WorkflowMessage - Enable strict parsing of annotations

[totten]

Overview

Not sure it's a good or bad idea. Spot-checked some local tests, and they seemed OK. Let's see what Jenkins thinks.

[civibot]

(Standard links)

• If you are reviewing this pull-request, you may wish to consult the test sites and the Review Standards (long template, short template).

[eileenmcnaughton]

@totten jenkins is positively cheerful - but does it matter that @braders patch isn't merged yet in terms of this one?

[totten]

@eileenmcnaughton The two patches might interact. (Not at a git level -- but functionally.) Since I think the other is probably more important (this one was on a lark), I'd suggest merging #25790 first and then re-run the test here.

[eileenmcnaughton]

@totten ok - let's merge the other - @colemanw has been ignoring it for 3 days - I doubt 2 more will matter + it would fail tests if it was a problem IMHO.

[eileenmcnaughton]

test this please

[eileenmcnaughton]

@totten it still passes

[totten]

@eileenmcnaughton Wow, ok. I guess we can do this then.

• Flipping this to "strict" is better from a core dev POV -- i.e. more likely to prevent bugs.
• But, generally, "strictness" toggles should be available to downstream -- they represent a pressure-relief-valve for someone mixing versions of core/extensions in random ways.
• It's kind of hard to care about as a hypothetical.

IMHO, the stricter policy is better than the looser policy -- but if anyone ever gets bitten by it, then it's discussion-welcome/patch-welcome to improve "strictness" toggles generally.

[eileenmcnaughton]

OK - lets' merge this - I think stricter is better & we are better not to tie ourself in knots over what might be handy for extension writers - esp on functionality that probably isn't much in use in the wild