[Cfp-interest 2626] Re: [SC22WG14.22998] n3082 wording on page 4

[Jim Thomas]

> On Jan 22, 2023, at 12:59 AM, Jₑₙₛ Gustedt <jens.gustedt at inria.fr> wrote:
> 
> Hello,
> this paper proposes wording on page 4 that has a suspicious "must"
> 
>     If an optional unary `+` or `-` operator followed by a signaling
>     NaN macro is used as an initializer that must be evaluated at
>     translation time …
> 
> If understand this correctly this is intended to indicate that this
> applies only to constant expression of that form that are constrained
> by the standard to be evaluated at translation time.

The intent is for it to apply to expressions of that form that are required by the standard to be evaluated at translation time. “Must” is intended to mean “is required to”.

> I'd propose to
> use the following replacement
> 
>     If an optional unary `+` or `-` operator followed by a signaling
>     NaN macro is used as an initializer that is constrained to be
>     evaluated at translation time …

Is “constrained” the right word? Maybe the clearer way to say it would be 

If an optional unary `+` or `-` operator followed by a signaling NaN macro is used as an initializer that has static or thread storage duration or is declared with the constexpr storage-class specifier.

- Jim Thomas

> 
> But if it was just meant to indicate that this applies to any such
> constant expression whenever it happens to be evaluated (in an
> initializer) at translation time, the "must be" should just be
> replaced by "is".
> 
> Thanks
> Jₑₙₛ
> 
> -- 
> :: ICube :::::::::::::::::::::::::::::: deputy director ::
> :: Université de Strasbourg :::::::::::::::::::::: ICPS ::
> :: INRIA Nancy Grand Est :::::::::::::::::::::::: Camus ::
> :: :::::::::::::::::::::::::::::::::::: ☎ +33 368854536 ::
> :: https://icube-icps.unistra.fr/index.php/Jens_Gustedt ::

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.oakapple.net/pipermail/cfp-interest/attachments/20230122/950db86d/attachment.htm>