Bump version to 1.0.0 globally #2149
Conversation
gavofyork
added
the
A2-insubstantial
| @@ -109,9 +108,8 @@ pub fn elect<T: Trait + 'static, FR, FN, FV, FS>( | |||
| >>, | |||
| for <'r> FS: Fn(&'r T::AccountId) -> BalanceOf<T>, | |||
| { | |||
| let expand = |b: BalanceOf<T>| <T::CurrencyToVote as Convert<BalanceOf<T>, u64>>::convert(b) as ExtendedBalance; | |||
There was a problem hiding this comment.
CC @kianenigma : No extraneous whitespace within code.
| minimum_validator_count: usize, | ||
| config: ElectionConfig<BalanceOf<T>>, | ||
| pub fn elect<T: Trait + 'static, FN, FV, FS>( | ||
| validator_count: usize, |
There was a problem hiding this comment.
rename rounds to validator_count - to minimise use of jargon.
also wtf is this passing in a closure only to call it immediately?
There was a problem hiding this comment.
See the open discussions here. #1915 (comment)
I would personally not do that and pre-process the data in lib.rs and pass everything in as a vectors.
If you mean only validator_count; it was still a storage item and wanted to keep it consistent.
gavofyork
added
A0-please_review
There was a problem hiding this comment.
apologies for the typo fair. Rest looks ok. I would open a ticket to remove the closures compleatly. This could leas us into turning the phragmen impl a separate crate that is actually reusable, not something tailored to work with what the staking module is providing.
|
done. |
* Bump versionb globally * Rebuild and fix * Rename fixes * Rebuild * Minor fix and code formatting for validator election * Fix tests * More test fixes * Fix several bugs in phragmen elections. * Rebuild, remove pointless closures
Also some minor changes to staking/phragmen:
Fn()->usize, only to always call it once: just pass ausize!rounds(which is incredibly opaque jargon taking from Phragmen's original context) tovalidator_count.min_validator_countrepresents the minimum number of validators we need, not one-less, so we require the number of validators to be>=before we declare code-red, not>.validator_countrepresents the ideal validator count. We only need to do candidate selection if we have more than that, so the comparison there should be>, not>=.