On Mon, 15 Sep 2025 21:44:35 GMT, Roger Riggs <[email protected]> wrote:
> Is there any observable/measureable performance difference? The byte type is > a bit of a hack, while char is semantically more precise. There is negligible performance impact. I filed this change not for performance, but rather simplifying the code path for readability because in the domain of DigitList, `digits` is a byte array and always composed of code points 48 to 57. With this context, passing it directly as a byte avoided an extra level of conversion and should be reasonable to the reader. In general, I understand the result of `(digit + '0')` as a char is semantically more precise; I can withdraw this PR if you think it hurts readability (since the goal is the opposite of that). ------------- PR Comment: https://git.openjdk.org/jdk/pull/27300#issuecomment-3294205443
