On Wed, 21 Oct 2020 at 17:15, Scott Schurr <s.scott.schurr@gmail.com> wrote:
Hi Jonathan,

On Wed, Oct 21, 2020 at 6:10 AM Jonathan Wakely <cxx@kayari.org> wrote:
Re http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2020/p2234r0.pdf

Thanks for working on this, Scott. I approve of the overall goal, but have some comments on specific parts.

Thanks for your careful reading of the paper.  I appreciate your feedback on the suggested approaches.  I believe every one of them.  I'm happy to fix those examples which are fixable and remove any suggested approaches that are simply not workable.  Would you be willing to help with that?

I've got a lot on my plate already, but I can try. I've made a mental note to keep an eye out for discussions and revisions of the paper now that I'm interested in it.

I don't think anything in the paper needs to be removed outright. My comments were meant more to dampen enthusiasm for radical changes (sorry! :-) than to reject anything outright. My objections don't necessarily need to be added to the paper, but might be useful input for some of the work done in small teams as suggested at the end of the paper.



And, yes, I completely agree that participation and review by optimizer writers is critical to the successful outcome of the audit.  I'm delighted that you seem to have a sense for who these folks would be for gcc.  Of course we'll need clang and Visual Studio optimizer writers involved with the review as well.  Yup, that will certainly influence how quickly the work can proceed.

I'm quite happy that you approve of the overall goal.  Motivating the audit (or something like it) was really the point of the paper.


Consider me motivated.