From: Johannes Rössel (joey@muhkuhsaft.de)
Date: Wed Jan 12 2011 - 05:56:29 CST
On 2011-01-12 12:00, QSJN 4 UKR wrote:
> Font creation is not so expensive and complex process. Two points for
> each line and three-four points for each circle-quadrant in gliph's
> contour.
That may well be, a good font still needs much work to create, even 
though a single glyph might be drawn in a few minutes. There is more to 
font design than simply drawing lines and circles.
> Hinting is unnecessary for hi-resolution-devices.
Sadly, I have yet to see devices with resolution that comes close to 
paper. At least the devices I view webpages on don't really exceed 
around 140 dpi.
> In order to do a great job, we need special methods
> for separating it into parts. First of all, there must be a standard
> (PANOSE?), which determines the basic properties of the font: ex to em
> ratio, the thickness of the strokes, the radius of the roundings, the
> shape of serifs. These properties must be either the input parameters
> of the font (that is, the font does not contain any information about
> them at all, but then it does not contain anything that could be
> sold!), or to be strictly identical for some font family. Second, the
> user's applications should create a composite font from the fonts of
> some family (do not install each font separately, but take a gliph (a
> kerning pair if you wish) where it found (in Internet if you wish)).
First of all, there are reasons for fonts that don't encompass all of 
Unicode, both technical (OpenType iirc only supports 64k glyphs) and 
stylistic: Latin, Greek and Cyrillic might get away with very similar 
styles (even though there is little historic precedent it works quite 
well), but how would you apply serifs to Arabic, Hebrew, Katakana or 
Bengali? How is an "ex to em ratio" even defined in scripts that don't 
distinguish case?
Different scripts have different stylistic requirements and oftentimes 
cannot simply be lumped together in the simple and nice system of how we 
deal with Latin: Antiqua in serif and sans-serif or script, blackletter, 
etc. Those are artifacts and history of our writing system. Other 
cultures differ here.
Usually you choose typefaces for different scripts for how well they go 
together with your main typeface or even other reasons; but that's not 
really a reason to lump everything together in one font. Robert 
Bringhurst's The Elements of Typographic Style also has some thoughts on 
that problem.
Back to the point of the original post, though, I cannot really think of 
terribly many use cases of allowing users to firce their fonts on 
everybody else on a web site. If they desperately want their own browser 
to use them, then there are facilities for that already.
In any case, I guess this is far beyond the scope of Unicode.
Regards,
Johannes
This archive was generated by hypermail 2.1.5 : Wed Jan 12 2011 - 06:00:52 CST