Resolution indpendence

David De La Harpe Golden david.delaharpe.golden at
Fri Jun 27 19:51:51 PDT 2008

Glynn Clements wrote:

> If you start from the fact that hand-crafted bitmaps are more legible
> than rasterised vectors

Perhaps in principle on low-res displays, but given the advent of
effective-res-enhancing subpixel-rendering-to-ordered-displays for that
to be true in practice even on current ca. 100DPI displays you'd need
something like amiga colorfonts [1] to allow you to hand-craft
individual subpixels, or at least 1-bit bitmaps drawn in 3:1 [2] aspect
ratio [3], to match the legibility of hinted vector fonts.

And while very high-res displays could perhaps mean subpixel rendering
or antialiasing is not worth bothering with (though if you have the
today-typical oodles of processing power and the code already written,
might as well), when we have e-paper displays at hundreds of DPI
(_already appearing_, will presumably only get better assuming continued
lack of apocalypse), handcrafted bitmap fonts and lack of resolution
independence are getting less and less practical or desirable.

[1] bitmap fonts not limited to 1-bit bitmaps*

[2] or maybe 2:1 for that "sublcd" green/magenta perceptual
subpixel scheme

[3] of course assuming square pixels split into ||| subpixels, also not
necessarily valid assumptions...

* Caffeine meandering: vector colorfonts? Seems like they'd be
useful/fun for obnoxious wordart/presentations and video titling. Is
there any provision for color vector data in any vector font formats?
It's obvious there _could_ be, of course.  Though I guess modern video
titling might use 3D models for glyphs for the most part though...

More information about the xorg mailing list