I'd like to weigh the pros and cons of keeping the changes made by #26.
Some of the cons I can think of:
- Greatly increased compile times
- Introduced methods to types that are possibly outside of the scope of this library
exponentOfE()
raised(to:)
squareRoot()
log()
- The distributions aren't very well documented. Their use cases aren't immediately obvious unless one goes through the related Wikipedia articles.
- Most developers using RandomKit will likely not need these features
An alternative solution I would like to propose is another library that has a dependency on RandomKit and features this.
This would be a breaking change, causing the version number for RandomKit to be bumped up to 3, which I'm ok with doing. If this change is made, other helpful breaking changes should also be made.
I would like to know what @phimage's opinion is on this, since he made the changes.
I'd like to weigh the pros and cons of keeping the changes made by #26.
Some of the cons I can think of:
exponentOfE()raised(to:)squareRoot()log()An alternative solution I would like to propose is another library that has a dependency on RandomKit and features this.
This would be a breaking change, causing the version number for RandomKit to be bumped up to 3, which I'm ok with doing. If this change is made, other helpful breaking changes should also be made.
I would like to know what @phimage's opinion is on this, since he made the changes.