[lib] Referring to standard library requirements. #1263
Comments
jensmaurer
added
the
decision-required
|
Editorial meeting consensus:
|
jensmaurer
removed
the
decision-required
|
[time.clock.req] p4.2 is an example of a long list of table references. Would this case be better served with a section reference to [utility.arg.requirements]? |
|
A lot of references to CamelCase requirements are missing cross-references. For example, CopyConstructible appears 54 times, only 14 of these appearances have a "ref" nearby. Do we want to have a cross-reference after all mentions of CamelCase requirements? |
jensmaurer
added
the
decision-required
|
@zygoloid opined that we'll reference the subclause (at the end) if we have a large-ish list of CamelCase requirements. Otherwise, we reference the table. |
jensmaurer
removed
the
decision-required
|
This is partially mooted by the ongoing concepts work for the standard library. |
|
See #2176 and the new section "Requirements expressed by concepts" on the wiki. |
...instead of "satisfy". Partially addresses cplusplus#1263.
...instead of "satisfy". Partially addresses cplusplus#1263.
...instead of "satisfy". Partially addresses cplusplus#1263.
...instead of "satisfy". Partially addresses cplusplus#1263.
First, we have a mixture of "shall satisfy" (89) and "shall meet" (61). What's the preferred phrase?
Second, for CopyConstructible etc., we sometimes cross-reference the table number and sometimes the section. What's the preference here?
Third, for CopyConstructible etc., we sometimes say "shall meet/satisfy the CopyConstructible requirements" and sometimes we say "shall meet/satisfy the requirements of CopyConstructible". What's the preferred phrase?
For iterator requirements, we're now fairly consistently saying "shall meet/satisfy the requirements of a blah iterator (xref to section)", which is good.
The text was updated successfully, but these errors were encountered: