Sorry, Tagir, I'm with Remi on this one. Baking conventions into a language is dangerous, and must be proven either canonical in some way or harmless. (Floating point tokens are canonical by reference to C, javadoc was a harmless design decision in the context of Java.) Tagir's proposals for indentation control are very clever but neither canonical nor harmless.
Jim's proposal is carefully designed to be as minimal and simple as possible. I expect most of the feedback will be, like Tagir's, to the effect that Jim has missed some crucial feature, so that if only the proposal were less minimal in some way it would be better. But minimality itself is a feature, one to be protected and defended, especially against committees of people each of whom has their favorite non-negotiable. That said, I will (in another message) argue (in a different way) that Jim's proposal is very slightly too simple, in a way that makes it fail to meet its primary goal. Spoiler: I think I can prove that Markdown code quoting is appropriately minimal in its design, in a way Jim's is not. — John
