new annotation or pragma to mark functions that are intended to be only used during compile time
Hipreme
msnmancini at hotmail.com
Thu Feb 20 15:30:59 UTC 2025
On Thursday, 20 February 2025 at 13:55:33 UTC, Ilya wrote:
> Hey, I did some measurements with my initial implementation and
> even marking the obviously CT-only things in our codebase gives
> ~7% compilation time improvement for some of the object files.
> That's not super exciting, but still something we'd like to get
> (we are annoyed by the long compilation times a lot).
>
> What are the next steps? This thread is super quiet... Does
> that mean there is not enough interest in this? Should I try
> proceeding with the DIP PR? Or it doesn't make much sense?
I would be very happy to have that in D. That idea is super cool
and I've never even thought that was a problem to be solved,
thanks for your input!
I have a plenty of functions that are called in compilation time,
and beyond reducing the object generation time, one would of
course get reduced binary size, which is also fairly important to
those which are targeting WASM.
More information about the dip.ideas
mailing list