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

No Subject



cc; s.rahtz@elsevier.co.uk
In-reply-to: <199702031115.LAA10650@lochnagarn.elsevier.co.uk> (message from Sebastian Rahtz on Mon, 3 Feb 1997 11:15:08 GMT)
Subject: Re: psnfss and lw35nfss
Reply-to: bkph@ai.mit.edu
BCC: bkph


   Yannis Haralambous writes:
    > >esperanto?! i rest my case.... cant you think of a better reason?
    > Unicode 0135 LATIN SMALL LETTER J WITH CIRCUMFLEX,
    > 01F0 LATIN SMALL LETTER J WITH CARON

   Sorry, Yannis, i meant i wanted a real-world reason for having to have
   dotless j faked in all fonts. i can think of airy-fairy academic
   reasons for myself....

No, no.  This is not airy-fairy!  What *is* wrong with it is that those
character should be in the font, not the pieces they are constructed
from.  Have you guys forgotten why you went for T1 encoding :=)!

   The bottom line, as Thierry pointed out to me, is that Bernard's
   method is not demonstrated to be portable to eg Textures, so i think
   we have to hold off on it for now

It is not just Textures, it is any system that uses fonts that do
not have dotlessj without invoking a PS interpreter.  Don't be so
myopic please.  Using PS for everything is what you do on UNIX
and it makes everything device dependent.  

On Macintosh and IBM PC/Windows we try and do things in a device
independent way.  It should work on *any* display driver not just
display PostScript, and *any* supported printer not just PS.  Keep 
in mind that Windows e.g. currently supports over 3,500 output
devices/printers --- no need to write a separate DVI-to-XXXX driver
for each of these.  The advantage of this device independence bring
with it limitations: you can't assume PostScript!

Sigh, I feel like i am beating a dead horse :=)

Looking towards the future, it is also a problem with any system that
uses UNICODE since there is no dotlessj in UNICODE (yes i know,
UNICODE is a character encoding not a glyph encoding, but since there
is no decent glyph encoding we all know what is going to be used).

But actually I could care less.  This is purely academic for me.
(1) I don't use dotlessj, (2) the fonts I most use do have a 
dotlessj, (3) DVIWindo can reencode Type 1 fonts to expose the
dotlessj (even though they are not in Mac or Windows standard
character set) and (4) if	i needed jcircumflex or jcaron
I would use Lucida Bright Latin, Lucida Sans Latin, and Lucida
Sans Typewriter families, which cover all Latin code pages and
have those glyphs already wired in.

Regards, Berthold.