Received: by nummer-3.proteosys id <01C19443.8F82A3FC@nummer-3.proteosys>; Thu, 3 Jan 2002 11:44:03 +0100 Return-Path: <@vm.gmd.de:LATEX-L@DHDURZ1.BITNET> MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----_=_NextPart_001_01C19443.8F82A3FC" x-vm-v5-data: ([nil nil nil nil nil nil nil nil nil][nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil]) X-MimeOLE: Produced By Microsoft Exchange V6.5 x-to: LATEX-L%DHDURZ1.BITNET@uga.cc.uga.edu Content-class: urn:content-classes:message Subject: Re: fontdef issues Date: Sun, 5 Jan 1992 08:28:00 +0100 Message-ID: X-MS-Has-Attach: X-MS-TNEF-Correlator: From: "Don Hosek" Sender: "LaTeX-L Mailing list" To: "Rainer M. Schoepf" Reply-To: "LaTeX-L Mailing list" Status: R X-Status: X-Keywords: X-UID: 477 This is a multi-part message in MIME format. ------_=_NextPart_001_01C19443.8F82A3FC Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable -while i agree in general terms with what's been said about the -absurdity of calling a font with design size of 17.28pt a 17pt font, -i would remind you all that one of the more important reasons for -magsteps is to permit documents to be printed out at a larger size -than the size at which they were tex'ed. this feature has been -supported in at least some "past" dvi drivers, although i haven't -kept up with the current realities. Few drivers allow scaling of the document at DVI-processing time. Incidentally, it's worth noting that the PK file requested by asking for, say cmr10 scaled X in a document scaled Y is not necessarily going to be the same as the PK file requested by asking for cmr10 scaled X*Y/1000 in a document scaled 1000. The rationale for the magstep series is given in the TeXbook and personally I think that it's time to move on, at least in the LaTeX world. -another important use of magnified fonts is in making slides. the -proper fonts to use for this purpose are ones with a smaller design -size, optically enlarged =3D "magnified", rather than fonts designed to -the larger size. i'm not saying that the enlargements should -necessarily progress in steps of 1.2, but that is a relatively -efficient way to do it. Slides are another issue altogether. How the NFSS and SliTeX interface I don't know because a certain person whose name starts with "F" and ends with "h" and has "ank Mittelbac" in the middle has never released his version of sfonts.new (or at least not while I was looking). -i think i needn't belabor the point that when these two applications -are added to that of printing out documents with fonts at the design -size, the required bitmap font libraries can be much smaller than they -would be otherwise. I don't know that size is really an issue with 40M drives generally a minimum in the personal computing world and the fact that a well-chosen set of fonts can fit in less than 2M of disk space ("well-chosen" means every font mentioned in lfonts.new+every font chosen in plain.tex plus every plain TeX font _not_ preloaded at magsteps 0, 0.5 and 1-5). -re what exactly *is* a 10pt font, that decision is entirely up to the -font designer. as don hosek has pointed out, helvetica has a much -more generous x-height than does times, and correspondingly smaller -ascenders and descenders. this was intentional! and what don has -done by scaling helvetica so that the x-height is closer to that of -times has introduced other anomalies that a skilled typographer would -not appreciate. it is worth noting, i think, that new fonts designed -particularly for laser printer use have tended to adopt the "family" -approach of computer modern (cf. stone, for example). I don't know about the "anomalies that a skilled typographer would not appreciate" being a real issue with Adobe's Helvetica. Differences in cap-heights across families are not unusual (cf. cmtt vs. cmr) and a similar issue arises with weight, color, etc. As for the blunt issue of optical scaling, yes it sucks but there are no more anomalous results from my 10% reduction than there are from using phvr at 12pt for a section header. I stand by my strategy as it stands for Helvetica type and given equivalently proportioned lead type for Times and Helvetica, I would use 9pt Helvetica with my 10pt Times. -this whole discussion started with frank's observation that cmr17 is -actually designed for 17.28pt. i'm sure the reason is so that it will -be compatible with cmr10 fonts scaled to magstep3. Care has been taken so that x-height, cap-height and a few other dimensions are compatible across what NFSS calls families in CM at 17pt. -examination of the -parameter values in cmr10.mf and of the printed images of cmr10 scaled -\magstep3 alongside of cmr17 is quite instructive. this is the kind -of subtlety that hasn't yet been accommodated by most "modern" font -technologies. (i will readily admit my bias for "designed" fonts, as -opposed to the ubiquitous optical scaling now foisted upon the dtp -public.) Here as well. I do, however, put up with it since there just isn't a whole lot of designed type available in digital formats. Speaking of subtleties, if I may ride my hobby-horse a bit here, I'd just like to say that the results one gets by mixing fonts scaled from different point sizes to the same final size are disappointing at best. In reality, cmr17 and cmr10 at 17.28 won't share the stage frequently. However cmr17 and cmsl12 at 17.28 do and the difference in weights is noticable. Other combinations also exhibit similar problems. Many macro packages (not just LaTeX) mix different design sizes scaled to a single type size on the theory that less scaling is better and can produce monstrous output in multiple-face headings (fortunately the instances are infrequent). It's far better to have a section header uniformly magnified from 10pt than to have some magnified 10pt some magnified 12pt and some 17pt. -the bottom line is that there are still reasons to use magstepped -fonts. i'm just asking for careful consideration of the effect of -abandoning magsteps before making a decision. Magnified fonts are useful, but I really wish we'd abandon our eccentric series of sizes. -dh ------_=_NextPart_001_01C19443.8F82A3FC Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Re: fontdef issues

-while i agree in general terms with what's been said = about the
-absurdity of calling a font with design size of = 17.28pt a 17pt font,
-i would remind you all that one of the more = important reasons for
-magsteps is to permit documents to be printed out at = a larger size
-than the size at which they were tex'ed.  this = feature has been
-supported in at least some "past" dvi = drivers, although i haven't
-kept up with the current realities.

Few drivers allow scaling of the document at = DVI-processing time.
Incidentally, it's worth noting that the PK file = requested by
asking for, say cmr10 scaled X in a document scaled Y = is not
necessarily going to be the same as the PK file = requested by
asking for cmr10 scaled X*Y/1000 in a document scaled = 1000. The
rationale for the magstep series is given in the = TeXbook and
personally I think that it's time to move on, at = least in the
LaTeX world.

-another important use of magnified fonts is in making = slides.  the
-proper fonts to use for this purpose are ones with a = smaller design
-size, optically enlarged =3D "magnified", = rather than fonts designed to
-the larger size.  i'm not saying that the = enlargements should
-necessarily progress in steps of 1.2, but that is a = relatively
-efficient way to do it.

Slides are another issue altogether. How the NFSS and = SliTeX
interface I don't know because a certain person whose = name starts
with "F" and ends with "h" and = has "ank Mittelbac" in the middle
has never released his version of sfonts.new (or at = least not
while I was looking).

-i think i needn't belabor the point that when these = two applications
-are added to that of printing out documents with = fonts at the design
-size, the required bitmap font libraries can be much = smaller than they
-would be otherwise.

I don't know that size is really an issue with 40M = drives
generally a minimum in the personal computing world = and the fact
that a well-chosen set of fonts can fit in less than = 2M of disk
space ("well-chosen" means every font = mentioned in
lfonts.new+every font chosen in plain.tex plus every = plain TeX
font _not_ preloaded at magsteps 0, 0.5 and = 1-5).

-re what exactly *is* a 10pt font, that decision is = entirely up to the
-font designer.  as don hosek has pointed out, = helvetica has a much
-more generous x-height than does times, and = correspondingly smaller
-ascenders and descenders.  this was = intentional!  and what don has
-done by scaling helvetica so that the x-height is = closer to that of
-times has introduced other anomalies that a skilled = typographer would
-not appreciate.  it is worth noting, i think, = that new fonts designed
-particularly for laser printer use have tended to = adopt the "family"
-approach of computer modern (cf. stone, for = example).

I don't know about the "anomalies that a skilled = typographer
would not appreciate" being a real issue with = Adobe's Helvetica.
Differences in cap-heights across families are not = unusual (cf.
cmtt vs. cmr) and a similar issue arises with weight, = color, etc.
As for the blunt issue of optical scaling, yes it = sucks but there
are no more anomalous results from my 10% reduction = than there
are from using phvr at 12pt for a section header. I = stand by my
strategy as it stands for Helvetica type and given = equivalently
proportioned lead type for Times and Helvetica, I = would use 9pt
Helvetica with my 10pt Times.

-this whole discussion started with frank's = observation that cmr17 is
-actually designed for 17.28pt.  i'm sure the = reason is so that it will
-be compatible with cmr10 fonts scaled to = magstep3.

Care has been taken so that x-height, cap-height and a = few other
dimensions are compatible across what NFSS calls = families in CM
at 17pt.

-examination of the
-parameter values in cmr10.mf and of the printed = images of cmr10 scaled
-\magstep3 alongside of cmr17 is quite = instructive.  this is the kind
-of subtlety that hasn't yet been accommodated by = most "modern" font
-technologies.  (i will readily admit my bias = for "designed" fonts, as
-opposed to the ubiquitous optical scaling now = foisted upon the dtp
-public.)

Here as well. I do, however, put up with it since = there just
isn't a whole lot of designed type available in = digital formats.
Speaking of subtleties, if I may ride my hobby-horse = a bit here,
I'd just like to say that the results one gets by = mixing fonts
scaled from different point sizes to the same final = size are
disappointing at best. In reality, cmr17 and cmr10 at = 17.28
won't share the stage frequently. However cmr17 and = cmsl12 at
17.28 do and the difference in weights is noticable. = Other
combinations also exhibit similar problems. Many = macro packages
(not just LaTeX) mix different design sizes scaled to = a single
type size on the theory that less scaling is better = and can
produce monstrous output in multiple-face headings = (fortunately
the instances are infrequent). It's far better to = have a section
header uniformly magnified from 10pt than to have = some magnified
10pt some magnified 12pt and some 17pt.

-the bottom line is that there are still reasons to = use magstepped
-fonts.  i'm just asking for careful = consideration of the effect of
-abandoning magsteps before making a decision.

Magnified fonts are useful, but I really wish we'd = abandon our
eccentric series of sizes.

-dh

------_=_NextPart_001_01C19443.8F82A3FC--