Re: Character Styles (here: generic ones)

[Juergen Christoffel <unido!gmdzi!jc@seismo.css.gov>]

I greatly appreciate the (unofficial) answer from DLW to my message;
this is a way for the symbolics user community to get a feeling of where
symbolics is moving to. Of course, I'll still direct all bug reports to
customer support, thus not filling up DLW's mailbox.

> From: "Daniel L. Weinreb" <DLW@ALDERAAN.SCRC.Symbolics.COM>
>   The issue of "corresponding" screen and hardcopy fonts is something
>   we've spent a lot of time discussing and working on for a long time.

For me 'corresponding' is simply: looks about the same on screen and
paper.  This could be done with a tool like METAFONT. See discussion
below.

> From: "Robert W. Kerns" <RWK@YUKON.SCRC.Symbolics.COM>
>   Do you believe it should be a requirement that ALL character styles 
>   should be supported on ALL devices?  If so, could you give your
>   reasons. As you can imagine, this would be quite expensive,
>   including for users who define their own character styles.

The METAFONT approach:

   Simply use a generic font description instead of all those particular
fonts. From this you may generate almost all pixel sizes you'll want and
almost any variation you'll like (i.e. bold, slanted, italics and so
on). This gives you a complete family of fonts for both the screen and
various (even future) output devices. Thus we would have those fonts for
our LGP-1's and LGP-2's and even might get them for future LGP-3's when
they arrive.

This clearly provides for:
	a) compatibility with future devices (and not this LGP-1 to
	   LGP-2 incompatibility)
	b) An easy way for users to provide their own character styles
	   in a clean way. Font design really IS a job for experts and
	   doing it on the pixel level is the worst thing I know.
	   (Actually FED is more or less a rather nice toy)
	c) Storage efficiency: for each family this is one generic font
	   vs. e.g. 4 faces, 5 sizes for at least 3 different output
	   devices making up at least 60 fonts to keep in the approach
	   used now by symbolics.
	d) Once you have a generic font, you have the COMPLETE set
	   of variations. (Try making an italic version from some roman
	   font using FED.) And for serious use of fonts, you'll almost
	   need those standard faces.

METAFONT surely isn't the only solution. But it is in the public domain
and would provide a sound base instead of starting from scratch. It's
written in highly portable PASCAL and with a little work could be
integrated into the lispm environment, I expect.

   One drawback of this approach is, that one might not use the
LaserWriter's builtin fonts and had to download the bitmaps instead.
But if it's not via the serial line, this shouldn't be so slow.

   I admit that there might be problems generating (readable) small
fonts with this approach. It could be tried and should be solvable.

Conclusion: We have flavors and know their value.
            Now let's have meta-fonts and we soon will appreciate
            their value too.

JC