[range.view] Add complexity requirement normatively #2592
Conversation
|
Please fold your fixups by "git rebase -i master" and force-push for easier review. |
jensmaurer
added
the
changes requested
JohelEGP
force-pushed
the
range_view
branch
2 times, most recently
from
16a5346
to
b3d69e6
Compare
jensmaurer
removed
the
changes requested
| semantic, the two are differentiated with the help of \tcode{enable_view}. | ||
| \tcode{T} models \libconcept{View} only if the complexity | ||
| for each initialization and assignment required by \libconcept{Copyable<T>} | ||
| is constant time. |
There was a problem hiding this comment.
This may well be intended, but I don't see how it's editorial. Is this somehow justified by the existing text?
There was a problem hiding this comment.
Paragraph 1 above: "The View concept specifies the requirements of a Range type that has constant time copy, move, and assignment operators...".
Although I wouldn't mind sending this to LWG: we need to add the missing requirement that destruction is O(1) as well which is certainly not editorial.
There was a problem hiding this comment.
Let's send the whole thing to LWG then. If they want us to apply the non-normative change editorially we can, but I think it's generally better for holistic changes with a normative component to be applied together.
|
If there's agreement that introductory paragraphs are normative, this is subsumed by #3453. |
|
I'm not sure that we have such agreement, or that ISO even allows that. It would be nice if that were allowed, and if we could have "(informative)" subclauses for introductions, but we tried to find permission for that in the ISO drafting directives and did not succeed. I'm noticing that we are using informative subclauses in the LFTS, e.g. https://cplusplus.github.io/fundamentals-ts/v3.html#general.plans. |
|
Subsumed by #3453. |
And drive-by move description above the definitions.
Also, could a \pnum be missing after the list in the example?