upvote
15% faster on top of a base interpreter that itself got 40%-50% faster on the same timeframe.
reply