LWG Poll 24: P1739R4 Avoid template bloat for safe_ranges in combination with "sub… #3777
Conversation
| If \tcode{decltype((F))} does not model | ||
| \tcode{\libconcept{convertible_to}<D>}, | ||
| \tcode{views::take(E, F)} is ill-formed. |
There was a problem hiding this comment.
"model" + ill-formed don't work together; advised LWG in http://lists.isocpp.org/lib/2020/02/15451.php
| If \tcode{decltype((F))} does not model | ||
| \tcode{\libconcept{convertible_to}<D>}, | ||
| \tcode{views::drop(E, F)} is ill-formed. |
There was a problem hiding this comment.
"model" + ill-formed don't work together; advised LWG in http://lists.isocpp.org/lib/2020/02/15451.php
| If \tcode{decltype((F))} does not model | ||
| \tcode{\libconcept{convertible_to}<D>}, | ||
| \tcode{views::counted(E, F)} is ill-formed. |
There was a problem hiding this comment.
"model" + ill-formed don't work together; advised LWG in http://lists.isocpp.org/lib/2020/02/15451.php
source/ranges.tex
Outdated
| then \tcode{((void) F, \placeholdernc{decay-copy}(E))}; | ||
|
|
||
| \item | ||
| otherwise, if \tcode{T} models |
There was a problem hiding this comment.
another "models"; see http://lists.isocpp.org/lib/2020/02/15452.php
(similarly for the other occurrences below)
This seems to make the entire paper unimplementable, and it thus should be rejected.
|
@zygoloid , this paper incorrectly uses "models" pervasively, which makes it essentially unimplementable. I've emailed LWG, but I'm not sure if anything here can be resurrected editorially. |
jensmaurer
added
changes requested
|
P2101R0 (applied with LWG poll 10) adds "If the validity or meaning of a program depends on whether a sequence of template arguments models a concept, and the concept is satisfied but not modeled, the program is ill-formed, no diagnostic required." |
jensmaurer
removed
changes requested
zygoloid
force-pushed
the
motions-2020-02-lwg-24
branch
2 times, most recently
from
1367c8e to
462a47b
Compare
…range-y" view adaptors Also fixes NB DE 288 and US 272 (C++20 CD). [range.counted] Moved note to work with new wording.
zygoloid
force-pushed
the
motions-2020-02-lwg-24
branch
from
462a47b to
12a6300
Compare
I encountered this when editing in a different paper. I find the current approach quite problematic from a comprehensibility standpoint, and I said this in person at Prague too, but it's too late in the cycle to fix this for C++20 (would require someone going through all the library wording and fixing all the places that say "models" when they mean to apply a machine-checkable set of requirements). |
…range-y" view adaptors
Also fixes NB DE 288 and US 272 (C++20 CD).
[range.counted] Moved note to work with new wording.
Fixes #3726.
Fixes cplusplus/papers#503.
Fixes cplusplus/nbballot#284.
Fixes cplusplus/nbballot#268.
Notes: