Re: failing fonts

[Tom Kacvinsky]

The fixing of mminstance is to allow for parsing of "large" XUID
arrays.  The problem we are ssing with ftlint/ftview is a coredump
associated with the MM parser, related to the existence of some global
hinting information in /Private dictionary tucked away in the /Blend
dictionary.

I suspect the reason why the t1/z1 modules fail on an instance made by
mmintance is that mminstance does not remove *all* of the MM code.  It
simply comments them out.  I haven't checked carefully, but if those
modules don't parse out the comments, they might be fooled into
thinking the instance is an MM font (which it isn't anymore).

Tom

On Sun, 9 Jul 2000, Sivan Toledo wrote:

> 
> 
> > The MM font you list below does not work with ATM on my Macintosh.
> > I think the font is buggy.I haven't had a change to figure out
> > what the problem is.Does Ed Kohler's mminstance correctly
> > interpolate it and produce a "clean" instance?
> 
> Some clarification: mminstance used to fail on this font, I sent
> the font to Ed Kohler and he fixed mminstance. The problem with 
> mminstance was that XUID was large. I don't know if this sheds
> any light on the problems since I see that you have started to track
> down the bug. 
> 
> Sivan Toledo
>