29-Oct-1990 16:11:40-GMT,1413;000000000001
Return-Path: <rokicki@Neon.Stanford.EDU>
Received: from Neon.Stanford.EDU by math.utah.edu (4.1/SMI-4.0-utah-csc-server)
id AA09907; Mon, 29 Oct 90 16:11:35 MST
Received: by Neon.Stanford.EDU (5.61/25-eef) id AA09289; Mon, 29 Oct 90 15:11:16 -0800
Date: Mon, 29 Oct 90 15:11:16 -0800
From: Tomas G. Rokicki <rokicki@Neon.Stanford.EDU>
Message-Id: <9010292311.AA09289@Neon.Stanford.EDU>
To: karl%aten.cs.umb.edu@RELAY.CS.NET
Cc: rjl@MONU1.CC.MONASH.EDU.AU, rokicki%neon.stanford.edu@RELAY.CS.NET,
bnb@MATH.AMS.COM, rfw@MATH.AMS.COM, mackay@JUNE.CS.WASHINGTON.EDU,
beebe@csc-sun.math.utah.edu
In-Reply-To: Karl Berry's message of Mon, 29 Oct 90 10:11:00 EST <9010291511.AA06213@aten.cs.umb.edu>
Subject: dvips and psfonts.vmap
I'm going to hold off on releasing dvips 5.4 until we get a reasonable set
of names for PostScript virtual fonts. I'm happy to use anything; Karl
is obviously on top of things in this regard.
The most important, of course, is setting *concrete* names and vf files
for the basic 35 (there are 35, right?) fonts in the LaserWriter, since
these will probably account for more than 99% of the PS font usage. And
then making everything use these names, from previewers to dvi2ps to
dvi3ps, etc. And coming up with a set of vf files for those who don't
have these fonts so the previewer will display something (ie, a vf file
that maps Times-Roman back to cmr10).
-tom
29-Oct-1990 17:55:05-GMT,4408;000000000001
Return-Path: <eln272v@monu1.cc.monash.edu.au>
Received: from monu1.cc.monash.edu.au by math.utah.edu (4.1/SMI-4.0-utah-csc-server)
id AA10593; Mon, 29 Oct 90 17:54:45 MST
Received: by monu1.cc.monash.edu.au (5.61/1.34)
id AA25032; Tue, 30 Oct 90 11:53:04 +1100
Message-Id: <9010300053.AA25032@monu1.cc.monash.edu.au>
To: Karl Berry <karl%aten.cs.umb.edu@RELAY.CS.NET>
Cc: rokicki%neon.stanford.edu@RELAY.CS.NET, bnb@MATH.AMS.COM, rfw@MATH.AMS.COM,
mackay@JUNE.CS.WASHINGTON.EDU, beebe@csc-sun.math.utah.edu
Subject: Re: dvips and psfonts.vmap
In-Reply-To: Your message of Mon, 29 Oct 90 10:11:00 -0500.
<9010291511.AA06213@aten.cs.umb.edu>
From: Russell Lang <rjl@monu1.cc.monash.edu.au>
Date: Tue, 30 Oct 90 11:52:59 +1100
Sender: eln272v@monu1.cc.monash.edu.au
> Well, handling Adobe's Expert and Alternate fonts isn't too painful: we
> can just add new variants `x' (`e' is already taken) and `a'.
This solves a few ambiguities in the set of fonts mentioned in dvips,
but not all. See my list of suggested names below.
> > The StoneInformal Italics have two variants - Informal and Italic.
By rearranging the order of the double variants, I have reduced the
ambiguious names down to 1. The other names have an 'i' in the expansion
position which doesn't mean anything, yet.
Is this set of names acceptable? Should there be some mention made
in your TUGboat article of what to do in the case of multiple variants?
This reminds me of my toddler nephew trying to push a square peg into
a round hole. It just can't be done - the same for a totally consistent
short naming scheme for postscript fonts.
Tom suggested creating another set of virtual fonts so that
previewers can substitute fonts. Who is best placed to do this?
I have done a very simple version of this - take one of the Adobe
VF's, convert it to VPL format, add one line
(MAPFONT D 0 (FONTNAME cmr10))
then convert it back to VF format.
This sort of works - character spacing is dreadful, many warning messages
are generated about character codes not present, but it does allow the
previewer to subtitute fonts.
With a shell script, you could make a new set of VF files which map
all the PostScript fonts to cmr10!
At present, the pslatex included with dvipslib doesn't use the standard
magstep sizes - it uses exact pt sizes. This makes it difficult if you
try to find cmr10 at 14pt when only cmr10 at 14.4pt is available.
--
Russell Lang Email: rjl@monu1.cc.monash.edu.au Phone: (03) 565 3460
Department of Electrical and Computer Systems Engineering
Monash University, Australia
Names that do not conform to the proposed naming scheme are marked
with a '?'.
The StoneInformal Italics have two variants - Informal and Italic.
I have put the italic variant in the Expansion letter position.
Since 'i' is not used as an expansion code, this might be acceptable.
StoneSans-Slanted has variants Sans Serif and Oblique.
This one is confusing since 'pstrso' could be StoneSans Oblique,
or StoneSans extra condensed.
pstrn StoneInformal.tfm
pstrni ? StoneInformal-Italic.tfm
pstbn StoneInformal-Bold.tfm
pstbni ? StoneInformal-BoldItalic.tfm
pstsn StoneInformal-Semibold.tfm
pstsni ? StoneInformal-SemiboldItalic.tfm
pstrnn StoneInformal-Condensed.tfm
pstrs StoneSans.tfm
pstrso ? StoneSans-Slanted.tfm
pstr StoneSerif.tfm
pstri StoneSerif-Italic.tfm
pstsr StoneSerif-Semibold.tfm
pstsi StoneSerif-SemiboldItalic.tfm
pstb StoneSerif-Bold.tfm
pstbi StoneSerif-BoldItalic.tfm
Some of the Adobe Garmond names also have two variants -
Alternate or Expert and Italic.
For these cases, the Italic variant is placed in the Expansion letter
position.
rpadr AGaramond-Regular.tfm
rpadrx AGaramondExp-Regular.tfm
rpadra AGaramondAlt-Regular.tfm
padr
padra (padr with digits replaced with old style)
rpadri AGaramond-Italic.tfm
rpadrxi ? AGaramondExp-Italic.tfm
rpadrai ? AGaramondAlt-Italic.tfm
padri
padrai ? (padri with digits replaced with old style)
rpads AGaramond-Semibold.tfm
rpadsx AGaramondExp-Semibold.tfm
pads
rpadsi AGaramond-SemiboldItalic.tfm
rpadsxi ? AGaramondExp-SemiboldItalic.tfm
padsi
rpadb AGaramond-Bold.tfm
rpadbx AGaramondExp-Bold.tfm
padb
rpadbi AGaramond-BoldItalic.tfm
rpadbxi ? AGaramondExp-BoldItalic.tfm
padbi
29-Oct-1990 17:58:45-GMT,1289;000000000001
Return-Path: <LATEX-L@DHDURZ1.BITNET>
Received: from CC.UTAH.EDU by math.utah.edu (4.1/SMI-4.0-utah-csc-server)
id AA10608; Mon, 29 Oct 90 17:58:41 MST
Received: from DHDURZ1 by CC.UTAH.EDU; Mon, 29 Oct 90 17:58 MDT
Received: by DHDURZ1 (Mailer R2.07) id 0403; Tue, 30 Oct 90 01:53:56 CET
Date: Mon, 29 Oct 90 16:36:00 PST
From: Don Hosek <DHOSEK@HMCVAX.CLAREMONT.EDU>
Subject: Re: naming convention for the Pandora fonts
Sender: LaTeX-L Mailing list <LATEX-L@DHDURZ1.BITNET>
To: "Nelson H.F. Beebe" <Beebe@SCIENCE.UTAH.EDU>
Reply-To: LaTeX-L Mailing list <LATEX-L@DHDURZ1.BITNET>
Message-Id: <112FED1DA11F803231@CC.UTAH.EDU>
X-Envelope-To: Beebe@SCIENCE.UTAH.EDU
>(2) pnr10 pnbx10 pnsl10 pntt9 pnss10 pnsx10 pnsi10
I've seen these forms on, I think, Clarkson; they're misnamed versions
of the "true" names (bx for b) with the old sail reduction to 6 letters
naming scheme (the first three chars + the last 3 chars).
Karl Berry has put together an improved version of the naming scheme
referred to by Neenie which allows most typefaces to be described in
eight characters or less. It can be a little more oblique in the naming,
but it is functional.
-dh
btw, personally I find Pandora useless for my work; no italic
corrections and a rather heavy appearance.
-dh
31-Oct-1990 16:42:09-GMT,15154;000000000001
Received: from [128.110.192.2] by math.utah.edu (4.1/SMI-4.0-utah-csc-server)
id AA02468; Wed, 31 Oct 90 16:42:07 MST
Return-Path: <@RELAY.CS.NET:karl@aten.cs.umb.edu>
Received: from RELAY.CS.NET by [128.110.192.2] with TCP; Mon 29 Oct 90 08:56:01-MST
Received: from relay2.cs.net by RELAY.CS.NET id ae11871; 29 Oct 90 10:55 EST
Received: from umb.edu by RELAY.CS.NET id af10509; 29 Oct 90 10:45 EST
Received: from aten.cs.umb.edu by cs.umb.edu (3.2/1.34)
id AA16447; Mon, 29 Oct 90 10:11:02 EST
Received: by aten.cs.umb.edu (3.2/1.34)
id AA06213; Mon, 29 Oct 90 10:11:00 EST
Date: Mon, 29 Oct 90 10:11:00 EST
From: Karl Berry <karl%aten.cs.umb.edu@RELAY.CS.NET>
Message-Id: <9010291511.AA06213@aten.cs.umb.edu>
To: rjl@MONU1.CC.MONASH.EDU.AU
Cc: rokicki%neon.stanford.edu@RELAY.CS.NET, bnb@MATH.AMS.COM, rfw@MATH.AMS.COM,
mackay@JUNE.CS.WASHINGTON.EDU, beebe@SCIENCE.UTAH.EDU
In-Reply-To: Russell Lang's message of Mon, 29 Oct 90 15:57:14 +1100 <9010290457.AA01112@monu1.cc.monash.edu.au>
Subject: dvips and psfonts.vmap
Resent-Date: Wed 31 Oct 90 16:40:46-MST
Resent-From: "Nelson H. F. Beebe" <Beebe@[128.110.192.2]>
Resent-To: beebe@math.utah.edu
Resent-Message-Id: <12634288722.23.BEEBE@SCIENCE.utah.edu>
Well, handling Adobe's Expert and Alternate fonts isn't too painful: we
can just add new variants `x' (`e' is already taken) and `a'.
But
> The StoneInformal Italics have two variants - Informal and Italic.
I can't believe I didn't think of this before. Of *course* some
typefaces can have multiple variants. Damn, damn, damn. I don't see
any good solution to this. I suppose I could randomly allocate a
variant letter for `informal italic' and `sans slanted', which will
solve the immediate problem, of generating names for the fonts
distributed with Tom's dvips. But we will rapidly run out of letters
once we start having `Lucida Typewriter Small Caps Sans Serif Informal
Oblique', for example... And yet just appending the letters will lead
to ambiguities; for example, is
pstso
Stone Sans Oblique or Stone Sans Extra Condensed?
I am depressed.
karl@cs.umb.edu
(here is the latest version of the article, for those who haven't seen
it since the initial discussion some months ago)
% Font naming proposal.
%
%% @texfile{
%% author = "Karl Berry",
%% version = "1.2",
%% date = "29 October 1990",
%% filename = "fontname.tex",
%% email = "karl@cs.umb.edu",
%% address = "135 Center Hill Rd. // Plymouth, MA 02360",
%% codetable = "ISO/ASCII",
%% docstring = "This article describes a naming scheme for fonts that
%% is supposed to be implementable on all the computers
%% on which TeX runs."
%% }
%%
%
% Use the TUGboat macros.
\input tugboat.sty
% Simple abbreviations.
%
\def\host#1{{\tt #1}}
\def\filename#1{{\tt #1}}
\def\journal#1{{\sl #1}}
\def\fontletter#1{{\tt #1}}
\def\program#1{{\tt #1}}
% I use \table to make the lists of abbreviations. #1 is the category.
% \table also starts a new section.
%
\def\table#1{%
\head
#1
\endhead
\halign\bgroup
\indent\tt ##\hfil\quad
&\vtop{%
\advance\hsize by -\parindent % The \indent in the first column.
\advance\hsize by -1em % The \quad.
% One table has two-letter abbreviations.
\advance\hsize by -2\fontdimen7\tentt
\rm\noindent ##}\hfil
\cr
}
\def\endtable{%
\egroup % End the \halign.
\smallskip % I don't know what kind of space you want, but it needs
% something here.
}%
% Within the \table's, I use \entry for each row.
% #1 is the abbreviation, #2 the meaning.
%
\def\entry#1#2{#1\cr}
% I use \shortlist to make two lists of very short items that need no
% marker.
%
\def\shortlist{\list[\unitemized\tag{\empty}][\cols=2]}
\title * Filenames for fonts *
\author * Karl Berry *
\address * 135 Center Hill Rd. \\ Plymouth, MA 02360 *
\netaddress * karl@cs.umb.edu *
\article
As more typeface families become available for use with \TeX, the need
for a consistent, rational naming scheme for the font filenames
concomitantly grows. Some (electronic) discussion has gone into the
following proposal; I felt it was appropriate now to bring it before a
wider community. In some respects, it follows and simplifies
Mittelbach's and Sch\"opf's article in \journal{TUGboat}, volume~11,
number~2 (June 1990).
Here are some facts about fonts that went into the hopper when creating
this proposal:
\list
\item \TeX\ runs on virtually all computers, under almost as many operating
systems, all with their own idea of how files should be named. Any
proposal regarding filenames, therefore, must cater to the lowest common
denominator. That seems to be eight characters in length, not counting any
extension, and with case being insignificant. Characters other
than letters and numerals are probably unusable.
\item Most typefaces are offered by several vendors. The version
offered by vendor~A is not compatible with that of vendor~B.
\item Typefaces typically come in different weights (hairline to extra
heavy), different expansions (ultra condensed to wide), and an
open-ended range of variants (italic, sans serif, typewriter,
shadow,~$\ldots$). No accepted standards exist for any of these
qualities, nor are any standards ever likely to gain acceptance.
\item The Computer Modern typeface family preserves traditional typesetting
practice in at least one important respect: different sizes of the same
font are not scaled linearly. This is in contrast to most commercial fonts
available.
\endlist
Here is how I propose to divide up the eight characters:
||
FTTWVEDD
||
where
\list
\item |F| represents the foundry that produced the font, and is omitted
if there isn't one.
\item |TT| represents the typeface name.
\item |W| represents the weight.
\item |V| represents the variant, and is omitted if both it and the
expansion are ``normal''.
\item |E| represents the expansion, and is omitted if it is ``normal''.
\item |DD| represents the design size, and is omitted if the font is
linearly scaled from a single {\tt tfm} file.
\endlist
See the section on virtual fonts (towards the end) for an exception to
the above.
The weight, variant, and expansion are probably all best taken from the
original source of the typeface, instead of trying to relate them to
some external standard.
Before giving the lists of abbreviations, let me point out two problems,
to neither of which I have a good solution. 1)~Assuming that only the
English letters are used, two letters is enough for only 676 typeface
families (even assuming we want to use all possible combinations, which
is doubtful). There are many more than 676 typeface families in the
world. 2)~Fonts with design sizes over 100$\,$pt are not common, but
neither are they unheard of.
On to the specifics of the lists. If you adopt this proposal at your
own installation, and find that you have fonts with some property I
missed, please write to me (see the end of the article for various
addresses), so I can update the lists. You can get the
most up-to-date version of these lists electronically, by anonymous ftp
>From the host \host{ftp.cs.umb.edu}. I will also send them to you by
electronic mail, if necessary.
I give the letters in lowercase, which is preferred on systems where
case is significant. The lists are in alphabetical order by the
abbreviations.
\table{Foundry}
\entry{a}{Autologic}
\entry{b}{Bitstream}
\entry{c}{Compugraphic}
\entry{g}{Free Software Foundation ({\tt g} for GNU)}
\entry{h}{Bigelow \& Holmes (with apologies to Chuck)}
\entry{i}{International Typeface Corporation}
\entry{p}{Adobe ({\tt p} for PostScript)}
\entry{r}{reserved for use with virtual fonts; see below}
\entry{s}{Sun}
\endtable
\table{Typeface families}
\entry{ad}{Adobe Garamond}
\entry{ag}{Avant Garde}
\entry{ao}{Antique Olive}
\entry{at}{American Typewriter}
\entry{bb}{Bembo}
\entry{bd}{Bodoni}
\entry{bg}{Benguiat}
\entry{bk}{Bookman}
\entry{bl}{Balloon}
\entry{bv}{Baskerville}
\entry{bw}{Broadway}
\entry{cb}{Cooper Black}
\entry{cl}{Cloister}
\entry{cr}{Courier}
\entry{cn}{Century}
\entry{cs}{Century Schoolbook}
\entry{hv}{Helvetica}
\entry{gm}{Garamond}
\entry{go}{Goudy Oldstyle}
\entry{gs}{Gill Sans}
\entry{jo}{Joanna}
\entry{lc}{Lucida}
\entry{lt}{Lutetia}
\entry{nc}{New Century Schoolbook}
\entry{op}{Optima}
\entry{pl}{Palatino}
\entry{pp}{Perpetua}
\entry{rw}{Rockwell}
\entry{st}{Stone}
\entry{sy}{Symbol}
\entry{tm}{Times}
\entry{un}{Univers}
\entry{uy}{University}
\entry{zc}{Zapf Chancery}
\entry{zd}{Zapf Dingbats}
\endtable
\table{Weight}
\entry{a}{hairline}
\entry{b}{bold}
\entry{c}{black}
\entry{d}{demi}
\entry{h}{heavy}
\entry{i}{extra light}
\entry{k}{book}
\entry{l}{light}
\entry{m}{medium}
\entry{r}{regular}
\entry{s}{semi}
\entry{t}{thin}
\entry{u}{ultra}
\entry{x}{extra bold}
\endtable
In order of lightest to heaviest (more or less):
\nobreak
\shortlist
hairline
thin
extra light
light
book
regular
medium
\colsep
demibold
semi
bold
extra bold
heavy
black
ultra
\endlist
\table{Variant}
\entry{a}{alternate}
\entry{b}{bright}
\entry{c}{small caps}
\entry{e}{engraved}
\entry{g}{grooved (as in the IBM logo)}
\entry{h}{shadow}
\entry{i}{(text) italic}
\entry{l}{outline}
\entry{n}{informal}
\entry{o}{oblique (i.e., slanted)}
\entry{r}{normal (roman or sans)}
\entry{s}{sans serif}
\entry{t}{typewriter}
\entry{u}{unslanted italic}
\entry{x}{expert}
\endtable
If the variant is {\tt r}, and the expansion is also normal, both the
variant and the expansion are omitted. When the normal version of the
typeface is sans serif (e.g., Helvetica), \fontletter{r} should be used,
not \fontletter{s}. Use \fontletter{s} only when the typeface family
has both serif and sans serif variants. The ``alternate'' variant
(\fontletter{a}) is used by some Adobe fonts that have oldstyle digits.
The ``expert'' variant (\fontletter{x}) is also used by some Adobe
fonts.
\table{Expansion}
\entry{c}{condensed (by hand)}
\entry{e}{expanded (automatic)}
\entry{n}{narrow (automatic)}
\entry{o}{extra condensed}
\entry{}{regular, normal, medium (always omitted)}
\entry{w}{wide}
\entry{x}{extended (by hand)}
\endtable
In order of narrowest to widest (more or less):
\shortlist
extra condensed
condensed
narrow
regular
\colsep
extended
expanded
wide
\endlist
Expansion of fonts is sometimes done automatically (as in PostScript
{\bf scale}), and sometimes done by humans. I chose `narrow' and
`expanded' to imply the former, and `condensed' and `extended' to imply
the latter, as I believe this reflects common usage.
\head * Virtual fonts *
In concert with releasing \TeX~3.0 and \MF~2.7, Don Knuth wrote two new
utility programs: \program{VFtoVP} and \program{VPtoVF}, which convert
to and from ``virtual'' fonts. Virtual fonts provide a general
interface between the writers of \TeX\ macros and font suppliers.
In general, therefore, it is impossible to come up with a general scheme
for naming virtual fonts, since each virtual font is an individual
creation, possibly bringing together many unrelated fonts.
Nevertheless, one common case is to use virtual fonts to map \TeX's
default accent and other character code conventions onto a
vendor-supplied font. For example, \program{dvips} (by Tom Rokicki)
does this for fonts given in the PostScript ``standard encoding''.
