docs: add polyfill.md

[evolkmann]

The Event() constuctor is used in the project. To make this work in Internet Explorer, a polyfill is required.

The new document describes the polyfilling process.

PR Checklist

Please check if your PR fulfills the following requirements:

• The commit message follows our guidelines: https://github.com/scullyio/scully/blob/master/CONTRIBUTING.md#commit
• Tests for the changes have been added (for bug fixes / features)
• Docs have been added / updated (for bug fixes / features)

PR Type

What kind of change does this PR introduce?

• Bugfix
• Feature
• Code style update (formatting, local variables)
• Refactoring (no functional changes, no api changes)
• Docs

What is the current behavior?

Issue Number: N/A

Scully is currently not working in Internet Explorer.

What is the new behavior?

The new doc describes the necessary polyfilling process.

Does this PR introduce a breaking change?

• Yes
• No

Other information

Reference to usage of new Event():

scully/projects/scullyio/ng-lib/src/lib/idleMonitor/idle-monitor.service.ts

Lines 25 to 27 in 27b79dd

	private initApp = new Event('AngularInitialized', {bubbles: true, cancelable: false});
	private appReady = new Event('AngularReady', {bubbles: true, cancelable: false});
	private appTimeout = new Event('AngularTimeout', {bubbles: true, cancelable: false});

[SanderElias]

@evolkmann You are aware that those events are only used inside our puppeteer rendering process, and that that is inheritly not Ie?

[evolkmann]

Hi @SanderElias, I was not aware of that, but still it caused our app to crash. With the polyfill enabled, the error did not occur again. Here is a screenshot of the error report:

[SanderElias]

Thanks for the info. I will log an issue for this because those should not fire in normal operation. For now, the polyfill will fix your issue.

[SanderElias]

LGTM

[evolkmann]

Hi @SanderElias, thanks for merging. Just curious, is it planned to remove the firing of those events in the browser at some point? Or is it indeed right that this code is executed in the browser?

[SanderElias]

Yes, I made a note to take care of this, but it can take a while before I get around to that. That's why I merged your docs PR, so anyone running into this can fund out the cause in the docs.

[evolkmann]

Alright, thanks for clarifying!