Conversion of Uint to Decimal too verbose

[apollo-sturdy]

As mentioned in the below comment:

          How can I convert Uint256 to Decimal256 without being verbose? Right now I need this:

Decimal256::from_atomics(weight,0).map_err(|e| StdError::generic_err("Decimal256RangeExceeded"))?

Originally posted by @FrankBoltzmann in #1186 (comment)

I have two questions here:

1. Is there a reason not to implement From<DecimalRangeExceeded> for StdError? Doing so would remove the requirement to map the error. If we do this we should also remember to do it for Decimal256RangeExceeded, SignedDecimalRangeExceeded and SignedDecimal256RangeExceeded.
2. Is there a reason not to simply implement From<Uint128> for Decimal, etc? I assume almost everyone doing these conversions are interested in doing from_atomics(x,0) rather than from_atomics (x,18) so it would be a good choice in my opinion.

[webmaster128]

Thanks for bringing this up again. I think it got kindof lost for no good reason.

Here is my latest cheat sheet. Turns out there is no good reason other than from_atomics(n, 0) is doing the job. But I agree, there can be a TryFrom implementation doing that. @aumetra could you look into that?

Regarding the error type: no strong reason other than adding everything to StdError got a bit ugly over time.

[webmaster128]

Part 2. is covered by #2075. I pulled out the error conversion questions (1.) into a separate ticket: #2076