Nils Knappmeier
Nils Knappmeier
Sounds like a breaking change to me, and a pretty special use-case. It should at least be a feature switch. Even then... Couldn't you just add a post-processing step like...
I have become very cautious with changes and very cautious with the assumption that "nobody would use Handlebars this way". The lack of a formal language specification makes it difficult...
Oh, it just came to my mind that if this is *not* a bug, than it is a new feature that changes the language. I promised @wycats not to allow...
If you consequently use the `{{#with x as | y |}}` syntax, this should not be a problem I think, but I can see that it is confusing. I am...
Might be, yes I'm not 100% sure, but this is the only explanation I have
Yes, that was before I told everyone to point PRs to the 4.x branch.
This is an implicit goal for 5.0, but the issue description documents parts of the changelog for 5.0, so I am assigning it to the milestone
We could also check if superclasses are whitelisted, so you only need to whitelist "A" in order to get properties from "A" and "B" accepted. The other way around is...
We copied the typings from DefinitelyTyped at the beginning of the year, so I'm not sure about the reasoning behind this structure. Maybe the git history of the DefinitelyTyped repository...
Mhm, even with deprecating this feature, it should still function correctly in 4.x versions. I actually think, with #1571 merged, the current state is wrong (because the type for Decorators...