William Woodruff

Similarly: `REPNZ` is undefined in front of `XCRYPT`, `XSHA`, and `MONTMUL`.

> This is one of a few instructions that I've wanted to test on real HW but I don't have a VIA CPU. Unfortunately not -- this is solely from...

> So, did they ran the hardware test or not? No tests yet.

> And 4 months later, did they finally ran the hardware test? Or this is no longer a real issue for you so that you stopped carrying about? This is...

Yeah, here are the big alternatives that I'm aware of: * PCRE/PCRE2: Stable, feature-equivalent with C APIs. Both generations of PCRE do backtracking, but are probably *much* better tuned than...

> I would assume that std::regex does if the string is available at compile time. I'm not sure though. If only we were so lucky :slightly_smiling_face: -- Hana Dusíková [has...

> an LRU cache for dynamic expressions Yeah, this is another source of performance woes with the current implementation -- `std::regex` objects are initialized on demand, which might mean (tens...

> Could you try and document which clients support this? Yep, I'll try to collect that. It's not a real standard, but the `,isnick` qualifier I mentioned above appears in...

This would _potentially_ be useful for #1484 as well, although not strictly necessary! For our purposes, a schema like this (at the absolute minimum) would be sufficient: ```json [ {...

Thanks for following up! I'll go ahead and opine on #1377 as well, but FWIW: our ideal output for an API would be nearly identical to the one currently produced...