Cursor movement in Linux 16610
Cursor movement in Linux 16611
Hey Tim, Yes and I agree with you. The captured image will look correct with the ClearType-sub-pixel hinting...
Incorrect. I thought I explained this earlier, but perhaps you missed it because of all the crap in this thread.
Consider an LCD pixel, consisting of a red sub-pixel, a green sub-pixel, and a blue sub-pixel. If you want that red sub-pixel to have a value of 0x12, the blue one to have a value of 0x34, and the green one to have a value of 0x56, you can get that by setting the pixel to 0x123456.
This works because there is a consistent mapping between pixels and groups of 3 sub-pixels. On a CRT, there is no such consistent mapping. Software generally can't know on a CRT what pixel a given colored dot belongs to, and, furthermore, a given pixel can actually affect more than one dot of a given color.
Because of the precise relationship between pixels and sub-pixels on an LCD, ClearType can calculate anti-aliasing based on the sub-pixels. The result of that calculation is a set of *pixel* values to send to the display that will cause that display to set the proper values for each sub-pixel.
If you take a screen capture, you get those same pixel values. When that screen capture is then displayed on an LCD with the same physical arrangement of sub-pixels in its pixels, all the sub-pixel values get set to the values ClearType wanted, and you see the full effect of ClearType.
Cursor movement in Linux 16612
begin virus.txt.scr MIMF (flatfish) Many as Aftab Singh, allisonhunt1968 plus 1, Anna Banger, anonymous, Archie Watermann, Baba Booey, Babu Singh, bill.gates.loves.me...
-- --Tim Smith