TUGBOAT
Volume 18, Number 1 / March 1997
3 Addresses 4 Notice regarding 1997 TEX Users Group election General Delivery 5 From the President / Michel Goossens 5 Editorial comments / Barbara Beeton Update to PSTricks; Quote out of context— Colophon 6 Erratum: Amsterdam, 13 March 1996 — Knuth meets NTG members, TUGboat 17, no. 4, pp. 342–355
Dreamboat 6NTS&ε-TEX: a status report / Philip Taylor METAP OST Software & Tools 12 A GNU Emacs editing mode for METAFONT and sources / Ulrik Vieth
Philology 17 The Traditional Arabic Typecase, Unicode, TEXandMETAFONT / Yannis Haralambous 30 A Medieval Icelandic manuscript: The making of a diplomatic edition / Andrea de Leeuw van Weenen
Book Reviews 37 Writing with TEX,andTEX&LATEX: Drawing & Literate Programming, by Eitan M. Gurari / Michael D. Sofka
Tutorial / Surveys 39 Typesetting mathematics for science and technology according to ISO 31/XI / Claudio Beccari A L TEX 48 A LATEX Tour, part 3: mfnfss, psnfss and babel / David Carlisle News & Announcements 56 Calendar
Late-Breaking 57 Production notes / Mimi Burbank News 57 Future issues 58 TUG’97 Update TUG Business 59 New members of the TUG Board 61 Institutional members 62 TUG membership application
Advertisements 63 TEX consulting and production services 57 Index of advertisers TEXUsersGroup Board of Directors Memberships and Subscriptions Donald Knuth, Grand Wizard of TEX-arcana† TUGboat (ISSN 0896-3207) is published quarterly Michel Goossens, President ∗ ∗ by the TEX Users Group, P. O. Box 1239, Three Yannis Haralambous , Vice President Rivers, CA 93271-1239, U.S.A.. Mimi Jett∗, Treasurer Sebastian Rahtz∗, Secretary 1997 dues for individual members are as follows: Barbara Beeton Ordinary members: $60. Karl Berry Students: $40. Donna Burnette Membership in the T X Users Group is for the E Robin Fairbairns calendar year, and includes all issues of TUGboat George Greenwade for the year in which membership begins or is Alan Hoenig renewed. Individual membership is open only to Judy Johnson named individuals, and carries with it such rights Patricia Monohon and responsibilities as voting in the annual election. Arthur Ogawa A membership form is provided on page 62. Jon Radel TUGboat subscriptions are available to organi- Petr Sojka zations and others wishing to receive TUGboat in a J´ıˇri Zlatuˇska name other than that of an individual. Subscription Raymond Goucher, Founding Executive Director † rates: $70 a year, including air mail delivery. Hermann Zapf, Wizard of Fonts † Periodical-class postage paid at San Francisco, ∗ CA, and additional mailing offices. Postmaster: member of executive committee †honorary Send address changes to TUGboat,TEXUsers Group, P. O. Box 1239, Three Rivers, CA 93271- Addresses Telephone 1239, U.S.A.. All correspondence, +1 209 561-0112 payments, parcels, Institutional Membership etc. Fax Institutional Membership is a means of showing TEX Users Group +1 209 561-4584 continuing interest in and support for both T X E P. O. Box 1239 and the TEX Users Group. For further information, Three Rivers, Electronic Mail contact the TUG office. CA 93271-1239 USA (Internet) To obtain an address suit- General correspondence: TUGboat c Copyright 1997, TEXUsersGroup able for courier or delivery [email protected] Permission is granted to make and distribute verbatim services, call or email the Submissions to TUGboat: copies of this publication or of individual items from this publication provided the copyright notice and this permission TUG office. [email protected] notice are preserved on all copies. Permission is granted to copy and distribute modified World Wide Web versions of this publication or of individual items from http://www.tug.org/ this publication under the conditions for verbatim copying, http://www.tug.org/TUGboat/ provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute transla- tions of this publication or of individual items from this TEX is a trademark of the American Mathematical publication into another language, under the above condi- Society. tions for modified versions, except that this permission notice may be included in translations approved by the TEXUsers Group instead of in the original English. Some individual authors may wish to retain traditional copyright rights to their own articles. Such articles can be identified by the presence of a copyright notice thereon.
Printed in U.S.A. When a type design is good, it is not because each individual letter of the alphabet is perfect in form, but because there is a feeling of harmony and unbroken rhythm that runs through the whole design, each letter kin to every other and to all. Frederic W. Goudy Type Design: A homily (1961)
COMMUNICATIONS OF THE TEX USERS GROUP EDITOR BARBARA BEETON
VOLUME 18, NUMBER 1 • MARCH 1997 THREE RIVERS • CALIFORNIA • U.S.A. TUGboat TUGboat Editorial Board
During 1997, the communications of the TEXUsers Barbara Beeton, Editor Group will be published in four issues. One issue, Mimi Burbank, Production Manager still to be determined, will contain the Proceedings Victor Eijkhout, Associate Editor, Macros of the 1997 TUG Annual Meeting. Alan Hoenig, Associate Editor, Fonts TUGboat is distributed as a benefit of mem- Christina Thiele, Associate Editor, Topics in the bership to all members. Humanities Submissions to TUGboat are reviewed by vol- Production Team: unteers and checked by the Editor before publica- Barbara Beeton, Mimi Burbank (Manager), Robin tion. However, the authors are still assumed to be Fairbairns, Michel Goossens, Sebastian Rahtz, the experts. Questions regarding content or accu- Christina Thiele racy should therefore be directed to the authors, with an information copy to the Editor. See page 3 for addresses.
Submitting Items for Publication Other TUG Publications The next regular issue will be Vol. 18, No. 2. The TUG publishes the series TEXniques, in which have deadline for technical items has passed; reports appeared reference materials and user manuals for and similar items are due by May 16. Mailing macro packages and TEX-related software, as well is scheduled for June. Deadlines for other future as the Proceedings of the 1987 and 1988 Annual issues are listed in the Calendar, page 56. Meetings. Other publications on TEXnical subjects Manuscripts should be submitted to a member also appear from time to time. of the TUGboat Editorial Board. Articles of general TUG is interested in considering additional interest, those not covered by any of the editorial manuscripts for publication. These might include departments listed, and all items submitted on manuals, instructional materials, documentation, or magnetic media or as camera-ready copy should works on any other topic that might be useful to be addressed to the Editor, Barbara Beeton (see the TEX community in general. Provision can be address on p. 3). made for including macro packages or software in Contributions in electronic form are encour- computer-readable form. If you have any such items aged, via electronic mail, on diskette, or made or know of any that you would like considered for available for the Editor to retrieve by anonymous publication, send the information to the attention FTP; contributions in the form of camera copy of the Publications Committee in care of the TUG are also accepted. The TUGboat “style files”, for office. use with either plain TEXorLATEX, are available “on all good archives”. For authors who have no Trademarks network FTP access, they will be sent on request; please specify which is preferred. Write or call the Many trademarked names appear in the pages of TUGboat. If there is any question about whether TUG office, or send e-mail to [email protected]. a name is or is not a trademark, prudence dictates This is also the preferred address for submitting contributions via electronic mail. that it should be treated as if it is. The following list of trademarks which appear in this issue may Reviewers not be complete. MS/DOS is a trademark of MicroSoft Corporation
Additional reviewers are needed, to assist in check- METAFONT is a trademark of Addison-Wesley Inc. ing new articles for completeness, accuracy, and PC TEX is a registered trademark of Personal TEX, presentation. Volunteers are invited to submit their Inc. names and interests for consideration; write to TUG- PostScript is a trademark of Adobe Systems, Inc. [email protected] or to the Editor, Barbara Beeton TEXandAMS-TEX are trademarks of the American (see address on p. 3). Mathematical Society. Textures is a trademark of Blue Sky Research. TUGboat Advertising and Mailing Lists Unix is a registered trademark of X/Open Co. Ltd. For information about advertising rates, publication schedules or the purchase of TUG mailing lists, write or call the TUG office.
TUGb oat Volume No
Notice regarding T X Users Group election
E
for a temp orary p erio d e ective immediately to expire This notice was sentby email on April
when their regular term b egins to all TUG members for whom TUG has email
It is prop osed that individuals wishing to present addresses If anyone reading this notice here
themselves as candidates for President do so in the did not receive a copyby email please let TUG
following manner with the election b e held during the knowby sending a message to tug mail tug org
annual TUG business meeting which will take place including your correct preferred email address
during the TUG annual meeting in San Francisco July We intend to use email more frequently in the
August future to communicate with TUG members and
Any TUG member in good standing who will be need your help to keep the records up to date
present at the annual business meeting may submit a
The election for the T X Users Group Board of Di
E
valid nomination form and supp orting statement see
rectors has now ended Since there were fewer nominated
the announcementinTUGb oat or on TUG s Web
candidates than vacant p ositions in accordance with
site http www tug org nomination form ps no later
the Election Pro cedures the candidates who did submit
than Monday July the rst day of the meeting and
nomination pap ers are declared elected No ballots will
b e prepared to present his her program at the business
b e circulated n The op en p ositions as announced were
meeting currently scheduled for Thursday July
President and six members of the Board of Directors
All TUG members are reminded that except for
The following individuals will b e taking p ositions
certain transactions such as p ersonnel matters which
on the Board of Directors for year terms ending in
legally are privileged Board meetings are op en to mem
b ers as observers The Board meeting is scheduled to b e
Donna Burnette
held on Saturday and Sunday July at a lo cation
Mimi Jett
to be determined details will be p osted at the main
Patricia Monohon
meeting site and on the TUG Web pages when available
Arthur Ogawa
Candidates for President are encouraged to attend
Petr So jka
Nomination forms and supp orting do cumentation
should be submitted to the Elections Committee by
Their statements of candidacy are published in this issue
one of two metho ds by July sent to the ad
of TUGb oat
dress b elow rather than to the address given in the
Four Board members whose terms were due to end
TUGb oat announcement or July brought
in had agreed that they would b e willing for their
to the meeting and delivered in p erson to one of the
terms to b e extended to to accommo date a change
undersigned committee members If sending forms by
in the election schedule from annually to every two
mail or fax please con rm this to the Committee at
years see the election announcement TUGb oat
the email address b elow electronic copies of candidates
p Had there b een a ballot their names would have
statements may also be sent to this address any time
b een listed there for con rmation by the membership
b efore the meeting
However in the absence of a ballot it is prop osed that
In order for this business to b e transacted legally
in accordance with the Bylaws Article VI I Section
at least TUG members must be present to form a
the incoming President con rm the extension of their
quorum Bylaws Article I I I Section It is therefore
terms These Board members are
very imp ortant for members to attend this meeting in
Barbara Beeton
order to help de ne the future of our organization
Karl Berry
For the Elections Committee
Judy Johnson
Sebastian Rahtz Barbara Beeton
Ji r Zlatu ska
There is no candidate for TUG President This
Address for submission of nominations for TUG
p oses a problem Although the Bylaws and Election Pro
President
cedures provide for the lling of vacant Board p ositions
Barbara Beeton
by app ointment no provision exists for the absence of a
TUG Elections Committee
candidate for President There is not su cient time left
American Mathematical So ciety
b efore the annual meeting in July to solicit additional
P O Box
candidates and circulate ballots
Providence RI
In order to ensure that the necessary decisions
Fax
are arrived at in a demo cratic manner and that the
Email tug election mail tug org
incoming Board has some say in the matters which will
a ect them most directly Michel Go ossens the outgoing
President has app ointed the new members to the Board
TUGb oat Volume No
By the time you read this it will b e early May
and I ap ologise for the delay with whichyou receive
General Delivery
this issue We are doing everything p ossible to
get the next issue of TUGb oat to you in June If
From the President
everything works out as planned we hop e to b e able
to o er youa very useful surprise with that issue
Michel Go ossens
And last but not least do not forget TUG
The Elections are now b ehind us and it is
i e your Annual Meeting in July in San Francisco
a real pleasure to congratulate Donna Burnette
We lo ok forward to a huge turn out We counton
Mimi Jett Patricia Monohon Arthur Ogawa and
your presence
Petr So jka who will serveasTUG Board members
Michel Go ossens
until Summer I extend my b est wishes to
CERN Geneva Switzerland
Barbara Beeton Karl Berry Judy Johnson and Ji r
goossens cern ch
Zlatu ska whose terms were extended until Summer
Together they will form the New Board who
will formally take over from the current Board at
our Annual Meeting in the Summer However to
prepare for the changeover and to have a demo cratic
decision base which is as large as p ossible I have
Editorial Comments
app ointed the incoming members to the Board with
immediate e ect I am sure that together we shall b e
Barbara Beeton
b etter able to reach the necessary urgent decisions
Up date to PSTricks
which need to be taken to make TUG serve its
membership and the T X users b etter
E
Users of PSTricks will know that for some years
On the other hand we had no candidate for
there has b een a b eta release of the package with
TUG President and as explained in the previous
signi cant extra facilities The author Timothyvan
article we are prop osing an ad hoc pro cedure to
Zandt has had little time to work on the package
extend the nomination p erio d and elect a candidate
recently but he has now agreed to an interim release
at our Annual Meeting in July I sincerely hop e
of PSTricks PSTricks which merges in all the
that one or more go o d candidates will step forward
b eta material with the main package and corrects
to take up the O ce of TUG President TUG really
known bugs This release has b een co ordinated by
needs someb o dy who can motivate her or his collab
Denis Girou and Sebastian Rahtz
orators in an e cient enthusiastic yet professional
This release has b een installed on CTAN in
way and can act as a stimulus to guarantee that the
graphics pstricks the old release has b een moved
Board acts as a united Team to address the problems
to oldstuff pstricks
at hand
At the same time there is a new mailing list
It is with great regret that I have to announce
for users of PSTricks mo derated by Denis Girou If
that Mimi Burbank who has served on the TUG
you want to subscrib e send a message saying
Board for manyyears most recently as TUG s Trea
subscrib e
surer resigned from the Board for p ersonal reasons
to
As well she will not be assuming the function of
pstricks request mail tug org
O ce Manager as prop osed in the previous issue of
Quote out of context Colophon
TUGb oat I am pleased that Mimi will continue to
work on TUGb oat since as TUGb oat Pro duction
Prices for complete sets of Colophon have b een ad
Manager she is an extremely valuable member of
vertised at upward of and single sp ecimens
the TUGb oat Pro duction Team
of the rarest have b een listed at several thousand
To cop e with ongoing nancial business I have
dollars A rare journal Some sp ecial editions
app ointed Mimi Jett as interim Treasurer until the
No the word omitted from the previous sentence is
new Executive Committee is elected in July at
sp ecies and the Colophon it describ es is a slow
TUG At the same time Art Ogawa b ecame
moving wingless stag b eetle whose thirteen sp ecies
TUG s acting O ce Manager and Art accepted
are restricted to the mountaintops of the former
gracefully to take over a ma jor part of the O ce
Cap e Province of South Africa These rare b eetles
duties A more de nite prop osal ab out the new form
were given provincial legal protection in but
the TUG O ce will take will b e announced so on
an attempt to list them on the worldwide roster
TUGb oat Volume No
of endangered sp ecies failed and commercial insect
dealers continue to collect and sell them South
African authorities plan to p etition again for list
ing of Colophon in a category that o ers limited
international protection
Webster s New International Dictionary sec
ond edition de nes colophon as
Finishing touch Obs
An inscription placed at the end of a book or
manuscript often containing facts relative to
its pro duction as the scrib e s illuminator s or
printer s name the place and date of publica
tion etc as from title page to colophon
Print An emblem usually a devices assumed
by the publishing house placed on the title
page shelfback etc
Colophon was also an ancient city of Ionia
However until I read the book An Inordi
nate Fondness for Beetles I had never b ecome
acquainted with the six legged variety This book
is quite wonderful with stunning photographs and
excellent informative text I picked it up to lo ok at
the pictures and b ecame engrossed in the story it
had to tell Recommended even for the timid
Barbara Beeton
American Mathematical So ciety
P O Box
Providence RI USA
bnb ams org
Erratum Amsterdam March
Knuth meets NTG members
TUGb oat no pp
Barbara Beeton
As often happ ens when taking shortcuts something
may go wrong In this case it app ears on page
of TUGb oat where in the right column you
will nd the words M etaP ost instead of the math
symbol This of course is due to a rede nition of
mp without paying attention We ap ologize
1
Arthur V Evans Director of the Insect Zo o Natural
History Museum of Los Angeles County and Charles L
Bellamy Senior Curator Coleoptera Transvaal Museum
Pretoria South Africa photographyby Lisa Charles Watson
technical advisor Henry Galiano and illustrations byPatricia
Wynne A Peter N Nevraumont Bo ok Henry Holt and
Company New York ISBN
TUGb oat Volume No
Dreamb oat
N S " T X a status rep ort
E
T
Philip Taylor
The N S pro ject pro ject was created at the insti
T
gation of Joachim Lammarsch and under the aegis
of DANTE e V at a DANTE meeting in Hamburg
in The idea of the pro ject was and is to
p erp etuate all that is b est in T X while b eing free of
E
the constraints whichKnuth has placed on the evo
lution of T X itself For that reason the pro ject was
E
called the N S pro ject N S b eing short for New
T T
Typesetting System to emphasise that we are not
violating Knuth s wishes that T X remain entirely
E
his resp onsibility indeed wehave received Knuth s
blessing to pursue the ideas of N S and " T X and
E
T
he went so far as to make some suggestions which as
he put it he mighthave incorp orated himself were
it not for the fact that he had decided to freeze the
evolution of T X We are of course endeavouring
E
to incorp orate those suggestions although wehave
not yet fully succeeded
Before getting into to o much technical detail I
should like to explain why the group is called the
N S group yet the pro ject on which this rep ort
T
is centered is called " T X During the group s
E
early delib erations we to ok advice from Joachim
Schrod who prop osed that rather than attempt to
mo dify T X in Web we rst re implementT X us
E E
ing a mo dern rapid prototyping language such as
LISP CLOS or PROLOG Joachim s philosophywas
that T X in Web is essentially to o p o orly structured
E
to allow for radical change and a more highly struc
tured implementation with a well speci ed interface
between mo dules was essential if wewere to make
more than cosmetic changes to T X the system
E
The group agreed that this philosophy was
sound but realised that the e ort needed to re
implementT X from scratchwas greater than could
E
be achieved with voluntary lab our if success was
to be achieved within a sensible timescale it was
essential that the group be in a p osition to em
ploy a programmer or small team of programmers
who could undertake the re implementation As
the group had no budget of its own and as it was
nancially dep endent on the go o dwill of DANTE
it was reluctantly agreed that this re implemention
would have to be put on ice until such time as
the group had adequate nancial resources in the
meantime the group agreed to pursue a rather more
TUGb oat Volume No
conservative approach evolutionary rather than rev What are the features of " T X First and
E
olutionary and to put all of their e orts into this less foremost " T X is compatible with T X if
E E
radical pro ject you want to try " T X you can use it to pro cess all
E
I am delighted to b e able to rep ort that whilst of your legacy do cuments it will pro duce results
this pap er was b eing written the future of the N S identical to T X rightdown to compatibility at the
E
T
pro ject b ecame very much more certain during level of the TRIP test When you are con dent that
the February meeting of DANTE in M unchen the " T X can do everything that T X do es you can try
E E
Board and members of DANTE agreed to donate its extensions there are approximately of these
DM to the N S pro ject which we b elieve eachintended to make the task of programming in
T
should enable us to employa programmer for one T X somewhat simpler even when these extensions
E
year full time in the Czech Republic working are enabled that is when " T X is op erating in so
E
on the N S pro ject under the direct sup ervision called extended mo de " T X will still pro cess
E
T
of Professor Ji r Zlatu ska Dean of the Faculty of all existing T X do cuments in a manner identical
E
Informatics at Masaryk University in Brno to T X provided of course that the do cumentdoes
E
Despite this very signi cantchange in the sta not inadvertently refer to one of the new " T X
E
tus of the N S pro ject the group have not aban primitives And nally ifyou need to you can use
T
doned or even reduced their e orts concerning " " T X in enhanced mo de to get access to new
E
T X and we have agreed a tentative sp eci cation features which are simply to o radical to be
E
for " T X V which we will be discussing and en compatible with T X in the rst release there
E E
deavouring to implement during the coming months is only one enhancement T X X T based on
E E
It is our intention to release " T X V one year after the earlier bi directional typesetting system called
E
the release of " T X V that is to say during the T X X T by Don Knuth and Pierre MacKay
E E E
rst twoweeks of November Unlike T X X T T X X T requires no variant of
E E E E
To clarify the distinction N S refers to a DVI no new pseudo device driver and p erforms all
T
future pro ject to completely re implementT Xina of its op erations internally
E
mo dular fashion using a mo dern rapid prototyping To summarise " T X has three mo des of op
E
language and to investigate each mo dule in turn eration These are compatibilty mo de in which
to see how it might be improved examples of " T X b ehaves identically to T X including full
E E
p ossible mo dules include the user interface the user TRIP compatibility extended mo de in which "
programming language the typesetting engine itself T X provides approximately new primitives but
E
and or comp onents of the typesetting engine suchas within whichitcontinues to remain completely com
the line and page breaking algorithms patible with T X although it can no longer pass the
E
" T X on the other hand refers to work TRIP test b ecause of the presence of new primitives
E
in progress which seeks to develop the T X in Web and enhanced mo de within which strict com
E
implementation in a way which is and will remain patibility is sacri ced to allow additional features
compatible with T X itself The " of " which are in some way fundamentally incompatible
E
T X can be thought of as representing evolu with T X The choice between compatibility or
E E
tion extension enhancement and p erhaps extended mo de is made at the time a format is b eing
even Europ ean the varepsilon of its typeset form generated once the format is dump ed it contains
