Staking rate targeting and specific rewards.

[thiolliere]

I_NPoS

The total payout is calculated using I_NPoS function defined in http://research.web3.foundation/en/latest/polkadot/Token%20Economics/#payment-details

I_NPoS is implemented using some a piecewise linear function. This piecewise linear function has been generated with a test and is tested not to differ from actual float implementation with an error of 0.2%

Staking api

• The previous era reward accumulated at each new session is removed as reward is now computed from P_NPoS
• a root call add_points_to_validator is added so ppl rewarded can call it from inherent mecanism.

TODO:

• add inherent call to add_points_to_validator so ppl are actually rewarded:

  • In the branch of validity checking, we reward:

    • a parachain validator for each validity statement of the parachain block that it issues.
      should be made in polkadot, probably near https://github.com/paritytech/polkadot/blob/master/runtime/src/parachains.rs#L741

  • In the branch of block production, we reward:

    • the block producer for producing a (non-uncle) block in the relay chain,
    • the block producer for each reference to a previously unreferenced uncle, and
    • the producer of each referenced uncle block.

• traitify timestamp instead of direct use, note that contract and aura are using it directly as well
• timestamp trait: what api is to be used, returning 0 if timestamp is not yet inserted ?

related to #2781

[gavofyork]

@AlistairStewart could you take a look (and/or anyone else that knows how this should work)

[gavofyork]

seems broadly decent to me.

[gavofyork]

@thiolliere doesn't build?

[gavofyork]

@AlfonsoCev could you have a quick check that these code changes look like they have the correct logic?

[gavofyork]

@kianenigma are you happy with this?

[kianenigma]

Suggested change

	/// The value take into consideration desired interest rate and inflation rate (see module doc)
	/// The values taken into consideration are desired interest rate and inflation rate (see module doc)

[kianenigma]

(maybe my suggestion, but the sentence is ambiguous)

[thiolliere]

I improved a bit, actually it meant that I_NPoS function is defined using some constant which are the desired interest rate on ideal inflation rate and desired inflation rate

[kianenigma]

Could be used in the future into support. Looks like something that other modules can also use.

[thiolliere]

do you think having just this functions as is is interesting or should we provide a structure like RationalU32 { num: u32, den: u32 }

[kianenigma]

Based on your recommendation I think a RationalU32 with proper Mul impls will cover what I recommended. 👍

[kianenigma]

I can't know the details of the math portion, but seems to be reasonably tested. Otherwise looks good.

[AlfonsoCev]

@thiolliere Observation is on inflation.rs, on the definition of function I_full: for full generality, we should have that I_full is equal to I_left when x <= x_ideal, and equal to I_right otherwise. Right now the division is set at x=0.5, which is the current value of the parameter x_ideal, but since we defined that parameter we should use it.