On Thursday, March 6th, 2025 at 16:48, Bjorn Reese via Std-Proposals <std-proposals_at_[hidden]> wrote:

>
>
> On 3/7/25 01:32, Nikolaos D. Bougalis via Std-Proposals wrote:
>
> > The proposal briefly discusses that. I don't think having a separate function makes sense.
>
>
> Consider a new clock type. This is how other "time source" are handled
> (e.g. utc_clock, file_clock, gps_clock.) Even if you disagree, the
> proposal ought to discuss this alternative.

That is a good idea. I briefly considered that when writing the proposal up, but I don't remember why I dismissed the idea.

It would certainly make sense to go for a separate clock and, perhaps, also go some way towards allaying the concerns that others have had about how this clock would interact with reproducible/deterministic builds.

Best regards,

Nik Bougalis