[time.duration.nonmember] is missing formatting

[zygoloid]

More than half of the Returns: clauses in this subclause are in Roman rather than Teletype.

We also have inconsistency between this subclause and the next one: in [time.duration.comparisons](and in [time.point.nonmember]), we use italic teletype for the placeholder CT, but in this subclause (and in [time.duration.cast]) we use non-italic teletype for the placeholders CD and CR.

[tkoeppe]

@zygoloid: To clarify the complaint: Some itemdescrs are using \effects Returns rather than \returns, right? So the \effects should go away, too.

[zygoloid]

I'm not seeing any \effects in this section. Can you point to a paragraph where you see that issue? So far I'm only concerned by the issues I pointed out in the first comment.

[tkoeppe]

Never mind, I misunderstood!

[tkoeppe]

I'm unsure on how to proceed regarding placeholders. I am currently of the opinion that none of the use cases in utilities.tex should use placeholders, and the only legitimate placeholders I can think of are the F in do_F in locales and the key in fetch_key in atomics, where the placeholder actually stands for part of an identifier. By contrast, all the uses in utilities.tex are merely type aliases, usually introduced by something like "Let T be the type of...".

[jensmaurer]

Editorial meeting consensus: Do not italicize "CD" or "CT", these are actual types, not meta-placeholders.

[tkoeppe]

As a further example of what makes a placeholder, there is now the X in XTypes in [tuple.creation].

[jensmaurer]

I agree that CD represents a specific type (and thus should not be \placeholder), but CR(A,B) is a type meta-function that isn't actually C++, yet we use \tcode. Maybe we want to \placeholder CR here?

[zygoloid]

Expand the CR macro in its four uses.