@jensmaurer It's certainly a trivial change, but I see the question of "are extended integer types integral types?"/"Do the standard conversions apply to extended integer types?" more often than "do the standard integer types include unsigned and signed standard integer types" (not seen this one yet, luckly :P).

Specifying standard/extended subtly provides more information to the reader in the context of that single paragraph, which I believe is useful as paragraphs in the standard are typically not read sequentially.

My first reaction was that this change is incorrect: we should just define that "integer types" are "standard and extended integer types". But it appears that bool, char, wchar_t, and charN_t aren't actually "standard integer types". That seems like a normative error to me. Consider the (very few) uses of "standard integer type":

[conv.rank]/1.5: "The rank of any standard integer type shall be greater than the rank of any extended integer type with the same width."
[conv.rank]/1.7: "The rank of bool shall be less than the rank of all other standard integer types."

These both obviously meant to include all non-extended integer types.

[utility.intcmp]/1, /5, /9: "Mandates: Both T and U are standard integer types or extended integer types (6.8.2)."

I have no idea what the intent was here.

This needs CWG and LWG attention.

@zygoloid , I think we can remove this from CWG scope, because we have

conv.rank p1.8: "The ranks of char8_t, char16_t, char32_t, and wchar_t shall equal the ranks of their underlying types"
and we have transitivity via p1.10, so those are covered.

We should fix "[conv.rank]/1.7: "The rank of bool shall be less than the rank of all other standard integer types."
to avoid the hint that "bool" is a standard integer type.

Admittedly, that leaves open the possibility that bool has greater rank than an extended integer type, but that feels like design intent.

Regarding "[utility.intcmp]/1, /5, /9: "Mandates: Both T and U are standard integer types or extended integer types (6.8.2)."", there is a note at the bottom that clarifies the intent:

[Note 1 : These function templates cannot be used to compare byte, char, char8_t, char16_t, char32_t, wchar_t, and bool. — end note]

So, except for "other" in conv.rank, I think this works as intended and we should simply close this pull request.

We chose the wording in [utility.intcmp] specifically because it doesn't include bool, char etc.

The wording matches the intent.