emphasises that it is just a small evolutionary step within it a ag which indicates in whichmodeitis
from T X itself not a fundamental paradigm shift to op erate and it cannot b e used in the alternative
E
So work on N S will be started in the near mo de The choice b etween extended and enhanced
T
future but " T Xisnow After approximately three mo des on the other hand is made at run time
E
years in development and testing we released " T X even if an enhancement is enabled during format
E
at the Hamburg meeting of DANTE towards the end creation that enhancement will be automatically
of A few cosmetic changes were made just disabled b efore the format is dump ed and thus
subsequent to its o cial release but despite the when the format is used the enhancement will ini
very large number of accesses made on the " T X tially b e disabled Enhanced mo de can be entered
E
reference Web site I am delighted to say that we only from extended mo de not from compatibility
have received no rep orts of bugs This might of mo de
course simply mean that no one is actually using
The extensions of " T XV
E
" T X but I hop e it means that " T X is as far as
E E
p ossible bug free The extensions of the rst version of " T X can b e
E
classi ed into six distinct groups plus a seventh
TUGb oat Volume No
which is accessible only in enhanced mo de these expansion It then applies T X s input mech
E
groups are anism to this sequence of characters re tokenising
them according to the current catco de environment
Generalisation of the mark concept
currentgrouplevel is an internal read
Additional control over expansion
only integer which returns the current group level
Provision for re scanning previously read text
at the p oint of call in other words it returns the
Environmental enquiries
current depth of group nesting
Additional debugging features
currentgrouptype isaninternal read only
Miscellaneous primitives
integer which returns the type of the innermost
and
group as an integer in the range to These
Bi directional typesetting the T X X T prim
numbers can be converted to text strings using
E E
itives
de nitions provided in the " T X macro library
E
ifcsname has the same syntax as T X s
E
A brief sp eci cation of the elements which com
csname but is a Bo olean if yielding true
p ose these seven groups is given b elow followed bya
if and only if the putative control sequence is already
provisional sp eci cation of our plans for " T X V
E
known to " T X
E
In " T X V wetook the initial step towards
E
ifdefined is analogous to ifcsname but
eliminating from " T X the xed limits which be
E
takes as parameter a control sequence or active
set the current T X language we generalised the
E
character it yields true if and only if the control
concept of a mark into an array marks with
sequence or activecharacter is already known to "
elements in the rst release All of the related
T X
E
mark variables were similarly generalised into ar
lastnodetype isaninternal read only inte
rays of elements
ger which returns the type of the last no de on the
We added a further twenty three new primi
current list as an integer in the range to the
tives
upp er b ound may b e increased in a future version
protected is a pre x which can be used
These numbers can b e converted to text strings us
during macro de nition A macro de ned as pro
ing de nitions provided in the " T X macro library
E
tected will not expand during an edef a
eTeXversion isan internal read only inte
write orany similar op eration in which expan
ger which contains the integral comp onent of the
sion normally o ccurs It will however expand if it
combined version revisionnumber
reaches T X s stomach for example during the
E
eTeXrevision is a primitive which expands
actual typesetting pro cess
to yield a sequence of character tokens of category
detokenize isintended to b e used just b e
co de representing the fractional comp onent of the
fore a brace delimited token list a general text
combined version revisionnumber
in the terminology of The T Xbook It expands
E
showtokens is intended to be followed by
to yield a sequence of character tokens of category
a brace delimited token list and provides a simple
co de or corresp onding to the characters which
way of displaying the contents of a particular ele
comp ose the tokens of the unexpanded token list
ment of the marks family of arrays it has many
unexpanded is also intended to b e used just
other p otential applications
b efore a brace delimited token list but it expands
interactionmode provides read write ac
to yield the actual tokens of the token list If T X
E
cess to the current interaction mo de Assigning a
is p erforming an edef ora write or similar
numeric value sets the asso ciated mo de while the
op eration no further expansion takes place but if
current mo de may be ascertained by interrogating
these tokens reach T X s stomach for example
E
its value Symbolic de nitions of these values are
during the typesetting pro cess then they will ex
provided in the " T X macro library
E
pand normally
showgroups causes " T X to display the
E
readline is analogous to read but
level and type of all active groups from the p oint
treats eachcharacter as if it were currently of cate
at whichitwas called
gory co de or the text read can then b e scanned
tracingassigns if set to a p ositive non
and re scanned in di erent catco de environments
zero value causes " T X to display the value of reg
E
using another new primitive scantokens
isters b oth b efore and after assignment Standard
scantokens is intended to be followed by
T X displays only the new value not the old
E
a brace delimited token list and decomp oses the
token list into the corresp onding sequence of charac
ters as if the token list were written to a le without
TUGb oat Volume No
tracinggroups is an aid to debugging run beginR indicates the start of a region which
away group problems If it is set p ositive and non should b e set right to left
zero " T X traces entry and exit to every group endR indicates the end of a region which
E
tracingifs is an aid to debugging the ex should b e set right to left
pansion of conditionals If it is set to a p ositive non predisplaydirection isaninternal read
zero value " T X traces the owofcontrol through write integer which is initialised by " T X to indicate
E E
conditional statements the direction of the last partial paragraph b efore a
tracingscantokens when set to a p ositive maths display it is used to control the placementof
non zero value causes " T X to display an op en elements such as equation numbers and it can be
E
parenthesis and space whenever scantokens is directly set to alter this placement when appropri
invoked the matching close parenthesis will b e dis ate
played when the scan is complete
The " T X macro library
E
tracingcommands is de ned in T X to pro
E
vide di erent degrees of verbosity as its value is Although the ma jority of the e ort in developing
increased from zero to two in " T X we provide " T X has b een put into adding new functionality
E E
greater verbosity and detail when it is set to values to the T X language we also sp ent a little time in
E
greater than two developing a small macro library to accompany "
everyeof is one of Knuth s p ossibly go o d T X In essence this is a wrapp er for the Plain
E
ideas listed at the end of tex bug It is anal format source augmenting the existing de nitions
ogous to the other every primitives and where appropriate to supp ort the new primitives
takes as parameter a brace delimited token list the We also to ok the opp ortunityto add two fea
tokens of which are inserted when the end of a le tures whichwe thought might b e appreciated by the
is reached T X community at large we delayed the reading
E
middle is analogous to T X s left and of patterns and exceptions until a natural language
E
right primitives it sp eci es that the following handling mechanism had b een de ned so now pat
delimiter is to serve b oth as a right and left delim terns and exceptions are asso ciated with a particular
iter It will be set with spacing appropriate to a language rather than b eing de ned in limbo and
right delimiter with resp ect to the preceding atom we added supp ort for " T X library les so that one
E
and with spacing appropriate to a left delimiter with can now load mo dules as an alternativeto loading
resp ect to the succeeding atom complete les
unless allows the sense of all Bo olean con The language handling mechanism is not pred
ditionals to be inverted For example unless icated on the use of any particular natural language
ifeof yields true if and only if end of le has supp ort system it can be linked to Bab el for ex
not yet b een reached ample or to any other natural language handling
The nal class of primitives is comp osed of system In addition ho oks are provided so that
those that are accessible only when " T X is op er user co de can b e threaded b efore and after language
E
ating in enhanced mo de An " T X program enters selection and in this way we hop e to provide a
E
enhanced mo de when it assigns a p ositive non zero su ciently exible language handling environment
value to one of the enhanced state variables of to supp ort the needs of the ma jority of national
which the only one in the rst release of " T X T X user groups This is not to say that wedonot
E E
is TeXXeTstate Once this has b een assigned foresee a r ole for Omega on the contrary it is clear
a p ositive non zero value ve other primitives b e from the Omega discussion list that there is a very
come accessible beginL beginR endL real need for a system of Omega s complexity How
endR and predisplaydirection If these ever we b elieve that for typesetting environments
primitives are used when " T X is not op erating in in which access is needed only to the ma jor Western
E
enhanced mo de an error is rep orted languages " T X will prove su cient
E
TeXXeTstate is an internal read write in
Availability
teger which when set to a p ositive non zero value
enables use of the T X X T primitives When " T Xwas announced in late two refer
E E E
beginL indicates the start of a region which ence implementations were available Peter Breit
should b e set left to right enlohner s PubliC " T X which is a Turbo Pas
E
endL indicates the end of a region which cal implementation for the IBM PC family run
should b e set left to right ning MS DOS and Christian Spieler s VMS " T X
E
a Pascal implementation for the VAX VMS and
TUGb oat Volume No
AXP VMS family of machines Since " T X s re page break is taken These will b e made available
E
lease it was rst p orted to the Commo dore Amiga until the next page break in a new reserved box
and then to the Windows NT environments called pagediscards We do not consider it
During late February Bernd Raichle an p ossible at the present time to provide similar access
nounced that his p ort of " T XtoWeb C Version to material discarded during line breaking
E
was also available for release Bernd who works at In a similar area we hop e to make vsplit
Stuttgart University tells me that Eb erhard Mattes more useful by enabling the programmer to re
is planning to release a combined " T X ML TeX move the topsplitglue which is currently
E
for the IBM PC family in the near future inserted by T X we prop ose to implement
E
That summarises the present state of " T X in this in a very general manner by provid
E
the remainder of this pap er I will concentrate on the ing four list destructors one of which is al
ideas whichwe are considering for " T X V ready present in " T X V The four destruc
E E
tors are firstnodetype lastnodetype
Ideas for " T XV
E
removefirstnode and removelastnode In
We are lo oking at facilities for testing that a partic a future implementation we may allow access to
ular character exists within a given font these no des as well as the option of simply removing
iffontchar takes two parameters a font them
and a character number and yields true if and In T X parshape is really a write only
E
only if the character exists within the font It is quantity only the number of entries in the current
easy to de ne a macro ifchar which tests for parshape can b e subsequently interrogated In
the existence of a character within the current font " T X V weintend to provide read access to all of
E
There are four related primitives for nding the the dimensions of parshape although we are not
dimensions of a particular character in a given font yet certain whether we will do this through a single
fontcharwd fontchardp fontcharht control sequence parshapedimen or through a
and fontcharic each take two parameters a pair of control sequences parshapewidth and
fontandacharacter number and return the width parshapeindent No matter which is imple
depth height and italic correction resp ectively The mented the control sequence will b e indexed byan
e ect cannot b e achieved by setting a character in integer to return a particular dimension from the
a box and then measuring the box b ecause one or current parshape
more of the dimensions maybe negative As with In a manner analogous to " T X V s cur
E
iffontchar it is simple to de ne macros which rentgrouplevel and currentgrouptype we
p erforms the analogous tasks for the current font intended to provide in V new control sequences
We are debating whether to incorp orate a con for interrogating the current ow of control through
trol sequence iffont which would take as pa conditionals currentiflevel and current
rameter the external name of a font and return iftype We are also considering a third
true if and only if a font metric le of the same currentifbranch which will enable the pro
name can b e op ened not all members of the group grammer to determine whether the thread of ex
are convinced of the need for this so we would pansion is currently the then or the else
welcome your comments on this idea branch and also to determine if the current if
In " T X V we provided additional diagnos has not yet received su cient tokens to decide which
E
tics which rep ort the line number at which a group branch to take It may also b e p ossible to use this
was op ened if it has not b een closed at end of control sequence to determine which or has b een
program in V we prop ose to rep ort in addition taken in an ifcase but we are not yet sure that
the name of the le although implementation dif this is p ossible
ferences may prevent us from returning the path to Again by analogy with " T X V s show
E
the le We are also considering rep orting if a group groups weintend to provide showifs in V
mis matchis detected at end of le although this We are considering but have not yet agreed
may b e limited to a fairly simple test of group level up on a new class of alignment malign this if
rather than a full check to ensure that the le is implemented would provide a maths alignment
left at the identical group to that at which it was primitive
entered Continuing on the theme of removing xed lim
We hop e to avoid many of the problems which its from " T X we are lo oking into the p ossibility
E
currently b eset writers of output routines by provid of removing xed b ounds on the number of count
ing access to the items which are discarded when a
TUGb oat Volume No
registers dimen registers skip registers muskip reg to an otherwise unquali ed le sp eci cation on an
input openin or openout command We isters token list registers and b oxes If we are suc
envisage allowing the default extension to b e sp eci cessful in implementing these then we will probably
ed explicitly Here again your comments would b e remove the xed limit on the number of marks at the
welcome do you b elieve that this would b e useful same time
A recurring prop osal but one which we have During discussions in Brno Don asked us to
still not entirely agreed up on is the idea of an consider providing control over the lo oseness of the
evaluate primitive which would allow arith last line in a paragraph although T Xtyp esets this
E
metic to b e carried out in " T X s mouth Although to its natural length if parfillskip is in nite
E
we are all agreed that evaluate would b e very as it usually is Don said that traditionally this
b ene cial we also realise that to implement it in line was set to the same lo oseness as the previous
a way that will guarantee rep eatability across all line We are lo oking into ways of providing not only
platforms will require that it b e implemented with these two b oundary conditions but at a continuum
out recourse to the host s oating p oint arithmetic between those extremes p ossibly through a control
sequence finaladjdemerits This in turn implies either a very limited implemen
We are trying to save stack space in " T X tation or requires considerable time to implementa
E
V by eliminating redundant assignments that is full oating p oint package in software We are still
discussing which of these to adopt if either for " values which will be restored to exactly the same
T X V value on exit from a lo op No new primitive
E
During our discussions in Brno Don asked us will be required for this it is simply an internal
to investigate the idea of providing greater control optimisation
over the spacing of fractions so as to p ermit for We feel that the lost chars message which
example less dep endency on the use of kludges such currently go es into the log le is su ciently im
as sub strut We have not yet developed a p ortant that it should app ear on the console as
suitable mo del for this but we are continuing to well accordingly we are extending the semantics of
investigate the p ossibilities tracinglostchars sothatavalue greater than
Finally two fairly ma jor prop osals ML T X will cause the message to app ear on the console as
E
and p dfT X Mike Ferguson the author creator well as in the log
E
We are lo oking into an idea which would en of ML T X has given Bernd Raichle free rein to
E
able a programmer to add material for example oversee ML T X s future Bernd and Peter Breit
E
enlohner have made us aware that the present im crop marks to a box b eing shipp ed out even if
plementation has some de ciencies particularly in the output routine has b een rendered inaccessible
A
terms of the timing of certain op erations We hop e for example by a complex format suchasLT X
E
to b e able to provide a b etter implementation but We will probably implement this through a control
will ensure that users of ML T X are able to con sequence outpage analogous to output the
E
tribute to discussions on its sp eci cation b efore any routine asso ciated with which will b e entered at the
decisions are made p oint at whichabox is to b e shipp ed out This will
As to p dfT X this is a fairly recent pro ject also allow the programmer to overlay the b ox with
E
undertaken by Han The Thanh at Masaryk Univer material to b e placed at xed p oints on the page
sity in Brno We are very impressed with Thanh s We are still discussing the implementation of
work but are not convinced that the mo del whichhe outpage and would welcome your advice as
to whether you feel it should be recursive that has adopted is necessarily the b est way to pro ceed
is should a shipout called from within an in particular we are concerned that changes of the
outpage automatically invokea further instan magnitude required to supp ort his present imple
tiation of outpage unless the outer outpage mentation could introduce subtle bugs into T X
E
routine has already cleared the outpage token whichmay b e hard to detect and whichwould ad
versely a ect its stability Accordingly we are inter list register
ested in investigating a simpler mo del whichwould To allow for formats very di erent to the Plain
A
A
defer much of the pro cessing to a p ost pro cessor L T X A S T X L S T X family for example
M
E E E
M
similar to Sergey Lesenko s dvi p df but as Thanh ATML we are considering providing an alterna
has already p ointed out this would b e incompatible tive to the current default of app ending tex
1
A T X Markup Language presented at EuroT X
E E Papendal
TUGb oat Volume No
2
with adopting a variant of the HZ algorithm which
he b elieves could be incorp orated into his present
implementation We are still considering the options
in this area but would very much like to supp ort the
concept of p dfT X in some form
E
Finally wewould like to express our thanks to
all who have made this pro ject p ossible
To Professor Don Knuth without whom many
of us would never have met for his foresight
in creating T X and for his willingness to
E
discuss ideas for " T X despite his incredibly
E
busy schedule And to b oth him and his wife
Jill for making me feel so welcome to join them
during their week in the Czech Republic last
year
To Joachim Lammarsch for instigating the
pro ject
To DANTE e V for underwriting it
To Rainer Sch opf and Joachim Schrod for their
invaluable contributions during the rst year of
the pro ject
To Peter Breitenlohner without whom " T X
E
simply would not exist Peter has written vir
tually all of the Web co de and has b een resp on
sible for many of the ideas nowin" T X
E
To Bernd Raichle who has also b een resp on
sible for many of the ideas in " T X and who
E
volunteered to write the e TRIP test whichis
a daunting task
ToJi r Zlatu ska who as Pro ject Leader for the
N S pro ject has b een very patient in waiting
T
for the pro ject to get the nancial backing
it needs and who has in the meantime con
tributed much to the developmentof " T X
E
ToFriedhelm Sowa who is continuing to inves
tigate colour and user interfaces and who acts
as treasurer for the pro ject
To all the members of DANTE for making me
so welcome each time I attend their meetings
for their courtesy in sp eaking to me in English
and for their encouragement whenever I try to
sp eak a few faltering words of German And for
their magni cent donation of DM to the
N S pro ject which will enable the pro ject to
T
nally b ecome a reality
Philip Taylor
March
2
Electronic Publishing Vol September
Ab out micro typography and the hz program Her
man Zapf
TUGb oat Volume No
Software Tools
A GNU Emacs editing mo de for METAFONT
and METAP OST sources
Ulrik Vieth
Abstract
This article announces the release of meta mode el
a GNU Emacs editing mo de for METAFONT and
METAP OST source les meta mode el provides a
number of features commonly found in GNU Emacs
editing mo des for programming languages suchas
automatic indenting of source co de syntactic high
lighting symbol completion for partially typed key
words motion commands to move to the b eginning
or end of the enclosing environment or commands
to comment out or reindentenvironments regions
or bu ers An interface to running METAFONT or
METAP OST in a shell bu er from within Emacs is
currently under development and maybeintegrated
into meta mode el later
Introduction
The GNU Emacs editor is one of the most
widely used editors on Unix systems and some other
platforms As one of its most remarkable features it
provides a vast number of sp ecialized editing mo des
for a large variety of text formatting or programming
A
languages While supp ort for editing T XorLT X
E E
les has b een included in Emacs for many years
either as part of the standard GNU Emacs distribu
tion or through the optional AUC T X package
E
no such mo de sp eci c editing supp ort existed so far
for the somewhat esoteric programming languages
of METAFONT and METAP OST
When I started using METAP OST on a regular
basis in early shortly after I had completed
my rst p ort of METAP OST integrating it into the
Web C Kpathsea distribution I wasn t to o much
concerned ab out the lack of editing supp ort since I
was primarily interested in getting acquainted with
METAP OST so that I could get some data plotted
During the course of time however esp ecially after
I started gaining a little exp erience with Emacs Lisp
programming I b ecame increasingly unhappy
ab out b eing stuck with fundamental mo de for edit
ing METAP OST sources in Emacs
1
While this article only refers to GNU Emacs most of it
should b e applicable to the XEmacs editor as well Although
meta mode el was developed exclusively under GNU Emacs
it was tried to make sure that everything will also work under XEmacs
TUGb oat Volume No
So it happ ened one day in January that I turned out to be necessary to have two separate
initialization functions to be able to take care of b egan asking myself and also our lo cal Emacs guru
subtle di erences b etween the two mo des most no what it would take to write a new ma jor mo de for
tably p erhaps when it comes to the list of known editing METAFONT or METAP OST sources After
symbols for the completion function or the list of consulting the GNU Emacs Lisp Manual the task
shell commands to generate pro of sheets turned out to b e simpler than exp ected and pretty
Following the usual Emacs conventions b oth of much straightforward On the following weekend
these initialization functions provide ho ok variables when I had some spare time I sat down to b egin
metafont mode hook and metapost mode hook to writing what was to b ecome meta mode el By
allow adding extra setup or customization co de to coincidence it happ ened to be February
the individual mo des if desired In addition there is exactly twentyyears after the day on which genesis
