freetype
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Using the FreeType 2.0 Polygon Rasterizer in a GraphicsPackage


From: Just van Rossum
Subject: Re: Using the FreeType 2.0 Polygon Rasterizer in a GraphicsPackage
Date: Thu, 9 Mar 2000 11:50:08 +0100

At 8:52 AM +0100 09-03-2000, David Turner wrote:
>Thanks for the tip.

Same here.

>I must say that I know a lot about LibArt, actually I'm
>pretty looking forwards to seeing Raph Levien again at Guadec in a few
>days :-)

I knew nothing about LibArt, so I am enlightened. (But I couldn't find a
download link at that page :-( )

>LibArt is an excellent graphics library, but it's probably unsuitable to use
>it just to stroke FreeType outlines.. Main reasons are:
>
>  - LGPL license, thus incompatible with the FreeType one

That is a big problem indeed.

>  - there is no easy way to use its path stroker independently from
>    the rest of LibArt, which means linking a 50 Kb lib to do just that..

Ouch...

There's actually a slight problem with using the FT raster (interface) for
graphics: for strokes you need to know whether a subpath is closed or not.
For filled shapes this isn't so important, so it's logical this info isn't
available in FT. Oh, and TT outlines are closed by definition... So: in
order to use FT as a graphics primitives renderer (when a stroke raster is
added), there needs to be one flag for each subpath. The simplest thing I
can think of righ now would be to use the highest order bit of the
FT_Outline.contours array elements. Hm...

>  - ideally, the path stroker should be used to embolden fonts too (because
>    this is basically the same process). As far as I know, this is not
>    possible with the LibArt stroker..

Aaaaah!!! The nightmare of any serious typographer/type designer:
emboldened fonts... If I had *anything* to say about this, I'd say don't
implement this. It destroys *any* typeface.

More general: "say no to embolding" ;-)

Just





reply via email to

[Prev in Thread] Current Thread [Next in Thread]