Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site uw-beaver
Path: utzoo!watmath!clyde!bonnie!akgua!whuxlm!whuxl!houxm!vax135!cornell!uw-beaver!laser-lovers
From: laser-lovers@uw-beaver
Newsgroups: fa.laser-lovers
Subject: Re:  More news on Postscript printers
Message-ID: <584@uw-beaver>
Date: Fri, 25-Jan-85 20:35:21 EST
Article-I.D.: uw-beave.584
Posted: Fri Jan 25 20:35:21 1985
Date-Received: Tue, 29-Jan-85 07:30:36 EST
Sender: daemon@uw-beaver
Organization: U of Washington Computer Science
Lines: 56

From: Robert Morris <ram%vax2.uucp%umass-boston.csnet@csnet-relay.arpa>

re: making PostScript printers output CM fonts strikes me as
akin to making a Maserati (well, maybe a Porsche) masquerade
as a Volkswagen. PostScript has enough power to get back from
the PostScript engine sufficient metric information to in
turn build TeX font metrics and have TeX deal directly with
PostScript's native abilities.

People who want to circumvent Adobe font generation should
in my opinion have a good reason to do so, and CM sure
ain't such a reason (except, as above, if you want all
the cars in your garage to look like Volkswagen's). There
are reasons, some of which are presently the subject of
hot debate which only time and experiments will settle.
Adobe rasterizes on the fly and bitmaps are therefore
untuned. Some authorities claim you can not get the best
results on low-resolution devices when you rasterize on
the fly, but other's claim that the
Adobe results will satisfy all or most users. Another
problem arises when you have to match font widths for
two disparate devices, e.g. you have screen fonts which
are supposed to represent the printer fonts, but your
composition system uses the printer font metrics to 
determine line breaks in order to get the best result
for the printer. If the screen font is, say, too wide
then the screen text will be overcrowded and you will judge
that the setting is bad(unless you look at the printed output).
Similarly one might imagine such a problem if one set with
typesetter metrics and used the laser printer as proof device.
Your judgment of the quality of the final setting will be
in error if the fonts are badly mismatched (it's harder
to match screen fonts to laser than laser to typesetter
because a 12 point screen font is only about 12 pixels high
and 6-10 wide at typical screen resolutions. hence the
artist(or rasterization algorithm)
 has fewer pixels to push around to tune the fonts)

I have the impression that Adobe is prepared to offer
its customers (i.e. OEM's) screen fonts for typical
screens which solve this problem. Apple must have solved
it, or else either the Mac screens or the printed output
will look badly spaced (anyone seen a WYSIWYG application
with the Mac and Laserwriter?). I also have the
impression that Imagen is or will offer the corresponding
solution from Bigelow&Holmes for the Lucida fonts described
recently in this forum.

These being rather beautiful fonts whose outlines are
explicitly designed for laser printers, I would make the
same argument about a desire to use TeX with CM instead of
Lucida on an Imagen or whatever else you can get them on.
Your main motivation would have to be that you want your
new copies of the document to look as unsatisfactory as
your old ones. Imagen/Lucida and PostScript/{Times,Helvetica}
strike me as Porsche's in the garage.