You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
There are periods were there's both a high throughput of PRs waiting for merge and lots of shit that's broken. In those cases, we would like to prioritise fixes for non-deterministic issues, however not necessarily to crater PR throughput by moving them higher the priority scale (as the mergebot batches PRs of the same priority).
PR ordering should have a second priority which says that the PR should be in the next staging of that priority (if possible), could probably just be a boolean added to the ORDER BY clause.
A more complicated version would be to multiply the priority by 10, but keep the batching on just the decade. This way we have room to prioritize PRs inside a level. However we'd have to update the default values to add 5, such that 0 => 5, 1 => 15, 2 => 25 in order to be able to set a lower (or higher) priority within the same decade.
The text was updated successfully, but these errors were encountered:
There are periods were there's both a high throughput of PRs waiting for merge and lots of shit that's broken. In those cases, we would like to prioritise fixes for non-deterministic issues, however not necessarily to crater PR throughput by moving them higher the priority scale (as the mergebot batches PRs of the same priority).
PR ordering should have a second priority which says that the PR should be in the next staging of that priority (if possible), could probably just be a boolean added to the ORDER BY clause.
A more complicated version would be to multiply the priority by 10, but keep the batching on just the decade. This way we have room to prioritize PRs inside a level. However we'd have to update the default values to add 5, such that 0 => 5, 1 => 15, 2 => 25 in order to be able to set a lower (or higher) priority within the same decade.
The text was updated successfully, but these errors were encountered: