Ryan Burns

One more quick question, do I need to do something to get decomp-permuter to define PERMUTER when preprocessing? The example you posted just seems to take the else path when...

Ah, that's probably why. I need to keep the mwerks-syntax fixed-address declarations when computing the score, since these fixed-address declarations have different codegen than `*(type*) 0xaddress`. So that only works...

> I can imagine many cases where you care about the order Can you list some? Maybe I'm just drawing a blank here but I feel like this is unusual...

Thanks all for your feedback! I've updated the RFC to use the simplified order-preserving desugaring.

> Let's say the order is random. Each time you evaluate a derivation that uses the syntax, you would get a different derivation. > Or perhaps it's not random, but...

@Synthetica9: > Not sure I like this. There currently is no seperator in lists, and this would add that. > > I'd propose a `list select` instead, something like: >...

@Infinisil, very cool idea, thank you! I think your proposal is a great improvement when constructing combining attrs into a unified environment such that order truly does not matter. When...