When a column sits for a while, there are more errors because there are more browsers running the same tests. It would be nice to have a metric conveyed of how many pass/fail results have come in. That may help how the row is understood. For example. If no new commits come in for 2 days to trigger a new column. There may be 2000 pass results for a sim fuzz, and that is why an assertion came in that otherwise wasn't noted.
When a column sits for a while, there are more errors because there are more browsers running the same tests. It would be nice to have a metric conveyed of how many pass/fail results have come in. That may help how the row is understood. For example. If no new commits come in for 2 days to trigger a new column. There may be 2000 pass results for a sim fuzz, and that is why an assertion came in that otherwise wasn't noted.