Re: TrueType font changes size when rotated?

At 2:14 PM +0200 4/26/07, Tijnema ! wrote:

On 4/26/07, Richard Lynch <ceo@l-i-e.com> wrote:

On Wed, April 25, 2007 3:33 pm, Seth Price wrote:

I downloaded your image, enlarged it, and measured it out in photoshop. The distance from the center to the left side is 131px, and center to right side is 129px. Diameter to top is 129px, and to bottom is 130px. While it may not be an obvious ellipse, it's enough to throw off my

calculations.

And, really, if drawing text with a 1-pixel error margin throws you off, you're in trouble anyway...

It just shouldn't be like that :)

Yeah, in a prefect world it would be. But I think Seth, as well as the people he reported the bug to, are going to find that while we can compute exactly what any measurement should be, we also have to pass it through the user's browser and user's monitor to display it. I don't even want to get into how the various browsers display data, let alone how different monitors might. The combinations here are staggering. Granted, the OP is using a Mac and that comes as close as most can be, but even Apple's standards have been influenced by M$ (another OT subject). However, I'm actually surprised that the error was as small as it was -- we're talking about less than 0.5 percent by using his measurements. Also, please note that if you move the center of the image a single pixel to the lower-left, then the error is only a single pixel in one direction. Who's to say that the OP did not realize nor use where zero-zero actually was? I seem to recall that when one has to make a decision as to where a pixel is placed, we use the lower left corner as the coordinates and not the center of the pixel -- similar as we do for displaying characters -- is that not correct? In any event, if it were me, I would use the center of the locus of points to calculate error. Cheers, tedd -- ------- http://sperling.com http://ancientstones.com http://earthstones.com