i don't love adding the dependencies in a separate diff, would like to see them added when they're actually being used

Neither do I. But I would rather have critical diffs like D7022 be smaller so that a reviewer can have a more focused context when reviewing; and I think that alone is more valuable than consolidating diffs.

so that a reviewer can have a more focused context when reviewing

Yeah personally prefer dependency diffs/simple refactors to be their own diffs so they can quickly be accepted and moved on from.. also reduces clutter from "meatier" diffs

i frankly don't think this adds much clutter at all... also it's more tedious for me to switch between diffs to see how these dependencies are being used. just looking at this diff, i have no idea what any of these dependencies are for besides the one that @jon annotated, and even then i still have to look at another diff to see what handle_pake_to_grpc_err looks like. the lack of context in this diff makes it difficult to determine if a given dependency is appropriate or propose an alternative.

In D7020#208568, @varun wrote:

i frankly don't think this adds much clutter at all... also it's more tedious for me to switch between diffs to see how these dependencies are being used. just looking at this diff, i have no idea what any of these dependencies are for besides the one that @jon annotated, and even then i still have to look at another diff to see what handle_pake_to_grpc_err looks like. the lack of context in this diff makes it difficult to determine if a given dependency is appropriate or propose an alternative.

That's fair

I agree with @varun