As Sebastian said, this is probably best handled via an attribute.

You can do this today with #pragma unroll (on many platforms), although
generally you might get better results via PGO.

To the original proposal...

I can't think of any time I've said to myself "I desperately wish I had an
easier way to write this for loop". IMO, the way it works today is not
particularly onerous. And changing it just increases the cognitive load for
people who are new to the language.

Usually you either have an iterator you're working with and you don't care
about indexing (so use ranged for), or you're using indexing, and that's
either really simple or complex with oddball termination cases - I don't
see the win here. It feels like a Pythonism creeping in.

Simon




On Wed, Feb 12, 2025, 10:59 Sebastian Wittmeier via Std-Proposals <
std-proposals_at_[hidden]> wrote:

> That is the domain of the optimizer.
>
> Even inline functions does not mean any longer that they are inlined (just
> that they can be defined multiple times and put in headers).
>
> If you want to standardize a hint to the optimizer, a (standardized)
> attribute would be a good choice, as it does not change the meaning and
> behaviour of the code.
>
>
> -----Ursprüngliche Nachricht-----
> *Von:* Filip via Std-Proposals <std-proposals_at_[hidden]>
> *Gesendet:* Mi 12.02.2025 11:04
> *Betreff:* Re: [std-proposals] for-loops revision (simplification and new
> syntax)
> *An:* std-proposals_at_[hidden];
> *CC:* Filip <fph2137_at_[hidden]>; Proposals C++ <
> std-proposals_at_[hidden]>;
> Maybe to force the compiler to unroll the loop we could use:
>
> ‘’’
> inline for
> ‘’’
>
> I’ve seen similar syntax in zig.
>
> Cheers, Filip
>
> Wiadomość napisana przez Jonathan Wakely via Std-Proposals <
> std-proposals_at_[hidden]> w dniu 12 lut 2025, o godz. 10:46:
>
> 
>
>
> On Wed, 12 Feb 2025, 02:42 Henry Miller via Std-Proposals, <
> std-proposals_at_[hidden]> wrote:
>
>
>
> On Tue, Feb 11, 2025, at 12:03, Pau Miquel Montequi Hernandez via
> Std-Proposals wrote:
> > @Andrey Semashev
> >
> >> Quite honestly, looping with a counter is by far not the most common
> kind
> >> of loops you write in real code. It is the simplest kind of loop that
> they
> >> teach in clases
> >
> >
> > Looping with a counter might not be the most common kind of loop written,
> > but indeed it is the simplest kind of loop that they teach in classes.
> "*Make
> > simple things simple*" is a good principle that improves readability and
> > teachability of the language. Also this proposal can be seen as a way to
> > homogenize the loops, if something looks like a range, should be iterable
> > as a range; right now C++11's range-for loop is used for many constructs
> > that look and behave like ranges but leave aside the most simple of them:
> > just iterate from number A to nomber B, in my opinión this is an
> oversight.
>
> many developes have adopted a no loops ever policy. It turns out everytime
> you think loop you can reach for std:: where you will find an algorithm
> that is more expressive and safer to get the same result. I'm not 100%
> there but I'm close.
>
> As such the only use I have for loops is teaching how things work inside.
> for students being explict about exactly what is happening is comforting
> even if most often the loop always looks the same. Perhaps the syntax could
> be improved but is it worth making it easier to use something rarely used?
>
>
> Does adding several new ways to write a loop make things easier?
>
> Especially when some of the suggestions add or remove parts of the usual
> syntax. It seems to me that it would make it harder to read the loop header
> because you'd have to examine it more carefully to see what it was doing
> (especially the suggestion for multiple init statements).
>
> And it's all just syntactic sugar, not enabling anything new which can't
> be done (quite easily!) today.
>
> The range-based 'for' loop added in C++11 actually enabled something new
> which couldn't be done in C++03: iterating over arrays or containers with
> uniform syntax, without needing to write 'typename R::iterator' or
> 'typename R::const_iterator'. And even that became redundant with 'auto'
> and ' std::begin' added at the same time as the new 'for' loop.
>
>
>
>
> --
> Std-Proposals mailing list
> Std-Proposals_at_[hidden]
> https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals
>
>
> --
> Std-Proposals mailing list
> Std-Proposals_at_[hidden]
> https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals
>
> --
> Std-Proposals mailing list
> Std-Proposals_at_[hidden]
> https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals
>