You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
(no change necessary): push_back should be UB when the capacity is exceeded
(no change necessary): free-standing header
(no change necessary, already fixed on master): "The copy/move constructor is duplicated. cbegin, cend do not have const overloads. Has member and non-member swap."
So AFAICT there is nothing to do here, except for summarizing the meeting notes for LWG. In particular, I'll add the following to the changelog at the top of the document:
note LEWG decisions about push_back and free-standing in the changelog
include concern in the changelog about the inability to handle errors due to UB
include polls about whether this should go in C++20 or in Library Fundamentals v3
The recommendation seems to be to forward the revised document to LWG, I will do it for the next mailing (or if I make it, post-meeting mailing):
@tituswinters@AlisdairM I think there was a concern raised by @AlisdairM about whether this can be replaced with a vector and a stack allocator. It is unclear to me whether this resolved itself during the discussion, or whether this issue is still open and I should raise this concern to LWG.
The text was updated successfully, but these errors were encountered:
@tituswinters also mentioned a couple of times in the minutes that the proposal does not feel fully baked.
I'm as surprised from the votes wanting to put this in C++20 as from the votes wanting to put this in Library Fundamentals TS v3.
Unless I missed it, neither LEWG nor LWG have gone through the wording in detail. Shouldn't one of the groups, either LEWG or LWG, go through the wording in detail to make sure that this is the wording we want, before deciding where to put it?
Did this happened already and I missed it or is the wording "good enough" ? Or is this the task of some other group ?
This is my first proposal and I feel a bit lost about how the steps being proposed fit into the process.
The feedback from Kona is that:
So AFAICT there is nothing to do here, except for summarizing the meeting notes for LWG. In particular, I'll add the following to the changelog at the top of the document:
The recommendation seems to be to forward the revised document to LWG, I will do it for the next mailing (or if I make it, post-meeting mailing):
@tituswinters @AlisdairM I think there was a concern raised by @AlisdairM about whether this can be replaced with a vector and a stack allocator. It is unclear to me whether this resolved itself during the discussion, or whether this issue is still open and I should raise this concern to LWG.
The text was updated successfully, but these errors were encountered: