latin modern fonts at eleventh hour

8/14/2019 Latin Modern fonts at eleventh hour

1/37

Latin Modern fonts

at eleventh hour

Brno, November,

Bogusaw Jackowski


2/37

aaaaaaaaaaaaaaaaaa

cc ce ee ee ee ee eggg

y yy yBrno, November, Bogusaw Jackowski Latin Modern fonts at eleventh hour

Latin Modern fonts: what are they?

A olletion of fonts (urrently )in the POSTSCRIPT Type and OTF formats,

based on the Computer Modern fonts, equippedwith a rich repertoire of diacritical characters

AvailabilityCTAN, for example:ftp://ftp.dante.de/tex-archives/fonts/lm

current version: ..number of glyphs: (ca per font)

licence: openware, LPPL-basedmaintainers: B. Jackowski and J. M. Nowacki, GUST


3/37

aaaaaaaaaaaaaaaaa



Latin Modern fonts: what are they for?

The Latin Modern project, launched bylocal TEX users groups in Spring 2002,

aimed at the collection of multilingual Latin-basedfonts that could be used as a default font family

in the realm of TEX.

Moreover, being distributed in bothPOSTSCRIPT Type 1 and OTF formats,

the Latin Modern fonts can be used in any system

that supports these formats.


4/37

aaaaaaaaaaaaa



Why Latin Modern fonts?

Other possibilities:

Computer Modern fonts (D. E. Knuth, )obsolete because of the lack of diacritical characters

EC fonts (J. Knappen and N. Schwartz, )obsolete because of the lack of outline fonts

CM-Super family (V. Volovich, )

very large and difficult to maintain


5/37

aaaaaaaaaaaa



The Latin Modern distribution

POSTSCRIPT Type 1 fonts, produced byMETATYPE1 (METAPOST engine plusa bunch of AWK scripts)

OpenType fonts, prepared using Adobe Font

Development Kit for OpenType (implementedfeatures: liga, onum, cpsp, dlig, frac)

Support for TEX (.tfm files)

Support for dvips and pdfTEX (.map and .enc files)

Support for LATEX (.fd and .sty files) Substitutions (partial, of course) for CM, PL, CS, VN;

the idea: to use the original TFM files with the relevantLatin Modern PFB files and dvips requisites

METATYPE1

sources


6/37

aaaaaaaa



The Latin Modern distribution:

the collection of fonts Fonts occurring both in LMs and CMs (50)

lmb lmbx lmbx lmbx lmbx lmbx lmbx lmbx lmbxi lmbxolms lmr lmr lmr lmr lmr lmr lmr lmrlmri lmri lmri lmri lmri lmro lmro lmro lmrolmss lmss lmss lmss lmss lmsso lmsso lmsso lmsso lmssolmssbx lmssd lmssq lmssqolmts lmtt lmtt lmtt lmtt lmtti lmtto lmvtt

Fonts present in LMs and absent from CMs (19)lmbo lmso lmro lmssbo lmssdo lmssqbx lmssqbo

lmtso lmtk lmtko lmtl lmtlo lmtl lmtlolmvtk lmvtko lmvtl lmvtlo lmvtto

Fonts absent from LMs and present in CMsmath companion (!)cmdunh cmff cmfi cmfib cminch cmtex cmtex cmtex cmu


7/37

aaaaaaa



Latin Modern distribution:

dark and light typewriter fonts

lmtl10 (light): abcdefABCDEF012345

lmtt10 (normal): abcdefABCDEF012345

lmtk10 (dark): abcdefABCDEF012345 lmtlc10 (light condensed): abcdefABCDEF012345 lmvtl10 (light): abcdefABCDEF012345

lmvtt10 (normal): abcdefABCDEF012345

lmvtk10 (dark): abcdefABCDEF012345

plus all oblique variants


8/37

aaaaaa




encodingsThe LM package provides TEX font metric files for:

EC (Cork) encoding (ec-.tfm)

EL (European Letters aka dense) encoding, el-.tfm(originally introduced by the ConTEXt fraternity)

QX (GUST) encoding, qx-.tfm

Y&Ys TEXnANSI aka LY1 encoding, texnansi-.tfm

T5 (Vietnamese) encoding, t5-.tfm Text Companion for EC fonts aka TS1, ts1-.tfm


9/37

aaaaa




the collection of glyphs

Aa

& ^~*@Bb\ | { } [ ] Cc : , Dd $ E e 8 = ! F f

ff ffifflfi5 fl 4 Gg


10/37

aaaa

cc cee eee eee eee eee eggg




` > Hh - I i

J j Kk L l


11/37

aaa





S s ; 7 6 /

T t 3 2 U u _ V v W w

X x Y y Z z 0


12/37

aa



Who is to be blamed for the plentitude

of diacritical characters?Six hundred years ago people tried to avoid them:

Ztyechto kuow myey nauczenye kterak mazywbyty awey duy propyety /Prwy kuzlaty / ktoz da geden halerz

Z techto kusv mej naucenie, kterak mas zivbyti a svej dusi prospeti.Prvy kus zlaty: Ktoz da jeden haler. . .

Jan Hus (), Devet zlatych kusu


13/37

a



Who is to be blamed for the plentitude

of diacritical characters?

Mistr Jan HusOrthographia Bohemica

Alois Vojtech Sembera, Mistra Jana Husiortografie ceska. Vydal Alois Vojtech Sembera.Ve Vidni , s. (v Archvu hl. m. Prahy?)


14/37



What can be found in the Internet

about hacek? It was introduced by Jan Hus (probably)

HACEK can be acute to your heart...

The acronym HACEK refers to a groupingof gram-negative bacilli:Haemophilus species,Actinobacillus actinomycetemcomitans,Cardiobacterium hominis,

Eikenella corrodens,Kingella species.These organisms are responsible for acuteheart diseases.


15/37

a



What can be found in the Internet

about hacek? It was introduced by Jan Hus (probably)

HACEK can be acute to your heart...

Which, in a way, agrees with D. E. Knuthsanticipation:

Type design can be hazardous [...] Once you gethooked, you will develop intense feelings aboutletterforms; the medium will intrude

on the messages that you read. And you willperpetually be thinking of improvementsto the fonts that you see everywhere, especiallythose of your own design.

Isnt it a heart disease?


16/37

c cee eee eee eee eee eggg


How the Latin Modern family was created?

METATYPE1

source(and report)

PostScript

Type 1 font

Gawk

T1Utils

METATYPE1

source

METAPOST

output(tfm, eps, aux)

PostScript

Type 1 font

METAPOST

Gawk

T1Utils


17/37

cee eee eee eee eee eggg


The structure of a Latin Modern source:

a driver file (lmr10)

1. % LATIN MODERN font: a driver file for lmr102. input fontbase;3. vardef cm_pal = "cmr10" enddef;

4. input comm_mac; % common defs, CM params5. input comm_mph; % common header6. input lmr10.mpm; % metric data7. input lmr10.mph; % PS-oriented header8. beginfont9. input lmr10.mpg; % frozen glyphs

10. input comm_mpg; % common glyphs (mainly diacritics)

11. if known generating: % optimize proofing time12. input lmr10.mpl; % ligatures and kerns13. fi14. endfont15. %%%% EOF


18/37

ee eee eee eee eee eggg


The structure of a Latin Modern source:

a boring excerpt from the file comm_

mpg.mp(common glyphs)

. . .

%% \vb\- Aacute:\- \PICT{Aacute}\-acc_glyph(_A)(_Acute)(_Aacute);

%% \vb\- aacute:\- \PICT{aacute}\-acc_glyph(_a)(_acute)(_aacute);

%% \vb\- Abreve:\- \PICT{Abreve}\-acc_glyph(_A)(_Breve)(_Abreve);

%% \vb\- abreve:\- \PICT{abreve}\-acc_glyph(_a)(_breve)(_abreve);

%% \vb\- Abreveacute:\- \PICT{Abreveacute}\-acc_glyph(_A)(_breveacute)(_Abreveacute);

%% \vb\- abreveacute:\- \PICT{abreveacute}\-acc_glyph(_a)(_breveacute)(_abreveacute);. . .


19/37

e eee eee eee eee eggg


A few examples of less boring

accented characters


20/37

eee eee e

ee eee eggg



accented characters


21/37

ee eee e

ee eee eggg



accented characters


22/37

e eee e

ee eee eggg



accented characters


23/37

ee e

ee eee eggg


Compatibility issues

A subset of CMs font metric files can be safelyused with LMs PFB files; known exceptions:

lack of math (will be fixed soon)

a macro \hbar produces slightly worse result(green CM, gray LM, blue LM glyph):

hh


24/37

e e

ee eee eggg





a macro \hbar produces slightly worse result

there are some discrepancies between CS and LMType 1 (PFB) fonts, reported by Karel Pska;for example, the widths of guillemots differ,most significantly in 5-point fonts:

csr5.pfb: 5000/9=555.556; lmr5.pfb: 750but in csr10.pfb i and lmr10.pfb the respectivewidths are consistent:

csr10.pfb: 5000/9=555.556

lmr10.pfb: 5000/9=555.556


25/37

e e

ee eee eggg





a macro \hbar produces slightly worse result

there are some discrepancies between CS and LMType 1 (PFB) fonts, reported by Karel Pska:

csr5.pfb: H

lmr5.pfb: H


26/37

e e

ee eee eggg



A subset of CMs font metric files can be safelyused with LMs PFB files

In general, however, full compatibilitycannot be even dreamt of

the sets of fonts differ

the exact dimensions of glyphs cannot be retained(the infamous limit of 16 heights, 16 depthsand 64 italic corrections)

in the LM fonts some apparent flawsof the CM fonts have been corrected(both concerning metric dataand character shapes)


27/37

e e

ee eee eggg



So, why not to follow the Senior Typographer, Hermann Zapf,who quite recently decided to improve his Optima?


28/37

e e

ee eee eggg



What can we do to make the LM fontsusable for the Czech TEX community?

We have already equipped LMs inCzech double quotes

Accents (hacek) have been modified

CS encoding(s), i.e., the relevant .tfm files,can be easily added to the LM distribution

We are open to further suggestions


29/37

e

ee eee eggg


Compatibility issues:an example of strange kernings

that have been corrected

The file roman.mf reads:

% three degrees of kerning:k#:=-.5u#; kk#:=-1.5u#; kkk#:=-2u#;ligtable "k":

if serifs: "v": "a" kern -u#, fi"w": "e" kern k#, "a" kern k#,

"o" kern k#, "c" kern k#;

The kern pair k-a appears in TFMs of serif fontstwice: first with the value u#, next with

the value k#, i.e., 0.5u#; moreover, thereare no v-a, v-c, v-e, and v-o kern pairs in sansseriffonts, although there are k-c, k-a, k-e, k-o, w-a,w-c, w-e, and w-o kern pairs in these fonts.The problem of missing v-* kerns in sansserif variants

occurs in EC, CS, PL, VN, and CM-Super fonts!


30/37

e eee eggg


Compatibility issues:a few example glyphs (mostly accents)

that we decided to touch

} ^ _

solid green CM, outline black


31/37

ee eggg


Compatibility issues:an example feature

that we are not going to touch

In the LM typewriter fonts, the letters v, y, x werereported to be too thin; the comparison with the

original fonts shows, that this was D. E. Knuths design:

Font vtt10: CMs bitmap; CMs outlines; LMs outlines


32/37

ggg


Plans for the future

We are not going to equip the Latin Modernfonts with non-Latin alphabets, such ascyrillic, Greek, etc.

We are not going to increase the numberof so called optical sizescan you seeany difference?

eee


33/37

ggg




We are not going to increase the numberof so called optical sizescan you seeany difference?

eeecmr8 at 9pt cmr9 at 9pt cmr10 at 9pt


34/37

ggg




We are not going to increase the numberof so called optical sizes

In general, we are not going to go too far withthe improvement of the original D. E. Knuthsdesignwe will consider the changing of only

those features which would have hamperedthe extending of the Latin Modern familywith new diacritical characters


35/37

yBrno, November, Bogusaw Jackowski Latin Modern fonts at eleventh hour

Plans for the future: to do

Fixing bugs and misfeaturesas soon as possible

Reaching the basic stability of the metric dataof the packagebefore the release of the version 1.00(scheduled for Spring 2006, BachoTEX)

Adding a basic math companion (borrowedfrom CMs, to begin with)timing as above

The improvement of kerning (reluctantly), glyphshapes (redrawing all accents from scratch?),and hinting; improving OTFs by, e.g., the additionof new featuresnot in a hurry

Further extensions to the character set (e.g., Africanlanguages)after the release of the version 1.0

More math, small capsfar future?..


36/37

yBrno, November, Bogusaw Jackowski Latin Modern fonts at eleventh hour

Should you have remarks concerningthe current LM release, please consult

Harald Harderss www pageWishes for Latin Modern:

http://www.harald-harders.de/latex/lmodern.html

At the moment the page is unmaintained;

nevertheless, a lot of information can be found there.


37/37

The Latin Modern project is supported by TEX UsersGroups, in particular, by the German-speakingTEX Users Group DANTE e.V., the Polish TEX UsersGroup GUST, the French-speaking TEX Users Group

GUTenberg, the Dutch-speaking TEX Users Group NTGandlast but not leastTUG.

latin modern fonts at eleventh hour

Documents