Adding experimental command to list validators #328
Conversation
|
WDYT about sorting the rows by total staked instead of by validator name? |
|
@mfornet done. |
ilblackdragon
requested review from
janedegtiareva,
vgrichina,
stefanopepe and
mikedotexe
|
Also @bowenwang1996 @frol interestingly, indeed near/nearcore#2259 is not needed, but seat price is actually not everything. We also cut some validators from their seats, because usually with given threshold it will be over 100 seats filled. To reproduce this logic, we actually need to reproduce the random shuffle that our EpochManager does... In this PR seat number is not always correct and doesn't add up to 100. PS. Will need to move logic to near-api-js in a separate PR - near/near-api-js#297 to track it. |
|
Super welcome feature! I feel limited by the structure of the CLI, mostly because it can only see "the now" and not in the past. Example of needed features:
On the other hand, I'm not sure that displaying the whole Validator RPC endpoint would be helpful without some time-related queries (e.g., look at past blocks, to know my node performance over time). From this point of view, I'd rather build an application to capture the time series, trying to catch the anomalies. |
ilblackdragon
changed the title
@ilblackdragon do we actually want to reproduce this logic? It seems very difficult. We not only need to implement rust's random number generator in javascript, but also find the random seed used there, which means we also need to reproduce the logic of epoch id computation. I suggest that we not do it. |
|
@vgrichina Moving into near/near-api-js#300 @bowenwang1996 Well, we need to make it reproducible for other clients to be able to reproduce... If we don't think our current random function is reproducible (and I looked into it - it's very complicated) - we should change it. |
Agree. Probably some simple shuffling would suffice. |
Ref #188
Command:
Current output:
Output:
Output: