Warn on `stat()` and `..var..`-notation?

[yutannihilation]

After #3534, now we have 3 different ways to access the calculated variables:

• after_stat(var)
• stat(var)
• ..var..

Though it seems very difficult to remove other syntax than after_stat(), I think we should promote the most proper one a bit more eagerly to avoid confusion. Can we warn on the usage of stat() and ..var..-notation?

[thomasp85]

I think we should (promote it more aggressively), but I also think instantly deprecating the other flavours upon release is a bit too agressive. It will be mentioned in the release post, and all docs have been updated to use it (and stat() has been removed from the doc)

I think we should begin warning sometime after this release

[yutannihilation]

Thanks, fair enough. But, I'm a bit pessimistic about the impact of docs as I see many people even don't know the existence of stat() after 1.5 years passed since the release (version 3.0.0)...

Anyway, I agree with you it's not the time to add warnings in the upcoming release. Let's wait for a while.