[webkit-reviews] review requested: [Bug 21278] Rounding behavior of AffineTransform.mapRect should be documented : [Attachment 23995] New patch

[bugzilla-daemon at webkit.org]

Brett Wilson (Google) <brettw at chromium.org> has asked  for review:
Bug 21278: Rounding behavior of AffineTransform.mapRect should be documented
https://bugs.webkit.org/show_bug.cgi?id=21278

Attachment 23995: New patch
https://bugs.webkit.org/attachment.cgi?id=23995&action=edit

------- Additional Comments from Brett Wilson (Google) <brettw at chromium.org>
> But why is it more important to document this for mapRect than for
> mapPoint, which is equally ambiguous?

I didn't notice this one. I added a similar comment based on what the Mac
version does (lround). I left the 0.5 rounding ambiguous since that seemed like
overspecifying it.

> Also, I'm not sure that including the comment without any blank lines is
> readable. It might be nicer to have paragraphing that groups the comment
> with the mapRect function (either just the IntRect overload, or both of
them).

I agree with the whitepsace, but for some reason I thought you would prefer it
this way. I added separate comments for each since they actually have totally
different rounding behaviors.