}

[askvortsov1]

Makes sense!

[askvortsov1]

Why keep this untranslated?

[SychO9]

Didn't think it would need translation since it's a console message along with English error messages. These will most of the time be copy-pasted to extension authors.

[dsevillamartin]

I agree with that there's no reason to translate this.

[askvortsov1]

Ooo I like this!

[dsevillamartin]

Seems to work fine.

---

I do wonder if there's something we could do to make it able to be monkey patched by e.g. Sentry in order to report these errors. Only way these errors would be caught now would be with console logging enabled.

Think that's an issue more in general with exporting functions, however. I believe we've had some issues relating to that in the past? Unsure. Just some thoughts.

[SychO9]

Good point for a good use case. But yeah if it were to be done, monkey patching the function itself would be the best approach, which yeah isn't possible right now, but will be once the export registry is tackled, which is on the 1.x roadmap.