How can I get epsilon value for Decimal

https://en.wikipedia.org/wiki/Machine_epsilon

Hi @deterok, do you have a use case for requiring this?

The primary reason an epsilon isn't exposed right now is because unlike a floating point number which uses approximations for representing numbers we instead use a fixed precision format. If we were to instead define it as the smallest representable number then technically the machine epsilon would be either 5 × 10^−28 or 10^−27 since the maximum whole precision for this library is 28 (i.e. 29 is supported in some cases but not completely). However it'd be useful to understand the use case for this first to better answer the question.

I guess we can close this stale issue :sweat_smile: