On Jan 20, 2017, at 4:30 PM, Brian Burkhalter <brian.burkhal...@oracle.com> wrote:
>> That is worrying me since I don't follow these lines are part of that:- >> 1684 // Move to the position (xoff, yoff). Since l-is >> used >> 1685 // to index into the scanline buffer, the >> calculation >> 1686 // must be limited by the size >> 1687 l += xoff + yoff*width; >> 1688 l %= width; >> 1687 was already there but 1688 and the comment are new and 1688 looks wrong >> to me >> as it would seem to throw away the y it just added in ... > > Indeed, if xoff is in the half-closed interval [0,width), then (xoff + > yoff*width) % width == xoff. This does not however account for the accumulation into “l” which might negate my observation. Brian
