Assess the feasibility of removing the quadfloat requirement
quadfloat is not natively supported in clang. Currently, macOS/clang users are required to install NTL.
PALISADE currently uses extended double (regular double with an extended exponent) in parameter generation (for BFVrns* and obfuscation schemes) to support q that are larger than 2^1023. We may be able to solve the inequalities for parameter selection w/o using extended doubles [by working with logarithms].
Quad-floats are used to support RNS moduli larger than 2^58 (but less than 2^60) in decryption for BFVrns and evaluation for obfuscation. Explore if there is an algorithmic way to avoid quadfloats in this case.