Improving performance of the Library with Cython

[Bhargavasomu]

What is wrong?

There are many places in the code like, finding an intersection of the tangent with the curve etc. In other words, there are many functions whose variables make use of fixed specific datatypes (mostly int). I feel that Cythonizing these functions would improve the performance.

Maybe tackle this after we have benchmarks for the library.

How can it be fixed

Make the suitable functions as cython functions.

[hwwhww]

Agreed! We did consider using Cython to improve performance; especially that Serenity BLS signature aggregation would be one of the highest overhead.

[carver]

Maybe tackle this after we have benchmarks for the library.

I would go as far as saying this is blocked on #49 -- there is no point in starting to optimize performance without a benchmarking suite.