also a meta common mode hook that applies to b oth of T X to ok place according to Don Knuth s own
E
2
mo des as well as a meta mode load hook that is account I supp ose I couldn t havechosen a b etter
evaluated when meta mode el is rst loaded date to embark on this pro ject nor a b etter wayto
celebrate this very sp ecial anniversary
Features
Overview of meta mode el
Once the general framework for a ma jor mo de is
in place adding more features and mo de sp eci c Installation
functions b ecomes relatively easy since they can b e
From the technical p oint of view meta mode el is a
conveniently added one by one as needed The func
contributed Emacs Lisp package that rst needs to
tionality currently implemented in meta mode el
installed in a place where it can b e found by Emacs
can b e summarized in the following areas
i e either in a p ersonal or system wide Emacs Lisp
automatic indenting of source co de
library directory listed in the load path variable
To activate the features provided in meta mode el
syntactic highlighting a k a fonti cation
the package then needs to b e loaded which is most
completion for partially typed keywords
easily arranged for by adding a few lines of Lisp co de
other miscellaneous editing functions
like these
Additional functionality for running METAFONT or
autoload metafont mode meta mode
METAP OST and related commands for pro ducing
Major mode for editing Metafont sources t
pro of sheets in a shell bu er from within Emacs is
autoload metapost mode meta mode
currently under development and maybe included
Major mode for editing MetaPost sources t
into meta mode el later At present a preliminary
setq auto mode alist
test version is implemented in a separate Emacs
append mf metafont mode
Lisp package tentatively called meta buf el which
mp metapost mode
auto mode alist
may be integrated with meta mode el by making
clever use of the various ho ok variables discussed
to the p ersonal or system wide Emacs startup le
ab ove For example the load ho ok may b e used to
to have meta mode el autoloaded at the rst time
load meta buf el at the same time meta mode el
a METAFONT or METAP OST source le is op ened
is loaded while the common mo de ho ok may be
Initialization used to make the functions provided in meta buf el
available in the keymap
Once meta mode el is loaded the ab ove co de has
the e ect of invoking an Emacs Lisp function called
Indentation
metafont mode or metapost mode whenevera mf
The default keymap used in meta mode el maps the
or mp le is loaded which then pro ceeds to set up
TAB key to a function that reindents the current line
everything necessary when entering the new editing
using an appropriate indent level computed auto
mo de Much of this initialization co de is actu
matically Furthermore the RET key also reindents
ally identical for b oth METAFONT and METAP OST
the current line b efore jumping to the appropriate
mo de as far as it concerns routine tasks needed for
indent level on the next line This allows you
every Emacs editing mo de such as setting up a
3
to blindly type arbitrary META co de terminating
syntax table or installing a keymap and a pull down
each line with RET as you type and get a nicely
menu for the mo de sp eci c functions However it
3
We will henceforth use the term META whenever we
2
Donald E Knuth The Errors of T X reprinted as discuss features that are applicable to b oth METAFONT and
E
Chapter of Literate Programming p METAP OST
TUGb oat Volume No
indented source le from which the grouping level of their syntactic prop erties such as comment lines
and the control ow of conditionals and lo ops is or quoted strings
immediately apparent Since FontLock mo de is an optional package it
At present meta mode el recognizes all stan needs to b e loaded and activated rst With recent
dard META language constructs including if ::: versions of GNU Emacs this has b ecome very easy
for :::endfor and def :::enddef blo cks as well as as it is p ossible to turn on FontLockmodeaswell as
common variants like forever forsu xes vardef optional FontLock supp ort packages globally with
or even mo de def In addition it also recognizes just two lines of Lisp co de
standard macros introducing blo ck structures such
global font lock mode t
as b eginchar :::endchar in METAFONT as well as
setq font lock support mode lazy lock mode
b egin g :::end g and b egingraph :::endgraph
in METAP OST
Once FontLock mo de is globally activated like
Furthermore o ccurrences of b egingroup and
this it will automatically apply to any new editing
endgroup are also considered although this might
mo de that supp orts it In order to take advantage
actually b e the wrong thing to do if these are used
of fonti cation when writing a ma jor mo de suchas
unbalanced across di erent macro de nitions Users
meta mode el it su ces to set up a few syntactic
should therefore b e aware that it may o ccasionally
variables and put together a list of regular expres
b e necessary to adjust the indentation of their source
sions that match the various language elements we
les manually in some unusual cases
wish to have highlighted
Much of the Emacs Lisp co de used in the in
While putting together a regular expression to
dentation function in meta mode el was adopted
match a list of keywords is fairly easy writing go o d
from AUC T X s latex el which actually had a
E
patterns to match macro de nition headers presents
somewhat simpler job since it only had to lo ok out
quite a challenge since wehave to cop e with the rich
for begin ::: end environments or lonely items
variety of language constructs that are available in
while we have to handle a wider variety of META
the META languages For instance we have to be
language constructs Nevertheless most of the AUC
aware that there are not only straightforward unary
T X co de could be put to a very good use For
E
macro de nitions introduced by def or vardef in
example the co de that previously used to outdent
which the name of the function follows immediately
items could be adapted to handle o ccurrences of
after the de nition keyword but also binary op er
elseif and else within if ::: blo cks It would
ator macro de nitions introduced by primarydef
have b een p ossible to apply the same logic to exitif
secondarydef ortertiarydef in which the name
and exitunless in the middle of forever :::endfor
of the function is embedded in b etween the param
blo cks but this idea was rejected since it app eared
eter arguments Furthermore function or variable
to o di erent from common co ding style
names don t necessarily have to consist of alphab etic
In any event meta mode el allows easy cus
characters or underscores they might just as well
tomization of the kinds of META language con
consist of non word symbols such as or or
structs recognized by mo difying the default regular
op erator symbols suchas or
expressions either by using M x edit options or
If this isn t enough another complication arises
by writing a few lines of Lisp co de to put in the
when it comes to parsing seemingly straightforward
p ersonal emacs startup le Some familiarity
variable declarations that involvea list of comma
with Emacs regular expressions will b e unavoidable
separated arguments of arbitrary length To handle
however to customize meta mode el at this level
this case a simple regular expression isn t enough
instead it is necessary to write a sp ecial purp ose
Fonti cation
utility function to match the arguments
FontLock mo de is a minor mo de provided in GNU
While all this has caused many headaches dur
Emacs which allows mo di cation of the app earance
ing the development of meta mode el it app ears
of a variety of ma jor editing mo des for di erent pro
that the FontLock patterns currently implemented
gramming or text formatting languages The basic
are go o d enough to handle most common cases as
idea is to havea number of di erently colored text
can be veri ed by loading plain mf or plain mp
faces which are used to highlightvarious language
into GNU Emacs and turning on fonti cation
elements consistently throughout all editing mo des
Finally it should b e noted that there was one
suchaskeywords function or variable names refer
more case that required sp ecial attention namely
ences to external lenames etc In addition certain
T X co de embedded in between btex :::etex or
E
language elements are also highlighted on the basis
verbatimtex :::etex in METAP OST sources
TUGb oat Volume No
From the p oint of view of syntactic highlighting the extracted list but this didn t matter to o much
it seemed b est to treat this embedded T X co de just since it had to b e done only once anyway
E
like a quoted string as it isn t interpreted in anyway In any case the resulting completion lists in
by METAP OST itself but just passed on to MakeMPX meta mode el should be fairly comprehensive and
for typesetting However to ensure prop er parsing might actually serve to give a good overview of what
this interpretation also made it necessary to retain commands are available Thus if you ever wanted
the meaning of escap e character for the backslash to know what tracing options exist just type trac
although this do esn t agree with its usual meaning followed by M TAB twice and see for youself As this
of relax in the META languages example illustrates typing trac is su cient to get
a partial completion to tracing whereup on typing
Symbol Completion
another one or two letters will b e enough to resolve
Automatic completion of partially typed keywords the remaining ambiguities
or lenames is a concept found throughout most In comparison to the completion in AUC T X
E
parts of GNU Emacs as well as in some mo dern it should b e mentioned that meta mode el do esn t
Unix shells The basic idea is to save keystrokes currently provide any context sensitive completion
by allowing one to type just the rst few letters nor do es it prompt the user to ll in the arguments
and p erform completion on pressing M TAB resulting where applicable Instead it just o ers any known
in partial completion and a display of all p ossible symbols for completion that match regardless of
matches if no unique match is found whether they would makeany sense in that context
An appreciable side e ect of symbol completion Given the versatility of the completion function it
is that it provides a wayofspellchecking keywords would certainly be p ossible to implement some of
in a programming language which helps to avoid this by preparing a more involved completion list
some of the most annoying compilation errors and some supp orting functions if desired but there
To implement symbol completion when writing are no such plans for the near future After all one
a new ma jor mo de it takes two things a completion might reasonably assume that users of METAFONT
function that do es the actual job and a list of known or METAP OST will be programmers who may be
symbols that are o ered for completion exp ected to know what they are doing whereas
As for the completion function implemented in authors of T X do cuments don t necessarily haveto
E
meta mode el there is little to say It was directly be T X macro programmers and thus might require
E
adopted from AUC T X s latex el but the frame a little more help
E
work was considerably simpli ed since it app eared
Miscellaneous Functions
unnecessary to supp ort multiple completion lists for
di erent kinds of symbols in META mo de whereas As usual in Emacs editing mo des for programming
A
it did make sense to have them in L T X mo de languages meta mode el also provides a small num
E
As for the list of known symbols there is a ber of basic editing functions that are adapted to
slightly more interesting story to tell The idea was the mo de sp eci c semantics For instance there are
to have one comprehensive list of symbols for each motion commands to move to the b eginning or end
of METAFONT and METAP OST which should in of the previous or next environment or commands
clude all primitives and macros de ned in plain mf to apply the mo de sp eci c indentation function or
or plain mp optionally augmented by the macros the standard Emacs comment region function to
de ned in standard packages suchasgraph mp or each line in an environment a region or a bu er
boxes mp in the case of METAP OST So what s the As for what kinds of META language elements
b est metho d to get a complete list of primitives constitute an environment a somewhat di erent
The answer is simple Use the source Luke set of regular expressions is used than in the indenta
Ieventually ended up with a little bit of Unix tion function Only the outermost blo ck structures
shell hackery along the lines of such as b eginchar :::endchar in METAFONT or
b egin g :::end g in METAP OST are taken into
grep primitive a zA Z mf mp web
account for this purp ose whereas conditionals and
sed s primitive a zA Z
sort mf mp prim list
lo ops are disregarded In addition de nition blo cks
suchasdef :::enddef and variants thereof are also
to extract the information ab out primitives directly
considered as de ning an environment for the con
from the WEB sources Unfortunately extracting the
venience of editing more extensive macro packages
corresp onding information from the macro de nition
However this may lead to problems if lo cal macro
headers in plain mf and plain mp didn t work out
de nitions are nested inside b eginchar :::endchar
quite as well and required a little editing to x up
TUGb oat Volume No
blo cks in which case a command on an environ that may b e merged backinto meta mode el from
ment might b e incorrectly applied to the inner blo ck time to time when the new features have proved
rather than the outer one Unfortunately there stable
do esn t seem to b e a general solution to this other
Acknowledgements
than mo difying the default regular expressions
Anumber of features implemented in meta mode el
Keybindings
have b een signi cantly in uenced by features found
Most of the mo de sp eci c editing functions provided in various Emacs editing mo des for other program
in meta mode el are b ound to fairly standard key ming or text formatting languages among them
bindings also used in Emacs editing mo des for other in particular the AUC T X package from which
E
programming languages For example M C a and I drew much of the indentation and symbol com
M C e are b ound to the motion commands applica pletion functions Emacs Lisp co de was b orrowed
ble to environments while M a and M e are retained and adapted to the new purp oses wherever p ossible
for the motion commands applicable to sentences thereby sharing all the good ideas in the true free
primarily for use in comment paragraphs Likewise software tradition while at the same time avoiding to
M C q reindents an environmentof META co de while reinvent the wheel unnecessarily Putting it all to
M q is retained as the function to re ll text in a gether and supplying the necessary knowledge ab out
comment paragraph META language features to write appropriate font
A complete listing of mo de sp eci c keybindings lo ck patterns and regular expressions however is
in meta mode el can b e obtained using the Emacs the main ingredient for which I take resp onsibility
help system Furthermore if Emacs is run under entirely myself I hop e METAFONT and METAP OST
a windowing system such as X all mo de sp eci c users using any avor of Emacs will enjoy it
editing commands are also accessible from a pull
References
down menuentitled Meta that gets installed in the
menubar when entering METAFONT or METAP OST Debra Cameron Bill Rosenblatt and Eric Ray
mo de This menu lists all available mo de sp eci c mond Learning GNU Emacs O Reilly Asso
editing commands along with their corresp onding ciates Inc September
keybindings
Richard Stallman GNU Emacs Manual Free
Software Foundation August th edition
Availability
for Emacs version
In the past preliminary versions of meta mode el
Kresten Krab Thorup GNU Emacs as a front
have b een made available by p ostings to the Usenet
A
end to L T X TUGb oat Octob er
E
newsgroup gnu emacs sources and the METAFONT
mailing list at metafont ens fr Asofversion
Rob ert J Chassell Programming in Emacs
meta mode el has b een uploaded to CTAN where
Lisp An Introduction Free Software Founda
it has found a place in the tex archive support
tion Octob er edition
emacs modes directory
Bil Lewis Daniel LaLib erte Richard Stallman
Shortly after releasing one of the early test
and GNU Manual Group GNU Emacs Lisp
versions I was contacted by Richard Stallman ab out
Reference Manual Free Software Foundation
signing a copyright transfer agreementto the Free
June edition for Emacs Version
Software Foundation to allow the integration of
meta mode el into the GNU Emacs distribution
Ulrik Vieth
which I have done now Therefore readers may
Heinrich Heine Universit at
lo ok forward to nding meta mode el as the default
D usseldorf
METAFONT or METAP OST editing mo de when the
Institut f ur Theoretische Physik II
next version of GNU Emacs eventually arrives
Universit atsstra e
While the functionality provided in version
D D usseldorf
of meta mode el is pretty stable now development
Germany
of some additional features will continue Most
vieth thphy uni duesseldorf d e
imp ortantly there are plans to implement an in
URL http www thphy
uni duesseldorf de vieth
terface to allow running METAFONT or METAP OST
in a shell bu er from within Emacs In order to
ensure stability however such development will b e
con ned to add on packages such as meta buf el TUGboat, Volume 18 (1997), No. 1 17
finally able) to impose new standards of simplified Philology typesetting,2 most of the time covering only the fun- damental properties of Arabic script, without any typographical enhancement. Was it the computers, The Traditional Arabic Typecase, Unicode, which have simplified Arabic printed script, or was
TEXand METAFONT it a deeper change in Arabic society and mental- Yannis Haralambous∗ ity? This is hard to say; nevertheless, even today, commercial computer typesetting systems are — a 1 Introduction few isolated exceptions apart — unable to reach the The first Arabic book, a 5 × 11 cm volume titled typographic quality of ’Almat¯˙ab‘ al’am¯arya.In
fig. 2, one can see different samples of printed Arabic
sw Pi (Book of the prayer of hours), was printed in 1514 by Grégoire de Grégoire in material, showing the evolution and simplification Venice and Fano, under the protection of Pope Leo of Arabic script; these examples are extreme cases: the 10th [1, p. 18 – 19]. It took about two cen- the first one is taken from a scholarly book printed turies for Arabic book printing to move to the East: in Lebanon (it contains almost all ligatures of the in 1727 the Ottoman printing agency was founded ’Almat¯˙ab‘ al’am¯arya typecase), the second from a in Constantinople and started printing using Dutch technical book printed in East Germany (a fewer types and technology [8, p. 156]. A similar institu- number of ligatures), and the third from a daily tion was founded in Cairo in 1821. newspaper printed in the U.K. (almost no ligatures). Undoubtedly a script like the Arabic one, hav- This paper describes the author’s solution to ing deep roots in calligraphy, was rather difficult this problem: (Al-Amal), a typesetting system to adapt to typography, a technique where strict based on TEX (actually TEX--XET), emulating the standardization and repetition of forms is neces- ’Almat¯˙ab‘ al’am¯arya typecase. This system (already sary. When Aldus Manutius created the first italic presented in [6] and [7]), has been recently extended font in 1501, out of manuscript calligraphic forms, to the complete set of Unicode Arabic alphabet he made a certain number of choices — and these characters; problems and open questions arising choices became a standard for occidental typogra- from this extension are discussed at the end of the phy. Similar choices had to be made for Arabic: paper. calligraphy had to be “tamed”, so that the results 2 The Cairo typecase would be homogeneous, reproducible, and flexible enough to be pleasant to the eye. Arabic letters have contextual forms, depending on
This standardization took place in 1906, in surrounding letters in the same word: a typical three 8
8 letter word will start with a letter in initial form,
A TA Cairo, when the (’Almat¯˙ab‘ al’am¯arya) typecase is defined. This typecase (see fig. 1), followed by a letter in medial form and, finally, by a divided in four parts (as opposed to “upper” and letter in final form (the hypothetical word consisting
“lower” case of the Occident), uses a total of 470 of three times the letter ‘gha’ is written G ). A characters. Astonishing as it may seem, this typeset- fourth form is used for isolated letters (this is also ting system has been kept in use until today: books the form used in crosswords or Scrabble-like games, typeset in a traditional way, all around the Arabic where letters have to placed in boxes, indepedently world, are still using the same set of characters, and of their context). Some letters appear only in iso- the same conventions and rules.1 In fig. 1, the reader lated and final form (and sometimes even only in can see the four parts of this typecase. isolated form), so that the letters immediately fol- The reader knowing the technical limitations lowing them must be written in initial (or isolated) of computerized typesetting can already imagine form, although they are located inside the word. the effect of computers on the Arabic script: not These are the basic contextual rules of the being able to cope with the complexity of the Cairo Arabic script: they are independent of style and typecase, the computer industry has tried (and was medium, and are applied in all cases, without ex- ception; they are as basic as the dot on the Latin ∗ lowercase ‘i’, or the horizontal bar of the ‘t’. The author would like to thank Michel Goossens, for — among other things — having given him access to [11], an But besides these contextual forms, ’Almat¯˙ab‘ extremely exciting book which has motivated this and forth- al’am¯arya also combines letters into ligatures,not coming developments. 1 In [4, p. 102 – 103], a book published in 1880 (!) the reader can find 30 rules for typesetting Arabic, which are still 2 For more information on the Arabic script and the strictly applied today by traditional typographers. computer see also [3] and [10]. 18 TUGboat, Volume 18 (1997), No. 1
Figure 1: The Cairo typecase—cases 1 (left, bottom), 2 (left, top), 3 (right, top) and 4 (right, bottom) unlike the ‘f’ + ‘i’ → ‘fi’ phenomenon in Latin al- some common Arabic letter, we mean all charac- phabet typesetting. In fig. 3, the reader can compare ters having the same base form as letter “foo” but
the same text with and without ligatures. The first different dots and other diacritics (for example, the
case is more-or-less the best one can obtain from ba-like family is the set of characters , , , , a standard commercial Macintosh and Windows- etc.). based Arabic typesetting system. The text of the After each entry of our list we give the coordi-
second figure is typeset in (Al-Amal), a type- nates of the type positions in the Cairo ’Almat¯˙ab‘ setting system developped by the author, and based al’am¯arya typecase (see fig. 1). These coordinates upon the combined use of a Lex/Yacc preprocessor, are notated in the following chess-like way: (a) TEXandMETAFONT; it follows the traditional type- the number of case (1–4), (b) the column (A–N, setting rules of the ’Almat¯˙ab‘ al’am¯arya typecase. counting from right to left3 with a subscript when the slot is splitted into two parts, the first part being 3 Typesetting rules for the Cairo typecase the one on the right), (c) the row, (1–8 counting In this section we give a short description of the most bottom to top). important ligatures and variant characters found in the Cairo ’Almat¯˙ab‘ al’am¯arya typecase, and their 3.1 Mandatory ligatures
use. This set of rules is described in [4, p. 102 – 103] L1. a lam-like letter followed by an alif-like letter:
8
y i $0 and has been confirmed by careful examination of , , ,etc. 1E3, 1E4, 1F3, 2A2, various printed texts of different origin. In fig. 3 2A11. the reader can compare the same text (actually L2. the second part of the word Allah (God): M . the text of the first example of fig. 2) typeset via 4A8. the Al-Amal system, with and without ’Almat¯˙ab‘ al’am¯arya ligatures. 3.2 Typographical ligatures
In the following we start by giving the manda-
A ý" L3. a lam-like letter followed by meem: ô, ,
tory ligatures (those that are part of every font),
ý ð , ,etc.1B8, 2E4, 2F4, 2G4.
then we give the second level ligatures (those that
L4. a ba-like letter followed by a ra-like one: , ,
are characteristic of the Cairo typecase), and finally ,etc.2B1, 2C1, 2D1, 2E1. we give the variant characters (form changes applied to single characters). By “foo-like”, where “foo” is 3 Remember, we are reading from right to left! TUGboat, Volume 18 (1997), No. 1 19
Figure 2: Samples of printed Arabic: Beirut 1963 (top), Leipzig 1981 (left), London 1992 (right).
t r
L5. a ba-like letter followed by a final noon-like one: L11. a ba-like letter followed by meem: , , ,
lsT cP q A A , , ,etc.2F1, 2K2. , , , , ,etc.2H2, 2I2, 2J2.
L6. a lam-like letter followed by a ha-like one: 2H1, L12. a ba-like letter followed by a gim-like one, and
A [P \P ZPY ^Q [Pü 2I1. eventually a meem: ZP , , , , , ,
etc. 2L2, 2M2, 2N2. H
L7. a ba-like letter followed by a ya-like one: ,
8
L13. a lam-like letter followed by a ya-like one: 8 ,
K w k , ,etc.2J1, 2K1.
etc. 2C4, 2D4.
s L8. a gim-like letter followed by meem: h, ,etc. L14. a kaf-like letter followed by an alif-like, or a
2L1. 8
/ H lam-like, or a final kaf-like: * , , ,
L9. a lam-like letter followed by a gim-like, and etc. Such a two-letter ligature can be extended
A ÅPR m ÅP ÅQR 2M1, eventually a meem: , , ,etc. to a three-letter or even four-letter ligature,
2N1. by adding a ya-like letter, or a ha-like letter, hl
L10. a ba-like letter followed by a ha-like: h , , or lam-like letter, or lam-alif-like ligature, etc.:
$ / A " hlw ,etc.2F2, 2G2. , , ,etc. 3H2, 3I2, 3L4, 3M4, 3H5,
20 TUGboat, Volume 18 (1997), No. 1
8
lG sr A y X b d A j A A y b g r h w n
Sw l A nY w A j k w A A w h A nY w tA qt s
8
Y d tY w y i j A g t A D tA GA
i b tHr A j h YA t A tqA h A A
8
n w Yn qf w ky A t btd s w
Y d d r b j y stn d yA A T A d A ny kt w
A y d y A l X d n w tY r g h A d A r
stT A A A d
8
lG sr A y b d A lm A A y b g r h w n
l A nY w A H w k A k w A A h A l nY w mtA qtbs m
8
w my g m g tb r r A A tA D mnSw GA
gA hm A YA t t r tqA h A m A m tY mY d klm A
b ktA tHr
8
n w Yn qf w l t A ly r ky A y mys w lmbtd y r y
r yb g d y stn d yA A T A l A yny tm d kt w
gd y A ly n w md g tY r f h A d A l Y d HA y r A A d mstT A
Figure 3: Samples of text typeset with Al-Amal, with (top) and without (bottom) Cairo typecase ligatures
3I5, 3J5, 3K5, 3L5, 3M5, 3N5, 3L6, 3M6, or a lam-like followed by two meems: 4C11, 3N6, 3K8, 3M8. 4A2, 4B2. L15. a ba-like letter, followed by a meem and a gim- L21. a sin-like, or ˙sad-like, or fa-like, or ayn-like, or
like letter: 3L2, 3M2, 3N2. gim-like, followed by a gim-like and eventually
f r
L16. a lam-like letter followed by a lam-like letter by a meem: , , , , , , , ,
r m m and eventually by a meem or a gim-like letter: , , , ,etc. table 4, columns L – 3E3, 3F3, 3G3, 3E6, 3F6, 3G6. N,rows2–8 andcolumnsH–K,rows2–
L17. a kaf-like letter followed by a meem and even- 4.
K V A tually other letters: , , , ,etc. 3H3, L22. a lam-like letter or lam-meem-like ligature, fol-
3I3, 3J3, 3K3, 3L3, 3M3, 3H6, 3I6, 3J6, lowed by a gim-like letter, and eventually a
A m
3K6, 3L6, 3M6. meem: , , , , ,etc.4C3, 4H6. md
L18. a meem followed by a gim-like letter and even- L23. the name “Mohammad” 4B8.
d tually a meem: , , ,etc. 3E4, 3F4, 3G4, 3E5, 3F5, 3G5. 3.3 Variant forms
L19. asad-like, ˙ ha-like, fa-like or kaf-like letter fol- V1. an initial ba-like letter in front of a sin-like,
r r r
lowed by a gim-like one: , , , , , ˙sad-like, ayin-like, waw-like or ha-like one grows
ly w h w A etc. 3H7, 3I7, 3J7, 3K7, 3L7, 3M7, 3H8, higher: sm , , , , ,etc. 2H6, 3I8, 3J8, 4I3. 2I6, 2J6, 2K6, 3A18, 3B18. L20. a ba-like, or lam-like, followed by meem, or a V2. a medial ba-like letter between two ba-like let- meem followed by a ba-like, followed by meem, ters, or in front of a sin-like letter grows higher:
TUGboat, Volume 18 (1997), No. 1 21
st qy t sr b , , , , , ,etc.2H5, Prof. Aqha, Univ. of Illinois) typeset in Al-Amal. 2I5, 2J5, 2K5, 3A28, 3B28, 3A7, 3B7, 3C7, In most of the cases, the extension to Unicode has 3D7, 3E7. been a straightforward task. Nevertheless, in some V3. an initial or medial gim-like letter in front of an cases the fact of applying a ligature or even just a
alif-like or lam-like letter takes a rounder closed contextual form similar to those of the basic Arabic
A A form: ly , , ,etc. 1K3, 1K4, 1L3, alphabet brought up ambiguities. These will be 1L4, 1M3, 1M4. discussed below. V4. an initial meem in front of a ra-like letter, a 4.1 Cases where contextuality leads to ha-like letter or a ya-like letter gets smaller and
8 confusion between characters
h non-hollow: , , ,etc.4C8, 4D8. 1. Letters fa and qaf. In basic Arabic, letters fa
V5. a ra-like letter following a gim-like, ta-like,˙ ayn-
(and its artificial derivative va )andqaf
like, fa-like, kaf-like, ha-like letter or a meem, have different forms: the former is longer and
j
takes a more calligraphic form: r , , flatter, while the latter is rounder and deeper.
, , , , , , ,etc. This difference is visible only in the isolated and
2L5, 2M5.
f q final forms: compare and . Since 4 Porting the Cairo case to Unicode these letters differ mainly in the number of dots (one for fa, two for qaf, three for va), the shape The first plane of ISO 10646-1, also known as difference is of minor importance, and in some Unicode, provides characters for the following lan- modern Arabic typefaces it is totally ignored. guages: Arabic (modern and classical), Farsi, Urdu, The problems arise with Unicode characters Pashto, Sindhi, Ottoman Turkish, Baluchi, Kash- 06A7 (arabic letter qaf with dot above) miri, Kazakh, Lahnda, Dargwa, Uighur, Turkic, and 06A8 (arabic letter qaf with three Berber, Hausa, Malay, Adighe, Ingush, Kirghiz 4 dots above), which use the basic shape of [12]. letter qaf, and have the same number of dots Similarly to European languages which have as fa and va. These characters are used in diacritized letters of the Latin alphabet to adapt Maghribi Arabic. In initial and medial forms, them to their phonetic needs, the languages stated in as well as in ligatures involving these forms, the previous paragraph have added diacritics to the they are indistinguishable from the basic Arabic letters of the basic Arabic alphabet. There is a slight letters fa and va. difference though: historically, Arabic alphabet was
first written without dots;5 so in a sense, dots are 2. Letters ta, noon and ya. In basic Arabic, already “diacritics”. It is only natural that these letters ta and noon have different forms: languages have first tried to use new combinations the former is longer and flatter and the latter is of dots and letter forms: almost every combination rounder and deeper. Once again, the difference
of basic form and sets of one, two, three, or even can only be seen in isolated and initial forms: four dots, over or under the word has been used to compare and . Since these letters obtain new characters. differ mainly in the number of dots (one above
The author has expanded the system to fornoon,twoaboveforta,etc.)theshape cover all these characters derived from the basic difference is of minor importance. Arabic script; in fig. 4 the reader can see an example Unicode characters 06BB (arabic letter of Sindhi text (kindly provided to the author by rnoon), 06BD (arabic letter noon with 4 This set of characters is quite complete; nevertheless, three dots above) use the letter form of the the author encountered characters not provided in Unicode, Arabic noon and the dots of the Urdu letter in four cases: for typesetting the Qur’¯an, a ba-like letter tteh and the Arabic letter tha. These letters without dot is needed [2, p. 102 – 103] (one new character), are used in Sindhi and Malay. Their initial and for typesetting old manuscripts, all characters are needed without dots (2 new characters, in ba-like and qaf-like forms), medial forms, as well as all ligatures involving Salem Chaker’s proposal for the transcription of Berber into initial and medial forms are indistinguishable Arabic script [5] (one new character), and Ahmed Lakhdar’s from the Urdu and Arabic counterparts. proposal for the writing of African languages [9] (7 new characters and 6 new diacritics). The situation is even more complicated since
5
Take for example letters (‘b’), (‘t’), (‘th’ like the Arabic letter ya shares the same initial in ‘thought’); they differ only by the number and position of and medial forms as ba, noon and friends: dots. Originally, these letters were all written without dots,
. Nevertheless, the isolated and final forms and the reader had to guess their pronounciation from the context(!). of this letter are significantly different from
22 TUGboat, Volume 18 (1997), No. 1
L yA w L w O w A y A y Pnd A L nd yw w
w M A y yA A L w O y i w A
yd r d d w A L m A A yd yd
A yA A A w A L nd w g g y yw
Po A t y m w A L yw A d yA Pw Po r A
r i h yw r w O t M fA n ng
r w w yd y w AN M AN L r w yd M A AN
A N nd d yA L A y i mg O Po A w A w w L
d yA w w A A L snd w w A y L w m Z r
P A A y n M fA nd L M nd Po bd ly N nd
yw t y M rN M bD stq Po n M
Figure 4: Sindhi text typeset in Al-Amal
those of the ba and noon letter shapes. Once The author has tried to design a ligature of
again, in basic Arabic the number and position isolated form , but the result is not entirely of dots is sufficient for determining the letter satisfying.
(ya carries two horizontaly aligned dots below). 2. The Uighur character 0675 (arabic letter Unicode character 067B (arabic letter high hamza alef) can hardly take part in a beeh lam-alef-like ligature: the hamza would be too ) has the form of ba and carries two ver- 6 tically aligned dots below; this is also the case far to the right. of 06D0 (arabic letter e) which carries the 3. The fact that ’Almat¯˙ab‘ al’am¯arya ligatures same set of dots, but has the form of an Arabic have been designed without taking into account ya. Furthermore, 06D1 (arabic letter yeh Indic characters, makes many ligatures with
with three dots below) carries three dots non-standard dots ambiguous: is »Q the combi-
below, exactly as does Arabic letter tha: the nation of t and (06A5 arabic letter feh
former has the letter form of a ya, while the with three dots below and the standard ¹ latter the one of a ba. Arabic hah) or of × and (06A1 arabic let- ¯ ter dotless feh and 0686 arabic letter Arabic and Sindhi letters kaf. 3. In Arabic, tcheh)? Theoretically, one can distinguish the letter kaf is written with an oblique ascender them by slightly moving the dots to the right
stroke in initial and medial form, and with a ¸QY in the former case ( »Q vs. ); but still the two hamza-like diacritic in final and isolated form. forms are very close graphically, and it may be 06A9 arabic let- Sindhi uses a kaf-like letter ( difficult to the reader to distinguish them at ter keheh ) which has oblique ascender strokes first sight. in all forms and no hamza-like diacritic. This letter is indistinguishable from the Arabic kaf, 5 Technical details in initial and medial forms, as well as in all 5.1 Preprocessing ligatures involving these forms. The extended Al-Amal system consists of four mod- 4.2 Cases where ligatures obstruct proper ules, as shown in fig. 5: diacritization of characters 1. re-encoding into the (extended) Unicode encod-
1. The Pashto ring (as in ) is incompatible with ing;
the ba-like + gim-like ligature (for example ]Q ) 2. standard contextual analysis and processing of and the initial/isolated ba-like + meem ligature the mandatory ligatures;
(for example ). Either the ring must be 6 NottomentionthefactthatintheQur’¯an one finds designed like “a drop that hangs” — a dubious a lam-alef ligature with a central hamza, not included in esthetic result — or the ligature must be broken. Unicode. TUGboat, Volume 18 (1997), No. 1 23
Cairo Re-encoding Contextual Typesetting with ligatures Input to Unicode analysis typecase ligatures
Typesetting without ligatures
Preparation to output Typesetting Output
Figure 5: The Al-Amal internal structure
3. Cairo typecase ligatures processing (optional); [6] Yannis Haralambous, Towards the revival of and traditional Arabic typography... through TEX, 4. output preparation (conversion into TEXcode). Proceedings of the EuroTEX92 conference, Prague, 1992 The first three modules are independent of TEX. To avoid ligatures one simply removes module 3 [7] Yannis Haralambous, Typesetting the Holy from the processing chain. Preprocessors have been Qur’¯an with TEX, Proceedings of the 2nd Inter- written in C, using Flex and Bison tools: writing a national Conference on Multilingual Comput- grammar for Arabic ligatures avoids tedious pattern ing (Latin and Arabic script), Durham, 1992. matching. [8] Klaus Lagally, ArabTEX — Typesetting Arabic with Vowels and Ligatures, Proceedings of the 5.2 The fonts EuroTEX92 conference, Prague, 1992 The Al-Amal have been designed in the METAFONT [9] Ahmed Lakhdar-Ghazal, Caractères arabes dia- language, to benefit from the maximum possibilities critiques selon l’ASV-CODAR (pour imprimer of optical scaling. Many ligatures have been split in les langues arabes), Institut d’Études et de several parts and are re-combined by TEX (this is es- Recherches pour l’Arabisation, Rabat, 1993. sentially the task of the preprocessor module). One [10] Pierre MacKay, Typesetting problem scripts, can consider these fonts as glyph containers, provid- BYTE 11, 2 1986, 201–218. ing glyphs which T X combines into characters and E [11] Roland Meynet, L’écriture arabe en question, ligatures. This approach has allowed minimization Dar el-Maghreb Éditeurs, Beyrouth, 1971. of storage space and time needed for design the font. The author was able to produce all possible Cairo [12] The Unicode Consortium, The Unicode Stan- typecase ligatures on the Unicode Arabic character dard, Version 1.0, Vol. 1, Addison-Wesley, 1991. set, using only six 8-bit (partially filled) font tables,7 8 consisting of less than 1500 glyphs. See tables 1 – 6. ⋄ Yannis Haralambous Atelier Fluxus Virus, 187, rue References Nationale, F-59 800 Lille, France [email protected] [1] Josée Balagna, L’imprimerie arabe en occident, URL: http://pobox.com/~yannis Maisonneuve & Larose, Paris 1984. [2] Syed Barakat Ahmad, Introduction to Qur’anic Script, Curzon Press, London, 1985. [3] Joseph D. Becker, Arabic Word Processing, Communications of the ACM, 30 (7), 1987. [4] Claus Faulmann, Das Buch der Schrift, en- thaltend die Schriftzeichen und Alphabete aller Zeiten und aller Völker des Erdkreises,Vi- enna, 1880 (reprint by Franz Greno, Nördlin- gen, 1985). [5] Yannis Haralambous, Un système TEXberbère, Études et documents berbères, 11 43–54, 1991.
7 In a forthcoming implementation of Al-Amal to Ω,these fonts will be merged into one 16-bit (virtual) font, and contextual analysis, as well as Cairo typecase ligatures, will be handled by Ω Translation Processes. 24 TUGboat, Volume 18 (1997), No. 1
"x0 "x1 "x2 "x3 "x4 "x5 "x6 "x7
"1x
"2x
"3x
A C D E F G
"4x
H I K L M N O
Q R S T U V W
"5x P
X Y Z
a b c d e f g
"6x
h j k l m n
q r s t u v w
"7x p
x y z
"8x
"9x
"Ax
"Bx
"Cx
"Dx
"Ex
"Fx
"x8 "x9 "xA "xB "xC "xD "xE "xF
Table 1: Table of the amal0-10 font (Basic glyphs). TUGboat, Volume 18 (1997), No. 1 25
"x0 "x1 "x2 "x3 "x4 "x5 "x6 "x7
"1x
"2x
"3x
A B C D E F G
"4x
H I J K L M N O
Q R S T U V W
"5x P
X Y Z
a b c d e f g
"6x
h i j k l m n o
q r s t u v w
"7x p
x y z
"8x
"9x
"Ax
"Bx
"Cx
"Dx
"Ex
"Fx
"x8 "x9 "xA "xB "xC "xD "xE "xF
Table 2: Table of the amal1-10 font (Ligatures I). 26 TUGboat, Volume 18 (1997), No. 1
"x0 "x1 "x2 "x3 "x4 "x5 "x6 "x7
"1x
"2x
"3x
A B C D E F G
"4x
H I J K L M N O
Q R S T U V W
"5x P
X Y Z
a b c d e f g
"6x
h i j k l m n o
q r s t u v w
"7x p
x y z
"8x
"9x
"Ax
"Bx
"Cx
"Dx
"Ex
"Fx
"x8 "x9 "xA "xB "xC "xD "xE "xF
Table 3: Table of the amal2-10 font (Ligatures II). TUGboat, Volume 18 (1997), No. 1 27
"x0 "x1 "x2 "x3 "x4 "x5 "x6 "x7
"1x
"2x
"3x
A B C D E F G
"4x
H I J K L M N O
Q R S T U V W
"5x P
X Y Z
a b c d e f g
"6x
h i j k l m n o
q r s t u v w
"7x p
x y z
"8x
"9x
"Ax
"Bx
"Cx
"Dx
"x8 "x9 "xA "xB "xC "xD "xE "xF
Table 4: Table of the amal3-10 font (Ligatures III). 28 TUGboat, Volume 18 (1997), No. 1
"x0 "x1 "x2 "x3 "x4 "x5 "x6 "x7
"1x
"2x
"3x
A B C D E F G
"4x
H I J K L M N O
Q R S T U V W
"5x P
X Y Z
a b c d e f g
"6x
h i j k l m n o
q r s t u v w
"7x p
x y z
"8x
"9x
"x8 "x9 "xA "xB "xC "xD "xE "xF
Table 5: Table of the amal4-10 font (Ligatures IV). TUGboat, Volume 18 (1997), No. 1 29
"x0 "x1 "x2 "x3 "x4 "x5 "x6 "x7
"1x
"2x
"3x
A B C D E "4x
"5x
"6x
h i j k l m n o
q r s t "7x p
"x8 "x9 "xA "xB "xC "xD "xE "xF
Table 6: Table of the amalf-10 font (Vowels and diacritics). 30 TUGboat, Volume 18 (1997), No. 1
A Medieval Icelandic manuscript∗ the three or four manuscripts that the institute and The making of a diplomatic edition Helgi deemed suitable and most urgent. Apart from some fragments, the Icelandic Hom- Andrea de Leeuw van Weenen ily Book is the oldest extant Old Icelandic manu- In November 1993 my edition of the Icelandic script, dating from around 1200 and containing on Homily Book1 was published by the Stofnun Arna´ its 102 parchment leaves (204 pages) some 60 ser- Magn´ussonara ´ ´Islandi (SAM)2 after having been ‘in mons. By its age alone, this manuscript is of the press’ for a period of 19 years. If it had not been for greatest interest for the study of the Old Icelandic TEX, this period might easily have been extended language; but it is also considered to be an example indefinitely. Looking back, work on the Icelandic of good style. Homily Book can be divided into three stages: the Work on the transcription started in the sum- scholarly work, the attempts at printing before TEX, mer of 1973. At first I worked from a set of pho- and the typesetting with TEX. tographs, later I was able to use the manuscript itself.3 After the first year the transcription with the 1 The scholarly work critical apparatus was finished, and the introduc- My involvement with this edition, or with Old Ice- tion, which was going to concentrate on orthography landic scholarship in general, came about almost and morphology, was well under way. Meanwhile, by accident. I arrived in Iceland in 1971 with my the staff at the institute had been keeping an eye husband, who had taken a temporary job at the on my work, and had offered to publish the edition in one of their series as a combined facsimile4 and University of Reykjav´ık, and my two small sons; my 5 knowledge of Icelandic at that time could easily find diplomatic edition. I gladly accepted their offer, place in half a column of this journal. In order to but should perhaps have sensed the problems that escape the drudgery of diaper laundry I enrolled in would develop afterwards when the project meeting the “Icelandic for foreigners” program at the univer- was nearly exclusively devoted to the choice of pa- sity (H´ask´oli ´Islands), where I became enthralled in per, rather than to editorial principles, deadlines, my second year by the secrets of paleography and special requirements, and the like. Old Icelandic grammar. So when I had passed my I left Iceland in 1974 with the promise that examination for the Bacc. Phil. Islandicae degree, I typesetting the transcription would start next week. looked around for something useful in that direction Famous last words. During the next two years I to occupy me in my third and final year in Iceland. finished writing the introduction and fulfilled the requirements of the University of Utrecht for a mas- A suggestion by Helgi Guµmundsson, associate professor of Icelandic at the University of Reykjav´ık, ters degree in Old Germanics. As typesetting in Ice- to write a doctoral thesis and to choose an edition land still had not started, I typed the introduction, with a thorough grammatical analysis as the topic pasted the needed corrections into the transcription did not strike me as a realistic option. I had majored and handed the thesis in as typescript, thinking that inmathematics,sowouldhavetogoalongwaybe- it might well be several more years before the book fore getting to a doctorate in a completely different got printed; but I never suspected that it would take field. Nevertheless, he insisted that shortcuts could 17 more years — or that I would have to be my own be found and that preparing such an edition while typesetter. I had the right resources was a sensible thing. Al- though I did not believe him at the time, he turned out to be right. Anyway, I let myself be talked into this undertaking and after some consultations with 3 the SAM, I choose the Icelandic Homily Book from The Royal Library loaned the manuscript to Iceland and later kindly granted permission for it to be taken out of its binding to be photographed for the facsimile edition. 4 This required photographing the whole manuscript. For ∗ An earlier version of this article appeared in MAPS this it had to be taken out of its binding. This was done in 14 (1995) pp. 31–34. 1975. The delay in printing had also delayed the rebinding. 1 Andrea de Leeuw van Weenen, ed., The Icelandic Hom- In fact, I found that, in February 1996, the manuscript was o ily Book, Perg. 15 4 in the Royal Library, Stockholm, ´Islensk still in loose pages. Handrit/Icelandic Manuscripts, Series in quarto vol. III, Stof- 5 The edition is a diplomatic one. That is, it aims nun Arna´ Magn´ussonar ´a ´Islandi, Reykjav´ık 1993. pp. 436 at reproducing the manuscript. In a diplomatic edition of + 204 plates. the strictest kind, even the abbreviation marks would be 2 This institute in Reykjav´ık keeps most of the existing reproduced. Here the abbreviations are expanded in italics. Icelandic manuscripts and is devoted to their study and The facing facsimile allows the reader to observe the originals publication. of such abbreviation marks. TUGboat, Volume 18 (1997), No. 1 31
2 Typesetting, the years before TEX the book in Iceland, that had now certainly gone. In 1974 all typesetting on Iceland was still done in So it was early 1989 when the marked proofs were lead. The transcription for my project required a returned for the second time to the typesetter. But number of unusual characters, and it turned out that when I arrived in Reykjav´ık some months later, I not only did the typesetting firms not have these found that the machinery had again been replaced characters, they did not exist in the Monotype cata- and that the typesetter was planning to start from logue. So they would have to be specially cut for this scratch. Again. publication. The various firms that were approached By this time I had about 10 years’ experience were understandably reluctant to invest in this, as with computers and I was quite sure that conver- there was no guarantee that the characters could sion was possible. Moreover, I had at some stage be used for other books. These negotiations took requested and obtained copies of the typesetting files. Admittedly, it had not been easy to decipher several years — in the small Icelandic community, 8 firms could be approached only one at a time, and these, but I had copies on DOS disks of the original most took their time to think the proposition over. files and conversions (via a SNOBOL4 program) of In 1979 the news came that one firm had purchased these files to ASCII, where the typesetting codes ´ phototypesetting machinery of the matrix variety had been removed. At this stage the Stofnun Arna and that they were willing to start work on the Magn´ussonar was as opposed as I was to going transcription. Slowly, the proofs started to come. through the whole troublesome procedure again — But with them came a surprise. it was becoming clear to us that, with the methods I had believed that proofreading would be my of the institute, we would always be limping behind responsibility, but now I found that proofs of books the continuous advances in technology. to be published by the Stofnun Arna´ Magn´ussonar6 So the disks were sent to Iceland and in due were habitually read by three independent read- course of time new proofs arrived. But after the ini- ers—the editor of the edition, one of the senior staff tial joy that conversion to the new machine was pos- members7 and a junior staff member. Additionally, sible, a closer look brought great disappointment. the proofreading did not mean comparing the proofs The font used looked decidedly irregular and the with the typescript but with the manuscript or the kerning of the high ‘s’ ( @ ) was absolutely ugly. But photographs, thereby checking not only the work even worse, many errors had crept in. A systematic of the typesetter but also that of the editor. This study of the errors identified on the first few pages meant that proofreading took quite some time. For brought me to the conclusion, later confirmed: a the SAM staff, it was one of the many jobs they had conversion program had been written, and where it to do besides their own research. When we disagreed produced erroneous results, rather than correcting about a reading there were lengthy discussions by and rerunning the program, they had opted for man- mail, which usually were only resolved during one ual correction of the output file, but such corrections of my visits to Iceland. So, when in 1983, we were had not been carried out very systematically. finally in agreement about the corrections to be Considering the state of affairs and the possi- made and sent the corrected proofs to the typesetter, bilities for correcting the files, I decided that the it was a very unpleasant surprise when we were told best thing would be to get my hands on their files that he had just got himself a new phototypesetting and repair them by comparison with mine. As this machine and could not convert the material he had required only a physical conversion to DOS disks, on punch tapes to this new machine. But he would it seemed easy enough. Unfortunately, this could have the thing typeset anew as soon as he could. In not be done in Iceland, but had to be handled in the end this took a year. Denmark by the manufacturer of the machinery, And so the whole circus started again in the and after some phoning and explaining, two disks autumn of 1984: proofreading in triplicate. There arrived, which were not too difficult to decipher. were fewer cases to discuss between us three; on the As soon as I had corrected a couple of pages, I other hand, the work went a lot slower. I was both returned the disk, and waited with some optimism in the final stage of another project and taking up for a corrected proof. No such thing — only a pan- a new job which required a lot of reading up — if icky fax that the disk could not be read. Some there had ever been any feeling of urgency about weeks of multilateral discussion followed between the institute and the typesetter in Reykjav´ık, the 6 Det Arnamagnæanske Institut in Copenhagen follows the same policy. 8 The original 8-inch disks were written in a proprietary 7 In my case, Stef´an Karlsson, now director of the Insti- format. It required the help of a specialized publishing house tute and professor at the University of Iceland. (Brill) to convert the files into DOS format. 32 TUGboat, Volume 18 (1997), No. 1
technical staff of the manufacturer of the typesetting only minor corrections, so the relevant part could machine in Denmark, and myself in Leiden (the be inserted after each page automatically. And so, Netherlands). This discussion was not made any while EDMAC was available, it was not considered. easier by the lack of a common language. In the Proofreading of the transcription could be min- end, it became clear that the lack of expertise on the imal, as the original ASCII files had been corrected Icelandic end, combined with the distances involved, and only the correctness of the conversion and the made it highly unlikely that the problem would ever working of the TEX macros needed to be checked. be solved. In the previous stages, no attempt at typeset- By this time, 1990, I had gained some experi- ting the introduction had been made. Over the years
ence with TEXandMETAFONT , enough at least to I had had serious discussions with Stef´an Karlsson be confident that the job could be done, and luckily about the arrangement of some of the quires. As this not enough to foresee all the problems involved. involved drawing and redrawing the figures depict- Moreover, I had already keyed in the apparatus, ing those quires, I had at some stage done the neces- together with all the points raised in connection sary drawings within LATEX’s picture environment. with them in 10 years of correspondence, and even As a result I adopted LATEX for the production of the A made a few METAFONT characters needed there. So introduction; however, neither LTEX’s book style I wrote a letter to Iceland enumerating the possi- nor the NTG’s boek style were to the liking of the bilities open to us; these ranged from typesetting institute. I therefore had to write my own style from scratch (for the third time) via various methods file—or rather, to fiddle with boek.sty and its involving conversion to the new machine, to doing it attached files to get the required results. A small myself with TEX. I outlined the disadvantages and surprise was having to define a new strutbox,as advantages and the fact that, in my opinion, some the normal one suits 10pt text only. methods were so impractical and relied so much on Some of the many tables in the introduction factors beyond our control that I was not willing to would only fit in landscape. As they all required cooperate in them. Probably the members of the a full page, I took the easy way out and produced staff of the Stofnun Arna´ Magn´ussonar were then them separately. about as fed up with the whole thing as I was, so The introduction required even more special they agreed that I should have a go with TEX. characters than the transcription, as the various The edition had now been ‘in press’ for 15 years. abbreviation marks, which are expanded in italics This time was rather evenly divided in three periods: in the transcription, have to be represented. On trying to find a suitable typesetting firm/system the other hand, it had been decided already in 1974 (1974–1979), first phototypesetting system (1979– that the survey of the characters occurring in the 1983), second phototypesetting system (1983–1989). manuscript should contain drawings by hand of the various characters and their variants. Here small 3 Typesetting, the years with TEX gapsweretobeleft,tobefilledinbyhandinthe 3.1 Picking up the pieces final 1270dpi copy. Apart from the transcription, which by now had 3.2 Design constraints gone through two failed typesetting attempts, the book was also to have an extensive introduction The edition of the Icelandic Homily Book had two (215 pages in the finished product). Again the components: the introduction (written after the transcription was tackled first, and this time on the body of the text had been produced), and the text base of the machine-readable version produced in itself. For the text, there were to be a series of the second attempt. This had been converted to notes at the bottoms of pages (the critical and pa- a simple ASCII-based encoding scheme of my own leological apparatus), line numbers in the left mar- devising and had been used for searching, concor- gins, occasional margin notes in the right, complex A font combinations throughout, and forced line and dancing, etc. The features of LTEX were irrelevant page breaks. The introduction would be different forthispart,soplainTEX was used to produce this part. As I had to write a conversion program9 in structure, with section headings at various levels, footnotes, tables, and numerous citations from the anyway to convert the files in my code to TEXfiles, I could easily include line numbering as well. The text. It also had a preface, table of contents, and a bibliography. apparatus existed already in TEX form and needed The book had to appear in a series and was 9 planned as a combined facsimile and diplomatic edi- Again, the conversion was handled by a SNOBOL4 program. tion, with photographs and transcription on facing TUGboat, Volume 18 (1997), No. 1 33
pages (see illustrations). This meant that both page the \llap for the line numbers, were put directly breaks and line breaks in the transcription were pre- into the TEX files by the conversion program. determined by the manuscript, not by the software. As well, presentation of the manuscript required a 3.3 Font issues large paper size (30.3 × 23.3mm), which in turn The next problem concerned the special characters
meant using a 12-point font. It came therefore as that had caused us difficulties right from the begin- an unpleasant surprise that the Computer Mod- ning: , to name a few, and of course .The ern fonts which I wanted to use were significantly latter could be taken from the Icelandic font, but the
wider than the fonts used previously, and, more to others had to be made with METAFONT . Some were
the point, that the resulting lines did not fit the easily constructed: the high ‘s’ ( ) only required given page width. After much hesitation I decided removing the horizontal stroke from the ‘f’, and of to decrease the width of the characters by about course the introduction of quite a few new ligatures.
10%. As the line breaks are determined by the Others, however, required adding a diacritic to a
manuscript, I could have set the \hsize to a rather character: , , , . Still others required more
arbitrary large value had it not been for the biblical METAFONT skills: or . I must stress that I
references which occasionally had to appear in the was, and am, far from mastering METAFONT ,andI right margin. Setting \hsize to 175mm and setting remember with embarrassment the time that I had
the references flush right in the line resulted in only produced a version of an ‘o’ with a squiggle ( )that one or two places where line and reference clashed. looked acceptable in isolation, but different when In these cases a solution was found by moving the inserted in a font. Only when I made a test font reference one line down. with just two copies of this character did I realise The paperheight too was not unproblematic. that the ‘o’ was drawn with the pen inherited from Some manuscript pages had many more lines than the previous character, and that I had introduced a others, and there was a critical apparatus that also smaller pen for the tail part. had to be accommodated as a whole at the foot of Finally, I made roman, italic, bold, and bold the page and could not be allowed to float to the italic fonts which consisted of the same characters next page. If I chose a page height that would fit all as their cm counterparts, minus the Greek letters, pages, the majority would look ugly, as they would but with the addition of the special characters and have far too large a gap between text and apparatus. the small capitals needed (see below). In the part So after some experiments I choose a page length of the fonts taken over from the cm fonts I made a that fitted most pages with the apparatus at the small change to the character ø — not to its shape, bottom of the page. The few overlong pages had but to its height. The height of this cm character is a special page height and the apparatus directly the height not of the ‘o’, but of the diagonal stroke. following the text. This results in the accent above ø standing higher After some experimentation \vsize was set to than that over ‘o’: ø´´ o. By reducing the height of
270mm and pages arranged as follows: first a head- the ø to the height of ‘o’, the accents come at the
o line containing the folio number (this is suppressed same height: . in the illustration), then the body of the text, then The various fonts were produced in a 300dpi the apparatus part. For normal pages this took the version for proofreading and a 1270 dpi version for form: \vfill, text of the apparatus, and finally the final production. The parameetrs were taken 46.8mm vertical white space. For overlong pages, from the cm fonts as well, apart from the necessary the apparatus followed the text after a 2.6pt gap adaptation mentioned above regarding the width of and was followed by a \vfill. In both cases the characters. apparatus was printed with a linewidth of 145mm. Besides the problem of designing special char- TEX’s habit of stretching and shrinking spaces, acters, there were also problems with the integration much as I value it elsewhere, did not improve the of various typefaces. Due to the diplomatic nature readability of this text, so I disabled it by redefining of the transcription, roman characters, italics, and \fontdimension’s 2, 3, and 4 for all relevant fonts small capitals can occur within a single word, and (roman, italic, bold, and bold italic). As well, the this poses problems. distances between the lines were set to a fixed value. The transcription follows the manuscript in its Otherwise, the style file for the transcription use of small capitals, which are employed to indicate consisted only of macros to arrange the fonts at double consonants.10 Normally smallcaps are larger various sizes into families, and shorthands for the 10 This was one of the methods used by medieval Icelandic special characters. All other coding, for example, scribes to put as much text as possible on the expensive 34 TUGboat, Volume 18 (1997), No. 1
Figure 1: Photograph TUGboat, Volume 18 (1997), No. 1 35
Figure 2: Transcription 36 TUGboat, Volume 18 (1997), No. 1 than the corresponding romans, as can be seen when further conversions, TEXable text. Nor was the I use this strategy for English and write corect for second proofreading a waste, since all errors spotted ‘correct’, planing for ‘planning’. This makes the were corrected in the machine-readable text. page look very jumpy, so I scaled down the small- I found it possible for somebody who is far from caps. However, this was not completely successful. beingaTEXpert to produce this rather complicated Even with the large number of parameters for the edition with the support of the ever helpful TEX cm fonts, there seemed to be a relationship between community. In particular, I would like to thank Kees the thickness of various strokes. I feel that a small van der Laan and Piet van Oostrum, who were both capital that has to fit within a word should be very helpful, providing me not only with TEXtricks parameterized in a different way, but for that task I but also with their explanations. The final result lacked the time. looks far better than two of the attempts before The transcription also has italics and romans TEX, and certainly as good as the third. mixed within words. I had thought that the italic If I were to be confronted with the same prob- correction would take care of that problem, but it lem now, I would certainly opt for TEX again. Also, did not. So I had to figure out experimentally the many of the other decisions would be taken in essen- amount of kerning needed for each pair of roman- tially the same way. The only decision I might re- italic and italic-roman that occurred. Again, this consider is the choice of LATEX for the introduction. can certainly be improved upon by someone with I found the relations between the macros of LATEX a designer’s eye. I can only say that this kerning itself and its various style files hard to understand. is a great improvement over the results without It was therefore difficult to achieve the requested the kerning. As the TEX files for the transcription design and format changes. pages were produced by a conversion program, these The example of the Homily Book has served to explicit kernings had already been automatically persuade a number of colleagues that TEX can rescue inserted. their work as well. In most of those cases a book And while all of this had been resolved for has been produced with a word-processing program, the actual transcription pages, the introduction still the publisher wants camera-ready copy, but the only existed as a typescript and contained thou- requirements of the publisher (and sometimes even sands of quoted words from the transcription. I was that of simple readability, or of the conventions not looking forward to typing in all those explicit in the specific field) cannot be met by the word- kernings, so I decided to solve the kerning problem processing program. Up to now TEX has always by combining romans and italics in a single font provided the necessary functionality. while taking care of the kernings in the ligature ta- bles. The roman and italic smallcaps which occurred ⋄ Andrea de Leeuw van Weenen within the transcription were placed in this same Department of Comparative font. This arrangement meant that italics could Linguistics Postbus 9515 not be accessed by the usual \it command, but via 2300 RA Leiden macros: \ia for italic-a, and so on. Since at most [email protected] only one or two consecutive italic characters occur, this made the typing not too onerous. 4Conclusion
The flexibility of TEXandMETAFONT have made it possible to produce a publication which might otherwise never have made it to the printing press, as another typesetting + proofreading cycle would probable have taken even longer and left us even further behind in the technology race. Looking back, the first typesetting and its proofreading were completely wasted. The second typesetting was not, although in a very round-about way, since it yielded a computer-readable, and after parchment. Another was the frequent use of abbreviation signs. TUGboat, Volume 18 (1997), No. 1 37
2 figures and bibliographies. LATEX is introduced at Book Review the end of this chapter as an example style library. The book ends with eight appendices. The first six are divided by tasks and cover TEX commands Book review: Writing With TEX and in more detail. Tables and output routines are ad- TEX&LATEX: Drawing & Literate dressed here, as well as a more about symbols, boxes Programming,byEitanGurari and penalties. Appendix G is a complete catalog of Michael D. Sofka TEX commands, each with a short explanation and example. The final appendix is a short bibliography Eitan M. Gurari, Writing with TEX. McGraw-Hill, of TEX books, newsletters and electronic resources. New York, 1994, ISBN 0-07-025207-6. I have always been of the opinion that as a general rule a programming language should be T X&LAT X: Drawing & Literate Eitan M. Gurari, E E describable in about 100 pages. T Xiscomplex Programming E . McGraw-Hill, New York, 1994, ISBN enough that a short reference is difficult. In Writing 0-07-911616-7. With TEX, however, Gurari has made a decent 230- page attempt. This economy of language, however, In this pair of books Gurari has written primers for comes at the price of ease of learning and depth of A using TEXandLTEX for writing, programming and knowledge. If you have already learned the basics drawing figures. The books are short and to the of TEX and wish to know more, there are other point, with most commands being introduced and books which go into the details of macros, line explained in one or two paragraphs. If you are new and page breaking and output routines. If you are to T X, or do not like to experiment, these books E new to TEX, and not skilled with programming, may not be for you. If, on the other hand, you Gurari offers little help. On the other hand, Writing are looking for a short and comprehensive review With TEX is a decent TEX reference. Gurari’s of TEX, or if you want to draw figures and write explanations are clear, concise and independent, programs in TEX, Gurari is a good choice. which is what a reference manual should be. Writing With TEX is a TEX reference manual. A 1 TEX<EX: Drawing & Literate Program- It starts with a three page Getting Started chap- ming is a reference manual for two macro packages ter which explains how to run T X. Most of the E written by Gurari: DraTEX for drawing pictures, remaining book is divided into two parts. Chap- and ProTEX for literate programming. The book ters 2–10 cover TEX as a Formatting Language.This title is misleading since it is really about drawing includes simple text, Fonts, Layout of Pages and and literate programming (and specifically about Mathematical Formulas (each in five pages!), inserts, DraTEX and ProTEX). Very little of the book deals boxes, and basic programming constructs such as A generically with TEXorLTEX. Groups and scanning. The concepts and commands The book begins with a short Getting Started introduced, however, are only the essentials needed A chapter which explains how to run TEXandLTEX. for formatting. Chapters 2–5 cover the basics of TEX (including T X as a Programming A Chapters 11–16 is called E Mathematical Formulas and LTEX ) in a whirlwind Language. It covers Macros, Data Types, Selectors 20 pages. Given how peripheral this is to the actual and Auxiliary Files (file I/O), each of which is a macro packages, I wonder why these chapters are basic programming unit. These chapters are a little included—it is unlikely somebody will purchase longer, averaging ten pages each. The programming A TEX<EX who is not already familiar with TEX. is rounded out with two chapters covering macros Still, Gurari is good at providing short and accurate and character codes in more detail. This includes A explanations, and I dare say the five page LTEX \let, \edef, \csname ...\endcsname, token expan- chapter is enough to get started. sion, active characters, changing character codes Chapters 6–11 cover DraTEX primitives. These and other subtleties of writing TEX macros. include macros for line and curve drawing, painting Chapter 17 is titled Environments For Writing. and clipping, coordinate systems (including three- Macros are described for basic document structures dimensional viewpoints), repetition and data paths, such as \Chapter and \Section,aswellasenviron- creating objects with tables of data, and arithmetic ment blocks, lists, cross-references, table of contents, operations. DraTEX is indeed an impressive package which does all of its drawing in TEX. This does, 1 When chapter titles appear appear as part of a descrip- tion they will be italicized. 2 Version 2.09 38 TUGboat, Volume 18 (1997), No. 1
however, impose some limits on what can be drawn. or if you like the idea of providing a single, portable For example, rotated text is not possible without source document containing both text and figures, 4 special fonts or \special support, and the clipping DraTEX and AlDraTEX are a good choice. and three dimensional commands are limited. DraTEX and ProTEX are available on CTAN Chapters 12–20 cover High-Level Drawing Fa- and at ftp.cis.ohio-state.edu in the directory cilities. These are drawing templates provided by /pub/tex/osu/gurari/ (they have been updated AlDraTEX, a macro package based on DraTEX. The since the book and disk were published). This templates include macros for pie, XY and bar charts. includes an Examples.tex file, which is a ProTEX There are three chapters introducing diagramming, file of examples from the book. There is no better a chapter on tree diagrams and another on labeled way to sample the abilities of ProTEX and DraTEX graphs. Gurari has authored a textbook on compu- then to compose this file. The result is a typeset tational theory, and the graphs and trees provided description of the examples, and 68 example figures by AlDraTEX seem sufficient for such a book. and exercises. Chapters 21 and 22 cover Literate Programming On the whole, both books are well written and 3 using ProTEX. Literate programming is a program- cover a lot of material in few pages. There are, how- ming style invented by Donald Knuth, and used for ever, no answers provided for the exercises so expect writing TEX, Metafont and the suite of programs to spend time experimenting. I recommend these used for font management. Literate programs con- books only for experienced programmers who need sist of a mixture of code and code documentation. a concise TEX reference manual, or who would like A set of filters converts the input file into either the to usea TEX based drawing or literate programming source code for the program, or a typeset document package. describing the operation of the program. Gurari’s ⋄ ProTEX is a macro package which uses TEXasthat Michael D. Sofka Computing Information Services filter. ProTEX’s output is a typeset document and auxiliary files consisting of the code described in the Rensselaer Polytechnic Institute Troy, New York 12180-3590 document. The code may be Pascal, C or any other [email protected] language including TEX. The book ends with five appendices. The first two cover the Implementation of ProTEX,andthe use of PostScript Figures Within TEX. ProTEX implementation details are provided so that users can adapt ProTEX to their own needs. PostScript is introduced as a standalone page description lan- guage, and as a way to supplement the abilities of DraTEX. In this mode it is similar to Timothy van Zandt’s PSTricks. Appendix C is a Catalog of Commands covering TEX, LATEX, DraTEX, AlDraTEX and ProTEX. Each command is presented with an example of its output. It would be helpful if the commands were cross- referenced with their explanation in the book. The last two appendices are a bibliography and information on acquiring the macro packages. The bibliography includes references on drawing and literate programming. The macros are available on a disk which accompanies the book, but updates and examples are online. Regardless of the systems on which you work, there are a variety of affordable (frequently free) drawing programs available. For complex figures 4 these are usually a better choice. For simple figures, You will, however, require a version of TEX compiled with a large main memory array, and a fairly fast computer. 3 This is the third package I am aware of called ProTEX. I ran the samples on an RS/6000-250 with Web2c, and a The other two are Professional typesetting systems based on 100Mhz Pentium with MikTEX. On both systems perfor- TEX. mance was acceptable.
TUGb oat Volume No
the prop er way coming to a compromise between
readability and abstract typesetting rules
Tutorials Surveys
I will discuss here those few tricks that physi
cists and engineers not mathematicians must know
Typesetting mathematics for science and
in order to satisfy the international regulations and
technology according to ISO XI
to distinguish similar symbols with di erent mean
ings and ultimately in order to cop e with the ISO
Claudio Beccari
regulations and the recommendations issued by
Abstract
the International Union of Pure and Applied Physics
IPU
Mathematicians set mathematics into type di er
ently from physicists and engineers the latter re
Upright and sloping letters
quire some particular tricks in order to satisfy
The main and p ossibly the only di erence b etween
ISO XI and to distinguish similar symbols that
mathematical vs physical mathematics lies in
A
have di erent meanings The L T X commands
"
E
the use of upright and sloping letters Scientists and
to implement such tricks are shown and explained
technologists should use upright letters much more
Introduction
often than mathematicians
A
In math mo de L T X chooses normal letters
"
E
As D E Knuth p oints out very well in The T Xbook
E
from the math italics alphab et which includes
the strength of T X and its derived dialects
E
also the Greek lowercase ones Both Latin and
A
among which L T X outclasses all others lies
"
E
Greek letters have a sloping shap e the former b eing
in the abilitytotyp eset b eautiful mathematics the
in italics the latter just sloping to the right with a
variety of symbols the shap e of the op erators the
slop e angle that matches the one of the italics char
spacing b oth vertical and horizontal the sizes of
acters Just the upp ercase Greek letters by default
rst and second order sub and sup erscripts make
are upright but it would b e very easy although un
A
L T X the b est software available to day for type
E
usual to set them with a sloping shap e b ecause they
A
setting mathematics Of course L T X doesawon
E
app ear in the math italic font and in all the text
derful job also with plain text tables cross refer
italic and other slanted fonts
encing indexing and so forth but other programs
In the following I will call roman the upright
may p erform well with the latter tasks what other
shap e and italic the one that T Xies are used to
E
programs really cannot do is the excellentwork with
asso ciate with math italics The choice of the word
mathematics all this is not surprising since T Xwas
E
roman is not chance since sans serif characters
created by a mathematician for typesetting mathe
are not suited for physical mathematics b ecause sev
matics rst of all in his own b o oks
eral signs are not easy to distinguish in the absence
In this pap er I do not discuss how to typeset
of serifs compare I and l for example you cannot
A
mathematics since L T X takes care of most of it
E
tell which one is upp er case I and which one lower
very seldom the author needs minor corrections of a
case l
formula and when this happ ens it is usually to cor
Sans serif upright characters may be used in
rect some spacing when slanted op erators are to o
technical and or physical texts in order to mark
close or to o far away from the symbols they precede
ob jects that cannot b e confused with mathematical
or follow so that the slanting shap e of the op erator
symbols for example for the names of p oints in the
requires some degree of manual intervention in or
description of geometrical gures technical ob jects
der to x the spacing Several such cases are dealt
exp erimental setups and the like Therefore sans
A
with b oth in The T Xbook and in Lamp ort s L T X
E E
serif upright letters never app ear in a mathematical
Handb o ok
formula of a physicist or an engineer while math
Nor do I discuss the aesthetics of a typeset for
ematicians use sans serif fonts to represent certain
mula where several factors should follow one an
structures in category theory As a partial excep
other in such a way that the formula pro le is
tion sans serif sloping upp ercase letters are allowed
as smo oth as p ossible without valleys and p eaks
to indicate tensors of the second rank but this is the
Typesetting a complicated formula requires b oth
mathematical knowledge and a sense of aesthetics
but requires also esp ecially in didactic b o oks that
the relevant parts of the formula are highlighted in
1
Thanks to the reviewer for this information although
he she do es not sp ecify it nevertheless I supp ose that they
are sans serif sloping fonts
TUGb oat Volume No
only exception mentioned in the IPU recommenda index but R where E distinguishes an ob ject such
E
tions and stated in the ISO regulations clause as the emitter
for using sans serif fonts
Roman symbols
Italic symbols
Any other symbol that was not dealt with in the
According to the ISO regulations and the IPU rec preceding subsection must b e set in roman font the
ommendations italic symbols should be used only list of such roman entities is surprisingly long and
to denote those mathematical and physical entities unfortunately little known although ISO regulations
that may assume di erent values typically those and IPU recommendations are quite clear on this
symbols that play the role of physical variables but sub ject
also those physical constants that are not really
A
Numbers must b e set in roman type L T X
E
constant b ecause b etter measuring techniques may
do es this by default
pro duce up dated values
Numerical constants must b e set in roman type
Among such constants there is for example the
this is p erhaps the most neglected rule but it
elementary electric charge the charge of the proton
applies to e the base of nat
e C that is considered constantuntil
ural logarithms to the imaginary unit that
b etter measures will add other signi cant digits the
mathematicians and physicists call i while
same holds true for such constants as the velocityof
most engineers call j and so on The ISO
light c the Planck constants h and h the Boltzmann
regulations clause allow b oth sym
constant k and so on
b ols for the imaginary unit but b oth must b e
Every physical variable is represented by one
set in roman type The reason b ehind this is to
italic letter with as many mo di ers as needed such
avoid confusion b etween the base of natural log
as subscripts sup erscripts primes etc clause
arithms and the elementary electric charge b e
There are a few exceptions to this rule rep
tween the imaginary unit and the instantaneous
resented by the dimensionless parameters such as
current i or the instantaneous current density
the Mach number the Euler number and so on
j whose symbols are recommended by the IPU
that are sp eci ed with a two letter symbol by the
sections and
ISO regulations for example the Machnumber
Anybo dy can notice that e and i or j are
is represented by Ma the Euler number by Eu when
universally typeset in italic font in b oth mathe
suchtwo letter symbols are used equations must b e
matical and physical or technological texts this
written with sp ecial care so as to make sure that
observation gives a measure of how much the
Ma do es not represent the pro duct of the physical
rule I am sp eaking ab out is ignored at least by
variables M and a For the names of the nuclides
physicists and engineers
that may consist of two letters see the next section
The same rule should apply to numerical con
In pure mathematics two or three letter names
stants represented by Greek letters such as
are used in applications such as for example the
but it is necessary to get by
name of the Galois eld with n elements that is
with it b ecause it is di cult to have b oth the
represented with GF n But such applications are
upright and the sloping Greek fonts fortunately
not substantially di erent from what the ISO regu
enough the frequency of such symbols except
lations say ab out the names of sp ecial functions
is not high If b oth upright and sloping
clauses from to
Greek fonts were available an upright should
In the domain of Computer Science as well as
indicate the numerical constant
in Electronics authors and typesetters make frequent
while a sloping should indicate the physical
use of multiletter symbols but this tradition is ev
constant rad corresp onding to the
idently in contrast with the ISO and IPU statements
angle of
and should b e abandoned
Sub and sup erscripts must b e set in italics when
Physical units must b e set in roman type with
they representphysical quantities or mathematical
the additional requirement that a line break
variables clause otherwise they should b e
cannot take place b etween the measure and the
set in roman type for example C where T rep
T
unit of measure In order to emphasize the uni
resents temp erature M where i is a summation
i
tary nature of a physical constant the measure
and the unit of measure should be separated
2
Thanks again to the reviewer for p ointing out this topic
3
byanunbreakable thin space rather than bya
Please notice the di erence between the symbol of a
physical quantity and the name of an op erator or a function regular interword unbreakable space Besides
TUGb oat Volume No
b eing quite reasonable this p oint cop es with dards this particular one is almost completely
the high p enalty that T Xintro duces in a for neglected
E
mula within a pro duct of terms if T X has to
The ISO prescription and the IPU recommen
E
break a formula across lines it do es so close to
dation concerning the di erential op erator are
a binary or a relation op erator although very
not illogical since in physical and technological
reluctantly not within the pro duct of terms
mathematics it is essential to distinguish di er
and a physical quantity is the pro duct of the
ent mathematical ob jects within a formula tak
measure times the unit of measure
ing into account the nomenclature recommen
dations for physical quantities it is really neces
Mathematical op erators indicated with letters
A
sary to distinguish d from d when the latter
must be set in roman type L T X already
E
indicates one of the many physical quantities
do es this for a number of prede ned op erators
whose symbol is recommended to b e d thick
such as lim or sin The package amsmath
ness diameter relative density lattice plane
adds some more prede ned op erators but most
spacing degeneracy of vibrational mo de etc
imp ortant of all it adds a couple of commands
for using other op erators We will discuss this
Sub and sup erscripts that do not represent
topic in a following section
physical quantities or mathematical variables
It is worthwhile to mention that op erators should b e set in roman type the amsmath pack
do not require that their argument b e enclosed age makes available the command text for set
within parentheses if the op erand is made up of ting in roman type anyword or sentence within
no more then two ob jects b eing counted as mathematics also in sub and sup erscripts In
general though the problem is to distinguish one ob ject but prop er spacing is required on
the type of sub and sup erscripts in order to de b oth sides so you simply write sin t and do
cide how they should b e typeset In the b o oks not need to write sin t but you should write
I wrote I estimated that more than two thirds sin f t in any case parentheses are required
of the subscripts I used had to be set in ro when a formula contains several op erators with
man type while the absolute ma jority of su their arguments
p erscripts were mathematical expressions al
The names of sp ecial functions such as erf
though just a few b o oks have no statistical
error function or Ei exp onential integral
value the exp erience I gained allows me to say
or E incomplete elliptic integral of the sec
that this p ercentage of roman subscripts should
ond kind are treated by ISO XI the
b e considered typical so that roman subscripts
same as the names of op erators clauses from
are almost the default case in physics and tech
to and should be typeset in
nology
roman type although their argument s must
Chemical symbols coincide with the names of always b e indicated within parentheses See the
the nuclides if they carry prop er sub and su following sections for what concerns the de ni
p erscripts with resp ect to typography particles tion of new op erators
and quanta are treated as the nuclides They
A particular op erator the op erator of di eren
are made up of one or two letters and must b e
tiation should b e set in roman type but under
A
set in roman type section chemical equa
the L T X p oint of view it should b ehave dif
E
tions are in general quite complicated so that
ferently from other op erators as concerns spac
for setting into typeabookonchemistry it is
ing The use of roman type for the di eren
b etter to use sp ecialized packages the last one
tial op erator is another example of a highly ne
that was describ ed in TUGb oat is an ex
glected rule alb eit the ISO regulations clause
A
cellent example the L T X Companion de
are explicit on this p oint
E
scrib es another package ChemT X for the
The house style of the ma jority of publish
E
same purp ose
ing companies where the di erential op erator is
For simpler chemical formulas that do not re
a common italic d was evidently set up under
A
quire the graphic facilities of L T X the au
the in uence of the tradition of pure mathemat
E
thor can typeset chemical equations with the
ical typesetting b efore the ISO regulations were
usual math commands provided he she remem
published now manyyears have elapsed since
b ers to switch to roman type for the symbols of
their publication so that the ISO regulations
the chemical elements
should be widely applied and it is surprising
Roman numbers are seldom used in physics it surprises me at least that while the mo d
and technology although they nd their wayin ern world is so attentivetointernational stan
TUGb oat Volume No
chemistry where they may represent the sp ec The International System of Units SI is
trum of a z fold ionized atom or in sup erscript the only legal system of units to be used in
p osition the oxidation number in b oth cases those countries that undersigned the sp eci c
they are set in roman type clause bill In practice the SI is the only system of
units accepted by the scienti c community al
Roman numbers used for enumerated lists or
though engineers do it more readily than physi
for numbering front matter pages do not p ertain
cists the latter sometimes indulge in using one
to mathematics they are generally set accord
of the old cgs systems more for the p ossibil
ing to the publishing house style As a matter
ity of skipping some constants in the formulas
of p ersonal taste in these situations I prefer
dealing with electromagnetic phenomena than
small caps roman numerals to italic ones
for a real need of using centimetres centimeters
Other typesetting recommendations
and grams
The SI establishes the symbols for the fun
While typesetting physical mathematics it is con
damental and derived units and the legal deci
venient to remember that scienti c and technical pa
mal pre xes some non SI units keep their value
p ers haveaworldwide readership and the observance
while some others are outlaw In any case
of the ISO regulations and IPU recommendations is
every legal or accepted unit has its own symbol
essential for making one s text easily understo o d by
and only that symbol must b e used you write
readers of di erent language and culture Here are
h min s not hrs mins secs you write
some hints
a conductance of mS not a conductance
The decimal separator is the decimal comma
of m0
in the rest of the world clause and the
The symbols of the physical units are symbols
decimal p oint in the English sp eaking coun
not abbreviations therefore they must never
tries it is necessary to b e consistent with this
carry the abbreviation p oint after them and do
rule and to avoid mixed separators see the fol
not change in the plural clause so that
lowing p oint
it is correct to write cm while it is wrong to
Before and after the decimal separator digits
write either cm or cms with the abbre
may b e group ed in triplets separated only bya
viation p oint or with the plural of the symbol
thin space not by commas USA or lowered or
Unit symbols must always b e used when they
raised p oints Europ e clause in general
accompany the measure and they must b e set
Europ eans are more used to the North Amer
after the measure the full sp elled name with
ican habit of dividing triplets with commas
lower case initial even if the unit has an upp er
than North Americans with Europ ean habits
case symbol should b e used in the text when
In b oth cases a professional typesetter of a sci
preceded by general sp eci cations such as sev
enti c or technical text uses only thin spaces
eral few or when the unit of measure is
so as to avoid the trouble of interpretation to
called by name you write electric bulbs are
readers of di erent cultures Triplets may be
lab eled with the op erating voltage in volts and
avoided when b efore or after the decimal sep
the p ower rating in watts not electric bulbs
arator there are just four digits that is you
are lab eled with the op erating voltage in V and
should write USD or preferably to
the p ower rating in W
USD while writing USD should be
As for the plurals of the unit names not
absolutely avoided in Europ e it means one
of the unit symbols that are invariant as men
US dollar and cents
tioned ab ove every language has its own rules
In this pap er the North American decimal
and every country its own regulations b eware
separator is b eing used all over for Europ eans
4
not to use a plural form of another language
I recall that the p oint in the role of the deci
mal separator is allowed only in numerical con
Prop er scienti c prose do es not use the physical
stants written within a section of a program
quantity symbols in text mo de it rather uses
ming language
the full sp elled name p ossibly followed by the
symbol you write space s and time t are the
When anumber is less than unityin absolute
only physical quantities necessary to deal with
value this should b e explicitly indicated with a
zero preceding the decimal separator clause
4
This is not a common problem in English sp eaking coun
therefore it is necessary to write and
tries but it is a problem in other countries where English
plurals are often used in place of and resp ectively
TUGb oat Volume No
kinematics not s and t are the only physical Physicists seem to like the powers of en
gineers prefer the decimal pre xes established quantities necessary to deal with kinematics
by the SI Such pre xes representing p osi
Connected to the previous item the scientist
tive and negative powers of in addition
and the engineer should choose the correct
to several that represent the rst few p ositive
names for physical quantities the IPU recom
and negativepowers of cover an extremely
mendations o er a very wide nomenclature
wide range from to so that pow
list in English and French for virtually every
ers of can be easily avoided When choos
quantityofinterest in pure and applied physics
ing the right pre x it is necessary to remember
for every quantity a preferred literal symbol
that integer values of the measures are preferred
is indicated and the names of the quantities
and that zero es just after the decimal separator
should b e used in a consistentway rather than
should b e avoided you write pF rather than
inventing new names and new symbols for old
nF This rule holds true everywhere ex
ob jects simply b ecause of sp ecialized technical
cept in tabular formats where the unit of mea
jargons
sure is sp eci ed in every column head and is
The physicist s and the engineer s equations are
valid for all the table entries of that column
relationships between physical quantities the
section
latter are groups made up of a rst term that
Connected with the previous item there is
represents the measure and a second term that
the question of the number of signi cant dig
represents the unit of measure although they
its while leading zero es b etween the decimal
are not commutative they b ehave as factors of
separator and the rst nonzero digit are al
amultiplication The mathematical op erations
lowed only in tables trailing zero es should b e
p erformed on physical quantities op erate in the
avoided unless they are signi cant b ecause they
usual way on the numerical parts and in an al
carry information on the measure precision in
gebraic way on the units of measure giving rise
physics m is di erent from m b ecause
in some instances to derived units If a physical
the former quantity is precise to cm while
equation contains a physical constant not repre
the latter to mm Therefore physicists
sented by a symbol this constantmust contain
and engineers should b e careful to use just the
b oth elements measure and unit of measure
p p
number of signi cant digits inclusive of the
not Z You write Z
trailing zero es that are compatible with the
For the same reason it is not necessary or
exp erimental or technological data they are re
b etter it is de nitely wrong to write the units
ferring to
of measure after a coherentphysical equation
I need not emphasize that pre xes as well as
see also the next item
units of measure should be sp elled correctly
Measure equations should b e absolutely avoided
paying particular attention to upp ercase and
in professional scienti c texts measure equa
lowercase letters and to the op erators b etween
tions were somewhat p opular b efore the SI was
units nevertheless the following errors are quite
universally adopted now they should not be
common K kelvin in place of k kilo M
used any more They survived in those coun
mega in place of m milli or vice versa
tries where the English system of units is b e
m milli micro in place of n nano
ing used but since scienti cally sp eaking this
clause u in place of hz and db
traditional system of units is illegal mea
in place of Hz and dB K in place of K
sure equations have no reason to be used any
kwh KWH KWh kw h awful in place
more
of kWh or even b etter kW h notice the
thin space kc kilo cycles in place of kHz or
Since every physical quantity is a group formed
at least even if it is not fully SI kc s kilo cy
by the measure and the unit of measure it is
cles p er second
wrong to sp ecify the unit of measure after the
symbol of a physical quantity you write a p e
Decimal pre xes should be attached to the
rio dic function of p erio d T not a p erio dic
following unit without interposing any space
function of p erio d T seconds
comp ound units mayhave a thin space b etween
them or exceptionally a raised dot clause
5
The United States was one of the last countries to adopt
either separator is mandatory when units
the SI in the late fties but even after four decades have
may b e confused with pre xes p ositive or neg
elapsed in that country the SI seems to have di culties in
replacing the traditional units ative powers refer to the unit inclusive of its
TUGb oat Volume No
decimal pre x negative powers although al mands are de ned just once and you do not have
lowed are preferably avoided kW h is b etter to be concerned ab out anything else on the con
than kW h and b etter than kWh m=s is trary if you use newcommand you receive error
A
b etter than m s and b oth mean something messages when L T X executes them if the com
"
E
di erent from ms meters per square sec mand b eing de ned already exists and if you use
ond versus the inverse of square milliseconds renewcommand the error messages show up the
A
clause rst time L T X executes it I prefer the aster
"
E
isk form b ecause I get b etter diagnostic messages in
As mentioned ab ove the math italic typeface
case I forget some closing brace
is used for normal variables and roman type
face is used for letters and adjectives that do
Some commands for sp ecial units and a pre
not representvariables Other typefaces are sel
x are very handy the following de nitions are
dom used the IPU recommendations state
valid b oth in text and in math mo de
that sloping sans serif characters maybe used
DeclareMathAccent ring
for second rank tensors and b oldface italic char
mathalpha operators
acters for vectors lowercase and matrices up
providecommand angs
p ercase
ensuremath smash mathrm ring A
Currency units are not part of the SI and are
providecommand ohm
standardized by other regulations they are
ensuremath mathrm Omega
generally made up of three upp ercase letters
providecommand degree
the rst two of which are the two letter co de
ensuremath circ
for the nation while the third one is the ini
