Darrell Wright

Number parsing has a precise mode now, but performance has not been optimized

Right now I have put off doing SIMD for the Real number parsing and have got good performance parsing all the available significant digits of the result type, parsing the...

I would like to look at using `compute_float_64` for the non-constexpr code path of result types of double/float. Right now I am getting about 900-1800MB/s depending on context of the...

Is this a possibility? I would like to relicense using BSL for my library depending on this but I think date being MIT means that any binaries of my code...

The branches within 32B boundaries is the fix for an intel cpu defect that causes inconsistent performance dur to simple changes. https://github.com/kostya/benchmarks/pull/266

my understanding was the flag is because the microcode fix causes perf issues when conditionals are not aligned. the benchmarking issue is the difference can be large when something changes...

> Btw. I woudn't do any exceptions for microcode "mistakes" and would **not** allow things like `-mbranches-within-32B-boundaries`. The issue this fixes is that a large majority of intel CPU's cannot...

At is also faster than operator[] when I was checking. It has a const overload too, so if the object is const it would still work.

I am having a similar issue with receiving data as it comes in e.g. File uploads Maybe something that wraps the read_async/write_async events in the ioservice? Regarding the SSL, I...

The body is allocated locally and it takes time/memory. It would be great to have a per route limit or per app limit, but a definition is better than none...