Rename strict flag to strictMode

[NullVoxPopuli]

Turns out we've had a discrepancy between strict and strictMode 🙃

• RFC update (and context / background): Update RFC 496, typos, use correct field name (strict => strictMode) emberjs/rfcs#920

There is another option to rename all the strictMode things to strict, but I think this would be more disruptive.

[ef4]

This seems like potentially a good bug fix but the way you're describing and testing it here is deeply misleading.

There is never, ever a reason to use ensure-safe-component in strict mode.

The existing leaves strict mode templates alone test is all you need. If it turns out it's really supposed to use strictMode instead of strict, then that test can be updated to be spelled that way. But other tests beyond that are just testing the same thing in a weird and misleading way.

[NullVoxPopuli]

If it turns out it's really supposed to use strictMode instead of strict,

yeah, I've discovered that the RFC was the only one using strict -- I have an update to that here: emberjs/rfcs#920

[simonihmig]

For the record: I just synced with Preston, as I also came to the conclusion when debugging #1382 that the reason for this was this mix up of the name of the strict mode flag. In the addon's dist output I had strictMode, and as you explained the component helper should not even try to resolve its argument and complain about dynamic usage. But it did. And by patching resolver-transform.js and renaming strict to strictMode as in this PR the error was gone, and Embroider was correctly identifying the compiled template tag coming from the addon as a strict mode template, bypassing all those resolver checks!