providecommand celsius
tial of the currency name therefore you write
ensuremath mathrm circ C
USD not US for a text that should b e read
providecommand micro
abroad and keep the symbol just for the
ensuremath mu
national readership after all many countries
The rst declaration adds the ring accent
use the dollar as the national currency b esides
A
in math mo de L T X has the r command
"
E
the United States there are Canada Australia
for the ring accent in text mo de but in math
Hong Kong and many others but in general
mo de the ring accent is missing so that it is
they are not equivalent to one another the same
necessary to de ne it At the present state
holds true for the UK p ound $ and the Italian
A
of the L T X art mathematical signs let
"
E
lira whose international symbols are GBP and
ters accents are taken from the traditional
ITL resp ectively
knuthian cm fonts and accents are set over the
In contrast to other units currency units pre
accented letter esp ecially a capital one a little
cede the monetary value but the rule of the un
to o high with the dc fonts accented letters are
breakable thin space keeps its validity Trailing
made up of just one glyph and the p ositioning
zero es such as USD may be required
of the accent has b een carefully adapted to each
in order to sp ecify also the unit submultiples
letter by the font designer Maybe in the future
for legal and or commercial purp oses
when new glyph math fonts are available
A
L T X new commands
the p osition of the accents will b e more precise
"
E
This is whyIintro duced the smash command
A
Here I will describ e some very simple L T X new
"
E
in the de nition of the angstr om unit other
commands intended to help in the typesetting of
wise in text mo de a line containing such a unit
physical mathematics In general I will use the
would force some interline leading resulting in
command providecommand This instruction b e
a nonuniform line spacing The chance of inter
haves just as newcommand except that it de nes
fering with the previous line descenders is small
the new command only if it is unde ned if the com
but not zero See the following paragraph
mand already existed the new de nition is silently
The ring command is useful not only for
ignored In this way if you make yourself several
setting the unit symbol A but also for setting
packages that include such de nitions the new com
6 7
The UK p ound and the Italian lira have similar symbols The angstr om unit is tolerated by the SI in the sp ecial
£ and \ attentive p eople notice that the UK p ound symbol ized eld of light and optics the name of the unit in contrast
has one stroke across while the Italian lira has two strokes to other SI units derived from p ersonal names containing di
but if you ask Italians the great ma jority don t notice the acritics maintains its diacritics annex A clause
di erence section IV
TUGb oat Volume No
avariety of physical variables in telecommuni where the typesetter should correct the almost
p erfect setting p erformed byT X cations when they refer to the analytical signal
E
The remaining de nitions introduce handy
A
L T X provides an internal command for set
"
E
commands for common units that may b e used
ting text sup erscripts but we need commands
in b oth mo des thanks to the ensuremath com
that let you set sub and sup erscripts with the
mand There is a drawback with these de ni
prop er math sizes the package amsmath pro
tions that is the newly de ned units are set in
vides the command text for setting text in
roman medium uprighttyp e everywhere and do
math mo de and this command chooses the
not change family or series in text mo de
right size according to the sub or sup erscript
The ohm de nition is a little redundant be
commands Here I introduce a couple of com
cause Omega by default is de ned as a symbol
mands that are good for single word sub and
A
not as a letter so that with L T X it is not
sup erscripts
"
E
sub ject to the change in the math version ei
providecommand ped
ther normal or b old but if you rede ne it or
ensuremath mathrm
use a package where it is rede ned as a letter
providecommand ap
from the mathnormal alphab et the ab ove de
ensuremath mathrm
nition guarantees that the unit is set in roman
These de nitions that work in b oth mo des can type anyway
b e handy also for setting text sup erscripts such
The following macro lets you set the units
nd rd
as The names are abbreviations of the
whichever they are in roman type b oth in text
Latin words pedex and apex that refer to the
and math mo des with the prop er unbreakable
fo ot or the head p osition resp ectively
thin spacing
A
The L T X package amsmath provides a cou
"
E
providecommand unit
ple of commands for using a sequence of letters
ensuremath mathrm
as an op erator they are operatorname and
When you use unit in text mo de of course
operatornamewithlimits here I introduce a
you must not leaveany space b etween the mea
couple of new commands that de ne new op er
sure and the unit command
ators
Numerical constants represented by Latin let
But b efore going on it is b etter to recall a
ters may b e treated by the following macros
A
couple of details ab out op erators L T X deals
E
The number e
with two kinds of op erators the log like and
providecommand eu
the lim like ones the former treat subscripts
A
ensuremath mathrm e
and sup erscripts as regular ones that is L T X
E
The imaginary unit
sets them in smaller size at the right and low
providecommand iu
ered or raised resp ectively while the latter ac
ensuremath mathrm j
cept sub and sup erscripts as limits so that in
math display mo de subscripts are set under
In the imaginary unit de nition a physicist
neath the op erator and sup erscripts ab ove it
would probably substitute j with i but this
A
Moreover L T X sets spaces at the left and
is not the p oint b ecause the ab ove de nition
E
at the right of the op erator in di erent ways
is just an example of how the imaginary unit
according to what kind of mathematical ob ject
should b e de ned A more sophisticated de ni
falls close to the op erator such di erent spac
tion might refer to math op erators see b elow
ings are describ ed in detail in chapter of The
so that prop er spacing is left around such let
T Xbook Therefore to set p erfect physical
ters esp ecially around the imaginary unit In
E
mathematics it is advisable to de ne op erators
several sciences though a common expression
in the prop er way but owing to the two kinds
is the combination of e raised to some imagi
of op erators it is necessary to sp ecify in the def
nary p ower if you use j for the imaginary unit
inition whether sub and sup erscripts should b e
its descender butts against the e irresp ective
used as limits or not
of what de nition you use and in practice this
o ccurs even if you stick to the tradition of us
providecommand newoperator
ing italic fonts this is one of the rare o ccasions
newcommand mathop
8
providecommand renewoperator
In any case even if ohm was simply let to Omega it
would b e shorter to sp ell and clearer to understand renewcommand mathop
TUGb oat Volume No
Here the rst argument is the op erator name With resp ect to total and partial deriva
the second is its full description and the third is tives it might be useful to de ne some com
the T X declaration limits or nolimits A mands with the order as an optional argument
E
couple of examples are necessary according to by means of the usual providecommand con
the ISO regulations the integer part of a decimal struct
number is called ent clause so
providecommand deriv
that the following de nition is required
frac diff diff
newoperator ent
providecommand pderiv
mathrm ent nolimits
frac partial
partial
According to the ISO regulations clauses
and the real part and the imag
The rst optional argument is the derivative
inary part of a complex number are Re and
order the second is the function b eing derived
Im not and therefore we need the fol
and the third the derivation variable Partial
lowing rede nitions
high order mixed derivatives require a more so
renewoperator Re
phisticated de nition or direct setting with the
mathrm Re nolimits
frac command
renewoperator Im
With the help of these new commands it b e
mathrm Im nolimits
comes very easy to typeset the following equa
tions with the assurance that spacings are right
The di erential op erator must b e treated in a
and the di erential op erator is correctly set in
slightly di erentway b ecause it is a sp ecial op
roman type a sample of source co de follows
erator that requires di erent spacings on the
each equation
left and on the right moreover b eing made up
of just one letter it is necessary to use a little
d y d y
b c f x a
T X trick in order to guarantee that its mathe
E
d x d x
matical axis lines up prop erly The di erential
a deriv y x
op erator should be spaced as an op erator on
d log y d y
the left while on its right it should receive dif
d x y d x
ferent spacings and in particular it should not
be spaced from its argument A p ossible way
deriv log y x
of achieving this result is to de ne it as an op
z z
d x d y d z x y
erator that contains a negative spacing on the
x y
right in order to cop e with p ossible exp onents
pderiv z x diff x
some tests must b e p erformed along the line and
Z
1
this requires some low level T X programming
st
E
e f t dt L f t
makeatletter
eu st f t diff t
providecommand diff
ifnextchar DIfF DIfF
The pre x femto that stands for
def DIfF
turns out to be rather frequent in micro elec
mathop mathrm mathstrut d
tronics where it is generally used as a pre x for
nolimits gobblespace
the unit farad the couple fF requires some
def gobblespace
kerning so that the lowercase f do es not butt
futurelet diffarg opspace
against the upp ercase F up to now the avail
def opspace
able math fonts do not consider this couple
let DiffSpace
for implicit kerning information therefore the
ifx diffarg
typesetter should remember to insert an italic
let DiffSpace relax
9
At the time of writing the latest version of dc fonts is
else
and has a date of June this release of the dc fonts
ifx diffarg
cop es with many new kernings that take place with SI decimal
let DiffSpace relax
pre xes but this do esn t help much in math mo de b ecause
here cm fonts are used their latest version was re ned in
else
but the latest available source METAFONT les on the
ifx diffarg
CTAN archives carry the date of June let s hop e that
let DiffSpace relax
the new math fonts with glyph enco ding will b e available
so on fi fi fi DiffSpace
TUGb oat Volume No
correction between the lowercase f and any Some text with physical quantities The heat
sink with a thermal resistance of C W upp ercase non slanting left side letter that fol
sa
is supp osed to maintain the temp erature be lows examples f F f F f W f W f H
low the transistor maximum junction op erat f H but fA fA
ing temp erature when it op erates in an envi
At present a femto macro would solve the
ronmentat C
kerning problem but in the future such kern
theta ped sa unit celsius W
ings should nd their wayinto the very de ni
unit celsius
tions of the various fonts that are used in math
mo de
Use of angstr oms awave length of A
unit angs
Examples
In addition to the examples shown in the previous
Conclusion
section I rep ort here some more examples where
I have b een using the ab ove commands for several
the various features describ ed ab ove show the dif
A
years b efore the advent of L T X I had put to
E
ference b etween mathematical and physical equa
A
gether similar commands for L T X but the dif
E
tions Between parentheses you nd the source co de
ference lies simply in the easier way of de ning them
for some relevant part of the example
A
with the new version of L T X With the help of such
E
The Euler equation involves the ve most im
commands I have found it very simple to cop e with
p ortant mathematical constants
the ISO regulations and IPU recommendations at the
j
e
p oint that now I feel somewhat handicapp ed if I have
to write anything scienti c on a computer where the
eu iu pi
A
L T X implementation is lacking such commands
E
Actually in this case as p ointed out in the
At the same time they are so few and simple
previous section the typesetter should intro
that I did not consider the p ossibility of making up a
duce a small correction to the spacing
short package le where the do cumentation and the
j
e
insertion driver would b e much larger than the useful
eu iu pi
lines to b e submitted to the CTAN archives Any
A comp onent list
body can copy such commands from pap er and or
use them as guidelines for making one s own set of
R k R k
B F
useful commands p ossibly p erforming b etter than
h k h
ie fe
the simple ones that I suggested
C nF R k
E C
But most imp ortant of all I warmly suggest to
R ped C unit k ohm
pay attention to the international regulations and
The equations of a common emitter stage with
recomendations the results are worth the little extra
a feedback resistor on the emitter
e ort
I h V V
b ie b R
E
References
h R I V
fe E b R
E
Mathematical sign and symbols for use in phys
V h R I
c fe C b
ical sciences and technology ISO ISO
A
whose L T X Notice in particular the term V
R
E
E
Standards Handbook N International Organi
source co de is V R ped E
nd
zation for Standardization Geneva ed
A table
Dimensionless parameters ISO ISO
Stub length Decay time
Standards Handbook N International Organi
in mm in ns
nd
zation for Standardization Geneva ed
SI units and recommendations for the use of
their multiples and of certain other units
ISO in ISO Standards Handbook N
International Organization for Standardization
nd
Geneva ed
An equation with complex variables
Le Syst eme International d Unit es Bureau In
Re F j 8 and 8
ternational des Pois et Mesures Pavillon de Bre
th
renewoperator Re mathrm Re nolimits teuil S evres France ed
TUGb oat Volume No
Co des for the representation of currencies and
funds ISO International Organization
for Standardization Geneva
Symbols units and nomenclature in physics
in CRC Handbook of chemistry and physics
R C Weast M J Astle and W H Beyer edi
th
tors CRC Press Inc Bo ca Raton Florida
edition pp F F
Knuth D E The T Xbook Addison Wesley
E
Publishing Company Reading Mass
Lamp ort L A document preparation system
A
L T X User guide reference manual Addison
E
Wesley Publishing Company Reading Mass
second edition
Go ossens M Mittelbach F and Samarin A The
A
L T X Companion Addison Wesley Publishing
E
Company Reading Mass
Fujita S X MT X for drawing chemical
E
structural formulas in TUGb oat vol n
pp March
Haas R T and O Kane K C Typesetting
Chemical Structure Formulas with the Text For
A
matter T X L T X in Computers and Chem
E E
istry vol n pp
Claudio Beccari
Dipartimento di Elettronica
Politecnico di Torino
Turin Italy beccari polito it
TUGb oat Volume No
A
L T X
E
A
ALT XTour part
E
mfnfss psnfss and bab el
David Carlisle
Introduction
This third installmentofmy tour covers three more
distributions that are supp orted via the standard
A
L T X bug rep ort mechanism describ ed in Part
E
A
The mfnfss distribution provides L T X supp ort
E
for some p opular Metafont pro duced fonts that do
A
not otherwise haveanyLT Xinterface
E
A
The psnfss distribution consists of L T X pack
E
ages giving access to PostScript fonts
The third distribution in this part of the tour
A
is bab el which provides L T X with multi lingual
E
capabilities
The Mfnfss Distribution
The mfnfss distribution is something of a collecting
p oint for les in the distribution that have not got
anywhere else to go
FontPackages
A
These packages provide L T X interfaces to some
E
publicly available fonts They do not provide the
fonts themselves which are available from the fonts
tree in the standard CTAN archives
pandora The Pandora family of fonts designed by
Nazneen N Billawala is an alternative to the
standard Computer Mo dern fonts of Knuth
The family consists of a full range of text fonts
including sans serif and slanted
oldgerm The old German fonts designed byYannis
Haralambous There are three styles of text
font Schwabacher Fraktur and Gothic The
terms Fraktur and Gothic tend to b e used in
terchangeably by English sp eaking mathemati
cians such as the present author but the fonts
in this collection have clearly distinguishable
styles
There is also a font of initials highly ornate
upp ercase letters suitable for use as the rst
letter of a section If you wish to use this
in drop caps style you may also want to use
one of the contributed packages available on
CTAN suchasdrop ordropping that automate
TUGb oat Volume No
the setting of a suitable paragraph shap e and Scheme are installed then one could in principle
A
inserting the initial letter at the correct size switch to Times Roman in a L T X do cument by
E
simply sp ecifying fontfamily ptm selectfont
T Enco ded Concrete Fonts
Normally one would instead want to assign the new
A
Note The following two les require the old release font to one of the default L T X families Roman
E
of the dc fonts Hop efully there will so on b e an as used by rmfamily Sans Serif sffamily and
o cially supp orted release of T enco ded concrete Typewriter or Monospace ttfamily
fonts based on the recently released ec fonts At that The supp ort for PostScript fonts is split into
time these les will probably b e withdrawn from this two The CTAN fonts psfonts area contains ma
mfnfss distribution terial that is mainly automatically generated from
the Adob e font metric les that are distributed with
dccr mf Metafont source le used by the output
all Type fonts This includes the font metrics
les from dccrstd tex to generate Concrete
themselves the Font Descriptor les the map les
Roman fonts in T enco ding
used to make fonts known to the dvips driver and
dccrstd tex T X le used in the generation of
E
A
some basic packages to declare single fonts to L T X
E
Concrete Roman fonts in T enco ding It will
This is supplemented in macros latex packages
pro duce a number of mf les corresp onding to
psnfss by the hand written packages of the psnfss
Concrete Roman fonts in di erent sizes By
collection that load p opular combinations of font
mo difying the table inside this le further Meta
families or deal with mathematics
font driver les can b e generated The fd les
This section refers at various p oints to Post
for the Concrete Roman fonts can b e pro duced
Script or Type fonts but in fact the T X supp ort
A
E
with cmextra ins which is part of the L T X
E
for these fonts applies equally well to True Type or
base distribution
other scalable formats As long as T X has access
E
The Psnfss Distribution
to the font metrics the font format do es not matter
to T X it matters to the driver you use to print
A A
E
With the release of L T X L T X gained inbuilt
"
E E
the DVI le
supp ort for the use of alternative font families in
do cuments and in particular for the use of scalable
Psfonts
font formats such as Type PostScript or True
The CTAN psfonts area primarily contains the font
Type
A
metric and L T X font descriptor les organised by
E
The collection of packages co ordinated by Se
fontvendor as outlined b elow The basic format of
bastian Rahtz known as psnfss o ers convenient
the le structure is the same for each font family
interfaces to most of the more common font sets
so only the top level directories are given here
Most of the les here relate to font les re
except for the Adob e Times family which is further
named to a consistent naming scheme promoted
expanded as an example
and maintained by Karl Berry This enco des the
font vendor and details of the font such as its
FontVendors
weight style and enco ding into a compact name
The font sub directories of fonts psfonts are
that usually ts in the eight letter lenames used by
some common lesystems More information ab out
adobe Fonts sold by Adob e or built into PostScript
the font naming scheme can b e found on CTAN in
devices
info fontname It should be noted however that
bh Fonts designed by Bigelow and Holmes these are
the packages themselves such as the times package
mainly sold through Y Y
do not dep end on any particular font naming con
bitstrea Bitstream fonts
A
vention L T X isolates packages from the details
E
monotype Monotype fonts
of the external font les by the use of fd Font
textures Textures Fonts for the Blue Sky Research
A
Descriptor les which map the L T X NFSS mo del
E
Macintosh T X implementation
E
of fonts to the external font metric les
urw Fonts distributed byURW
In principle there is no real need for packages
xadobe Adob e exp ert font sets
A
to load text fonts into L T X For example once
E
xmonotype Monotype exp ert font sets
the font metrics and font descriptor les for Times
Each of the vendor directories contains sub
Roman which is ptm in the Karl Berry Naming
directories corresp onding to the font families sup
1
The fd les provided here load the original yinit font
p orted by the psfonts distribution Using the to ols
The CTAN archives also contain yinitas a mo di ed version
of this font provided one can generate T X supp ort les for most E
TUGb oat Volume No
other text fonts the selection here is really just a set tex This directory contains the font descriptor les
of examples which must be placed in the input path for
A A
The sub directories of the adobe directory are L T X so that L T X has available the infor
E E
mation ab out the available fonts For some
agaramon Adob e s rendition of a Garamond serif
font families this directory would also contain
Roman family Commercial
A
a L T X package that assigns the fonts to one
E
avantgar Avant Garde sans serif built into most
A
of the standard L T X font families such as
E
PostScript devices
sffamily Some packages suchastimes are
baskervi Baskerville a commercially available ser
not distributed here as they would clash with
ifed Roman family
the packages distributed as part of psnfss as
bembo Bembo a commercially available serifed Ro
describ ed b elow
man family
tfm The font metrics in tfm format These les
bookman Bo okman built into most PostScript de
contain all the information ab out letter sizes
vices
ligatures and kerning that T X needs to type
E
centaur Centaur a commerciallyavailable serifed
set text
Roman family
There are several les as each font in the
courier Courier built into all PostScript devices
original family is made available in several en
garamond Garamond Another Garamond serif
co dings the two main ones b eing the Classic
Roman family Commercial
T X enco ding used by Computer Mo dern This
E
gillsans Gill Sans a commercially available sans
A
is known as OT in L T X and as t in the
E
serif family
Karl Berry font naming scheme used here Sim
helvetic Helvetica built into all PostScript de
ilarly the les with names ending in t relate to
vices
fonts enco ded to the eight bit Cork enco ding
nbaskerv ITC New Baskerville another varianton
A
known as T in L T X
E
the Baskerville theme Commercial
vf The virtual fonts Most but not all drivers
ncntrsbk New Century Schoolb o ok built into
handle the re enco ding of the original fonts to
most PostScript devices
the enco dings that T X exp ects by means of
E
optima Optima a commercially available sans serif
the virtual font mechanism Some sp ecial fonts
family
such as Zapf Dingbats are not re enco ded and
palatino Palatino serifed Roman family built into
so do not haveavf directory
most PostScript devices
There is one very imp ortant thing to note ab out
symbol Symbol built into all PostScript devices
the ab ove list There are no fonts Almost all of
times Times Roman built into all PostScript de
the fonts psfonts area of CTAN is concerned with
vices
providing mechanisms for using fonts that you have
univers Univers a commercially available sans
obtained elsewhere The fonts may be built in to
serif family
your printer or may b e purchased separately There
utopia Utopia a commercially available serifed Ro
are a few freely available Type fonts In such cases
man family
there will b e an additional directory type which
zapfchan ITC Zapf Chancery A script font built
contains the font les normally in pfb format
into most PostScript devices
Standard PostScript Fonts
zapfding ITC Zapf Dingbats A symbol font built
into most PostScript devices
In addition to the ab ove directories the psfonts area
contains two zip les If you need the les and have
All the directories corresp onding to a font fam
not got unzip or pkunzip or winzip or then you
ily lo ok essentially the same each with the following
can get a copy of unzip from the CTAN supp ort area
sub directories
dvips Contains the map le for the dvips driver
lw nfss This zip archive expands to the subset of