In this case, each font consists of a ``virtual'' tfm file, which is
what TeX uses, a ``raw'' tfm file, which corresponds to the actual
device font, and a vf file, which describes the relationship between the
two.
This adds another dimension to the space of font names, namely,
``virtualness'' (or rather, ``rawness'', since it is the virtual tfm
files that the users want to see). But we have already used up all
eight characters in the font names.
The best solution I have been able to think of is this: prepend
\fontletter{r} to the raw tfm files; the virtual tfm files should be
named with the usual foundry prefix. For example, the virtual Times
Roman tfm file is named \filename{ptmr}, as usual; the raw Times Roman
tfm file is named \filename{rptmr}. To prevent intolerable confusion, I
promise never to give a foundry the letter \fontletter{r}.
This scheme will work only as long as the virtualized fonts do not have
design sizes; if they do, another foundry letter will have to be
allocated, it seems to me.
A pox upon the houses of those who decided on fixed-length filenames!
\head * Examples *
In closing, I will give two examples. First, the fonts in the Univers
typeface family were assigned numbers by its designer, Adrien Frutiger.
(You can see the scheme on, for example, page~29 of {\sl The Art of
Typo.icon.ography}, by Martin Solomon.) Naturally, we want to give
them names.
\list[\tag{\empty}\unitemized]
45 (light): \filename{unl}
46 (light italic): \filename{unli}
47 (light condensed): \filename{unlrc}
48 (light condensed italic): \filename{unlic}
49 (light extra condensed): \filename{unlro}
53 (medium extended): \filename{unmrx}
55 (medium): \filename{unm}
56 (medium italic): \filename{unmi}
57 (medium condensed): \filename{unmrc}
58 (medium condensed italic): \filename{unmic}
59 (medium extra condensed): \filename{unmro}
63 (demibold extended): \filename{undrx}
65 (demibold): \filename{und}
66 (demibold italic): \filename{undi}
67 (demibold condensed): \filename{undrc}
68 (demibold condensed italic): \filename{undic}
73 (bold extended): \filename{unbrx}
75 (bold): \filename{unb}
76 (bold italic): \filename{unbi}
83 (extra bold extended): \filename{unxrx}
\endlist
Second, here are names for the 35 standard PostScript fonts:
\list[\tag{\empty}\unitemized]
AvantGarde-Book: \filename{pagk}
AvantGarde-BookOblique: \filename{pagko}
AvantGarde-Demi: \filename{pagd}
AvantGarde-DemiOblique: \filename{pagdo}
Bookman-Demi: \filename{pbkd}
Bookman-DemiItalic: \filename{pbkdi}
Bookman-Light: \filename{pbkl}
Bookman-LightItalic: \filename{pbkli}
Courier-Bold: \filename{pcrb}
Courier-BoldOblique: \filename{pcrbo}
Courier: \filename{pcrr}
Courier-Oblique: \filename{pcrro}
Helvetica-Bold: \filename{phvb}
Helvetica-BoldOblique: \filename{phvbo}
Helvetica-NarrowBold: \filename{phvbrn}
Helvetica-NarrowBoldOblique: \filename{phvbon}
Helvetica: \filename{phvr}
Helvetica-Oblique: \filename{phvro}
Helvetica-Narrow: \filename{phvrrn}
Helvetica-NarrowOblique: \filename{phvron}
NewCenturySchlbk-Bold: \filename{pncb}
NewCenturySchlbk-BoldItalic: \filename{pncbi}
NewCenturySchlbk-Italic: \filename{pncri}
NewCenturySchlbk-Roman: \filename{pncr}
Palatino-Bold: \filename{pplb}
Palatino-BoldItalic: \filename{pplbi}
Palatino-Italic: \filename{pplri}
Palatino-Roman: \filename{pplr}
Symbol: \filename{psyr}
Times-Bold: \filename{ptmb}
Times-BoldItalic: \filename{ptmbi}
Times-Italic: \filename{ptmri}
Times-Roman: \filename{ptmr}
ZapfChancery-MediumItalic: \filename{pzcmi}
ZapfDingbats: \filename{pzdr}
\endlist
Please contact me if you have any comments or additions.
\makesignature
\endarticle