Pino browser fixes

[davidmarkclements]

fixes #647 and #649

NOTE: MERGE and RELEASE hapijs/hapi-pino#79 FIRST

In #612 we added symbols to browser pino.

This is breaking backward compatibility, and rather than add additional bytes for a polyfill to the payload overhead that was already added by including all symbols I think we should remove the symbols from the browser as they were only included for a "very niche use case" in #612.

We can instead update hapi-pino to use Symbol.for since the serializers symbol is a global symbol.

This also includes the smallest possible polyfill for globalThis for a standard-compliant cross platform global, see #649 (comment)

[jsumners]

LGTM

[jfahrenkrug]

@davidmarkclements Not that I'm a reviewer or anything, but what do you think about adding a comment to polyfillGlobalThis() that links to https://mathiasbynens.be/notes/globalthis? I think if anyone comes across that code down the line they'd be grateful to have a thorough explanation of the zaniness that is going on ;-)

[mcollina]

I have a fear that this makes this change semver-major.
I would prefer it not to be.

[davidmarkclements]

yep. But I think this is a case of semver purity vs practicality. polyfilling symbols is extra overhead and not fully possible (e.g. typeof symbol). There is only one use case, and it's technically server side - so I think an argument can be made in this case for considering the breaking of browser compat a bug, and removing this to be the solution to that bug. Since this change along with the hapi-pino change will literally affect no one, as there are no other cases for pino.symbols in the browser.

[github-actions]

This pull request has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.