program This le can be app ended to
the psfonts adobe tree that corresp onds to the
psfonts map or used via a con guration le
Standard PostScript fonts as used in Adob e
to tell dvips where to nd the sp eci ed fonts
Laserwriter printers If you are only interested
A suitable con guration le is included in the
in using fonts built into your printer and not
directory
in using downloaded fonts then just get this
Other drivers will need similar information le rather than the large collection of metrics
but p erhaps in a di erent format in psfonts adobe
TUGb oat Volume No
lw pk This zip archive contains bitmap fonts for horrible if placed next to each other at the same
the Standard PostScript fonts in the usual PK nominal size as done by this package Helvetica
format understo o d by most dvi drivers This hasamuch larger x height the height of the
enables do cuments using Type fonts to be lower case letters than Times Roman so if sans
previewed with dvi previewers that can not use serif and Roman text are mixed in line then
outline font formats For example xdvi or the the sans serif lo oks much to o big This is not
emtex drivers so much of a problem if the sans serif is only
used for headings Courier is just to o wide
Tools and Extra Packages
when placed alongside Times Roman whichis
There are a few remaining directories in psfonts a particularly compact font
To partially comp ensate for these problems
A
ts The L T X textcomp package and related utili
E
the pslatex package written by me but cur
ties for accessing fonts in the text companion
rently distributed as a contributed package not
A
enco ding known as TS in L T X These include
E
A
part of the core L T X distribution is an alter
E
the TC fonts that are distributed with the EC
nativetothe times package It loads Helvetica
fonts and suitably re enco ded fonts from the
scaled by and loads Courier by way of a
standard Type font sets This enco ding con
virtual font that condenses it by scaling the
tains many non alphab etic symbols that should
horizontal direction only by pslatex also
match the current text font rather than the
contains a copy of the mathptm package see
math font It includes currency symbols su
b elow so installs a Times Italic based font set
p erior digits dagger signs etc
for use in mathematics
mathcomp A contributed package for using the text
palatino Declares Palatino as rmfamily and
companion fonts in math mo de
Helvetica and Courier as sffamily and
tools The source for the scripts and utilities used
ttfamily
for generating all these les
helvet Declares Helvetica as sffamily Do es not
change the other families
Standard Psnfss Packages
avant Declares Avant Garde as sffamily Do es
By contrast to the packages and font descriptor les
not change the other families
in the psfonts distribution the psnfss distribution
newcent Declares New Century Schoolb o ok as
contains hand written les These are either used
rmfamily Avant Garde as sffamily and
to set up p opular combinations of the standard
Courier as ttfamily
fonts or load alternative font sets for mathematics
bookman Declares Bo okman Roman as rmfamily
Due to the nature of mathematics fonts these latter
Avant Garde as sffamily and Courier as
packages are typically much more complicated in
ttfamily
ternally than the one or two line packages that load
chancery Declares Zapf Chancery as rmfamily
text fonts For the user however this complexity
should not b e apparent
The ab ove packages only a ect text fonts not
The rst set of packages all generated from the
mathematics psfonts dtx contains one sp ecial
source le psfonts dtx load combinations of the
package written by Alan Je rey which do es a ect
A
Basic Adob e PostScript font set into L T X
the math setup
E
times As one might guess this declares Times
mathptm This package uses a set of virtual les that
Roman as rmfamily For mainly historical
use various built in or freely available fonts to
reasons this package also declares Helvetica
make a set of fonts suitable for replacing the
as sffamily and Courier as ttfamily so
standard Computer Mo dern Math fonts In the
A
e ectively ensuring that all text but not math
current release b old fonts and so the L T X
E
ematics is set in the basic PostScript font set
boldmath command are not supp orted The
pslatex package referred to ab ove contains an
This is a convenience for the user who wants
essentially verbatim copyof mathptm
to replace all the text fonts by references to the
basic Adob e fonts It is an advantage to do One may use mathptm as an example of the
this if you want to pro duce device indep endent co ding needed to make virtual fonts for mathe
and small PostScript do cuments for distribu matics based on other text italic fonts How
tion The disadvantage is that Times Roman successful this will be dep ends to a certain
Helvetica and Courier despite b eing the stan extent how visually compatible are the sym
dard PostScript combination lo ok particularly b ols that are gathered from the various real
TUGb oat Volume No
fonts that are used by the virtual math fonts lucid Declares Lucida Roman and Lucida Sans as
the Roman and sans serif families and Adob e There are often good reasons for making such
Courier again as the monospaced font fonts the main one b eing that do cuments using
freely available fonts may b e more easily placed
lucmath Lucida has a matching set of mathematics
on the Web in PostScript form however the
fonts suitable for T X use This package makes
E
result is never likely to be as good as using
the required de nitions to make these known to
A
fonts that have symbols that are designed to
L T X
E
be visually compatible For mathematics use
Lucida Bright
within T X that currently restricts use to Com
E
puter Mo dern or the commercial MathTime or
A newer and more extensive Lucida family also
Lucida Bright font sets describ ed b elow
designed by Bigelow and Holmes but in this case sold
by Y Y is known as Lucida Bright and Lucida
The psfonts dtx source le contains one other
A
New Math The L T X supp ort describ ed here was
package
E
written by Sebastian Rahtz and myself
pifont This declares the Zapf Dingbats font which
lucidabr dtx This package replacing the earlier
contains an assorted mixture of symbols and
lucidbrb and lucidbry packages changes the
also de nes new user level commands to access
A
L T X defaults for b oth text and mathematics
these symbols See the package do cumentation
E
A
to use the Lucida Bright and Lucida New Math
or The L T X Companion for details
E
font collections It has numerous options to
Freely Available Type Text Fonts
control di erent asp ects of the package and
to control which of the fonts to use Lucida
The next set of packages are contributed by Peter
Bright contains several font families including
Dyballa In fact these are just one line packages
fax and casual etc as well as variant forms
loading the appropriate font Most of the co de is
of the math italic alphab et
in the fd les which are generated from the same
A
The L T X package and the font descriptor
source le
E
les for the math fonts are generated from
charter De nes rmfamily to use Bitstream Char
this source le The font descriptor les for
ter
A
the Lucida text fonts in the standard L T X
E
nimbus Declares URW Nimbus Roman Regular and
enco dings are available from the psfonts area
URW Nimbus Sans Regular as rmfamily and
in the bh directory after Bigelow and Holmes
sffamily These are essentially free clones of
the creators of these fonts
Times Roman and Helvetica
The T X supp ort and font metrics are freely
E
utopia De nes rmfamily to use Adob e Utopia
available but the fonts themselves must be
Regular
purchased separately
A
lucidabr ins L T X installation le for Lucida
Commercial Text Fonts
E
Bright using the standardised Karl Berry
The following packages are generated from the
font names
source le adobe dtx They are a rather random
lucidabr yy Alternative installation le Use this
selection from the large catalogue of fonts sold by
instead of lucidabr ins if you plan to install
Adob e
the fonts with their original font names as sold
garamond Garamond as rmfamily Optima as
by Y Y In this case you do not need the fd
sffamily and Courier as ttfamily
les from the psfonts area
basker Baskerville as rmfamily
lucidabr txt Introduction and installation guide
mtimes Monotype Times as rmfamily
for this package
bembo Bembo as rmfamily Optima as sffamily
MathTime
and the ever p opular Courier as ttfamily
The MathTime fonts are pro duced by Michael Spi
Adob e Lucida
vak T Xplorators They are sold by Y Y The
E
A
The following two packages relate to the original
L T X supp ort was written byFrank Mittelbach and
E
Lucida font set designed by Bigelow and Holmes
myself
and sold by Adob e They are generated from the
mathtime dtx The mathtime package is mainly con
alucida dtx source le
cerned with mathematics setup although it
2
Not sure why this is generated from adobe source le selects Times Helvetica and Courier as the
TUGb oat Volume No
text fonts if they have not already b een set by The Bab el Distribution
another package The MathTime mathematics
The bab el package is distributed from latex
fonts are sp ecially designed to match Times
A
packages babel and is supp orted via the L T X
E
Roman but blend quite well with other text
bug rep orting address but has origins predating
fonts that are of a similar weight Computer
A
the current L T X release As well as supp orting
E
Mo dern mathematics tends to lo ok very light
A
L T X it contains supp ort for plain T X and
E E
if used with font families other than Computer
formats such as AMST X or eplain that are based
E
Mo dern The package has several options to
on plain Primarily bab el is the work of Johannes
control the fontchoices made
Braams with contributions for sp eci c language
mtfonts fdd The source for the font descriptor les
les bynumerous p eople
for MathTime mathematics fonts
A
Bab el consists of a kernel that extends L T X
E
mathtime ins Installation le Note that this le
with a mechanism for switching between sp eci ed
may b e edited in a couple of places dep ending
languages Part of this kernel related to hyphen
A
on whether or not you have the extended Math
ation must be loaded when the L T X format is
E
Time Plus font set which includes b old math
made to get the full b ene t of hyphenation tables
supp ort
for multiple languages For each language or re
mathtime txt Introduction and installation guide
lated group of languages supp orted by bab el there
for this package
exists a language sp eci c co de le This will o er
translations of the xed text strings used in the
Do cumentation and Other Files
A
standard L T X classes suchas Table of Contents
E
readme txt General introduction
Figure etc and may also o er language sp eci c
shorthands that make typing common constructs
psnfss e tex User level do cumentation on the use
easier for example the german option provides the
of these packages
construct ff to pro duce that would hyphenate
test tex Testing accents and other enco ding sp e
to f if it fell at the end of a line The language
ci c commands are working correctly using
le may also mo dify the typesetting to supp ort the
PostScript fonts
normal conventions of that language For example
test tex Test do cument that uses most of the
the french option mo di es the spacing around punc
Standard fonts
tuation marks in text
pitest tex Test of the pifont package
mathtest tex Test of the mathptm package
Bab el Kernel
makefile Unix make utility to automate installa
babel sty The main interface to bab el The user
tion of the packages
sp eci es all languages to b e used in a do cument
allpspk Unix script that makes a test do cument
as options to this package the last option sp ec
using a sp eci ed font family and then uses
i ed is the default language for the do cument
dvips and its asso ciated scripts to generate pk
So for example
versions of the fonts
usepackage french german babel
makepk Unix script that calls allpspk on some com
mon fonts
would enable the use of French and German
conventions within the do cument with the de
Psnfssx
fault language b eing German
Recently the psnfss collection has aquired a close
A
cousin psnfssx distributed as a contributed package hyphen cfg The standard L T X interface to hy
E
A
from macros latex contrib supported psnfssx phenation When the L T X format is b eing
E
This contains some lesser used or nonstandard pack made this le is input if it exists to setup
ages related to PostScript supp ort Of particular the required hyphenation patterns In the base
A
interest mightbe the ly les contributed by my L T X distribution there is no such le and
E
A
self in that directory which provide the L T X sup so a default action is taken which loads the
E
p ort for the texnansi enco ding promoted by Y Y original T X patterns for American English
E
bywayofan LY option to the fontinst package The bab el distribution provides this con gu
This psnfssx collection also contains some ob ration le generated from babel dtx which
solete versions of packages formerly in psnfss this de nes some core functionality and then reads
material is provided for historical interest only Use language dat to sp ecify which hyphenation
at own risk les to load
TUGb oat Volume No
language dat This le must be edited to sp ecify english The american USenglish and british
UKenglish options The option english which language hyphenation les to load and
refers to either British or American English the name of the external le which contains
dep ending on the lo cal installation the hyphenation table for each such language
and optionally a second external le typically
esperant The esp eranto option
containing hypenation exceptions Note that
estonian Supp ort for the Estonian language
hyphenation les must be sp eci ed here and
finnish Supp ort for the Finnish language
so loaded when the format is made This is
frenchb Supp ort for the French language the
a restriction of the underlying T X system
E
corresp onding options are french frenchb
Do cuments using other languages not sp eci ed
or francais If the french option is used then
here may still be pro cessed and bab el will
french ldf will b e used from the GUTenburg
translate any xed text strings but it will not
french package if it is available
be able to correctly hyphenate that language
galician Supp ort for the Galician language
A default hyphenation will b e used most likely
germanb The austrian and german germanb op
English which may or may not be suitable
tions
dep ending how far the language di ers from
English
kannada Supp ort for the Indian language Kan
nada
switch def This le is also generated from the
irish Supp ort for the Irish Gaelic language
same babel dtx source If bab el is used as
a package but was not used when the format
italian Supp ort for the Italian language
was made then the core functionality normally
lsorbian The lowersorbian option
provided by hyphen cfg will not be present
magyar The magyar hungarian options
The package will detect this and so input this
norsk Supp ort for the Norwegian languages with
le to provide the necessary de nitions
options norsk nynorsk
polish Supp ort for the Polish language
Language Sp eci c Files
portuges The brazil brazilian and portuges
The implementation of the language sp eci c co de
portuguese options
for each language within bab el is contained in les
romanian Supp ort for the Romanian language
with extension ldf language de nition les
These are not directly input by the user but sp ec
sanskrit Supp ort for the Sanskrit language
i ed as options to the bab el package Normally
transliterated to latin script
the option name is the same as the le name ex
scottish Supp ort for the Scottish Gaelic language
cept where noted b elow Some similar languages
slovak Supp ort for the Slovakian language
or dialects are supp orted by the same external le
slovene Supp ort for the Slovenian language
and some options are available in more than one
spanish Supp ort for the Spanish language
name such aliases are noted in parentheses in the
swedish Supp ort for the Swedish language
list b elow
turkish Supp ort for the Turkish language
Most languages also have a le with extension
sty however this is just o ered for compatibility
usorbian The upp ersorbian option
A
with older versions of Bab el and of L T X or for
E
welsh Supp ort for the Welsh language
A
use with plain T X based formats In normal L T X
E E
Bab el version sees the welcome re intro
usage only the ldf le is used
duction of supp ort for non latin scripts It is prob
bahasa Supp ort for the Bahasa language
ably fair to say that this supp ort is still more ex
p erimental than the supp ort for latin scripts One
basque Supp ort for the Basque language
problem not directly under bab el control is that
breton Supp ort for the Breton language
the T X enco dings for Greek and Cyrillic corre
E
catalan Supp ort for the Catalan language
sp onding to T for Europ ean Latin scripts have not
croatian Supp ort for the Croatian language
yet b een nalised or agreed Currently bab el uses
two lo cally de ned enco dings LWN and LGR
czech Supp ort for the Czech language
greek The greek option which utilises the kd
danish Supp ort for the Danish language
Greek fonts
dutch The dutch and afrikaans options
russianb The russian option which utilises the
3
Not in the current release planned for bab el LH fonts
TUGb oat Volume No
Two separate packages are currently in preparation T X Do cuments
E
which will be distributed together with suitable
tb The source of the original article that ap
fonts and hypenation tables from CTAN These
p eared in TUGb oat Volume No
will extend bab el with options for the Ethiopian and
tb The source of an up date article that ap
Ukrainian languages
p eared in in TUGb oat Volume
No
Compatibility Files
tb The source of an up date article that never
The distribution contains the following two source
app eared in TUGb oat but was presented at
les which generate les which enable the use of
EuroT X Arnhem
E
bab el with formats based on plain T X and also
E
A
the old L T X release
E
Example File
bbcompat The source for compatibilitymode les
language skeleton An example le that can be
Most languages are provided with a package
used to build new language de nition les from
with extension sty This just inputs the cor
scratch
resp onding language de nition le and should
A
never b e needed using the normal L T Xinter
Coming So on
E
face
Part of this tour will describ e the les of the
bbplain The source for the plain def le allowing
amsfonts and amslatex distributions of packages pro
the use of bab el with plain T X
E
duced by the American Mathematical so ciety
Installation Script and Font Descriptor
David Carlisle
Files
Department of Computer Science
Manchester University
babel ins Unpacks the bab el distribution from the
Oxford Road
do cumented source les
Manchester England M PL UK
cyrillic fdd Font descriptor les for Cyrillic fonts
carlisle cs man ac uk
in LCY enco ding
greek fdd Font descriptor les for Greek fonts in
LGR enco ding
Do cumentation
ASCI I Text Files
readme txt The distribution guide
install txt How to install Bab el
install mac How to install Bab el with OZT X
E
CyrillicFonts txt Further notes on the Cyrillic
installation
GreekFonts txt Further notes on the Greek instal lation
TUGb oat Volume No
Late Breaking News
Pro duction Notes
Mimi Burbank
Once again we re late That is the bad news The
go o d news is that the next issue of TUGb oat should
be reaching your desk in ab out a month with a
present included therein
We include in this issue another article from
the TUG Conference by Yannis Haralambous
This article requires sp ecial typesetting capabilities
unavailable to the pro duction team at this time and
so we asked the author to typeset it for us
Output The nal camera copy was prepared at
SCRI on several of the lo cal platforms to test the new
TeX Live setup see b elow Alphastation
running OSF v IBM RS workstation run
ning aix IBM SP RS no de running aix
Silicon Graphics SGI workstation running
Irix and a Dual Pentium Pro running Linux
i We used the TeX Live setup Version which
is based on Karl Berry s Web c T X implementation
E
version we also worked with Thomas Esser s
teT X package loaded from the TeX Live Version
E
CD ROM PostScript output at dpi was pro
duced using Radical Eye Software s dvipsk a and
printed on a QMS printer
I am so pleased to announce that TeX Live
Version is alive and doing well
Coming Next Issue
Well wehave a gift for you in the next issue The
TeX Live Version CD ROM is b eing pressed as
this issue go es to press and will b e included in the
next issue of TUGb oat along with the do cumen
tation The CD ROM contains a very large TDS
compatible tree of macros fonts and do cumentation
as well as a runnable T X system for a dozen vari
E
eties of Unix also included is a system for win
Windows or NT and packages for Windows
DOS and Macintosh ready to install
We also plan to add some material whichwas
held over from the last issue of T X and TUG News
E
a series of abstracts from Die T Xnische
E
Kom odie and Les Cahiers GUTenberg aswell as a
manual byB Jackowski on METAFONT practical
and impractical applications
aeaRayCp edieJn 0 1997 20, June Deadline Copy Ready Camera
rpitDaln a 6 1997 16, May Deadline Preprint
rlmnr aesDeMy2 1997 2, May Due Papers Preliminary
oicto facpac oatosMrh2,1997 21, March authors to acceptance of Notification
umsino btat ac 4 1997 14, March Abstracts of Submission
edie Revised — Deadlines
omte,15Cne ilRa,Pyot,M 26 USA. 02360 MA Plymouth, Road, Hill Center 135 Committee,
[email protected] rb otlmi oTG9 Bursary TUG’97 to mail postal by or to email send requests, bursary For need.
iat h demonstrate who cipants parti those and students of support for funds bursary arrange to hope We
email.
[email protected] o nomto via information for at Committee Conference the contact to welcome is browser web
ae n ogn nomto.Ayn akn a lacking Anyone information. lodging and ravel t for links with together posted, been has schedule liminary
http://www.tug.org/tug97 n oeta pre- a that note and , at page home Page Home ’97 our visit Please TUG
egvnott l ofrneattendees. conference all to out given be
saalbet aecneec tede h r tyn on staying are who attendees conference take to available is apst h etr optracut will accounts Computer center. the to campus
h etri eywl qipd ihsvrllcuehalls lecture several with equipped, well very is center The n iigfclte.Sutetransportation Shuttle facilities. dining and
on atteUiest n otebeach. the to and University the past town,
ca ec.Asnl u iltk asnesfo h ferry the from passengers take will bus single A beach. ocean emnlbidn nteEbraeothrough Embarcadero the on building terminal
h aiiyi bu 52 iuebsrd rmtecne o center the from ride bus minute 15-20 a about is facility The a rnic,btentw n the and town between Francisco, San f
us osn ncmu saalbet ofrneattendees. conference to available is campus on housing guest
rnic a n h rmtcdwtw kln,adi hr walk short a is and skyline, downtown dramatic the and Bay Francisco rmGle aePr.Economical Park. Gate Golden from
onanCneec etr oae na5-cehltpre hilltop 55-acre a on located Center, Conference Mountain uewt iw ftePcfi ca,San Ocean, Pacific the of views with fuge
n xiigcte ntewrd h ofrnewl ehl tTh at held be will conference The world. the in cities exciting and nvriyo a rnic Lone Francisco San of University e
E ewudlk oetn nivtto oT to invitation an extend to like would We sr rudtewrdt onu noeo h otbeautiful most the of one in us join to world the around users X
a rnic,California Francisco, San
oeMuti ofrneCenter Conference Mountain Lone uy2t hog uut1st August through 28th July
XComesHome XComesHome XComesHome
E
E E XComesHome XComesHome
T
E
T T E
T T XComesHome XComesHome
E E T T
XComesHome XComesHome
E E T T
E XComesHome XComesHome T TE
TUG’97
E
XComesHome T TEXComesHome
E T T
XComesHome EXComesHome
T T
E E XComesHome XComesHome
T T
E E
XComesHome XComesHome T T
E
T T E
T
XComesHome XComesHome
E E
E XComesHome XComesHome XComesHome
TUGb oat Volume No preliminary draft Apr
Personal statement
New members of the TUG Board
As a b oard member I would be interested in
As announced earlier in this issue the number
actively involving members of our T X community
E
of candidates for op en p ositions on the Board of
in extending the installation standardization pro cess
Directors was fewer than the number of p ositions
to include the VMS op erating system
hence according to the Election Pro cedures the
candidates who did submit nomination pap ers were
Mimi Jett
declared elected Since there will therefore be no
ETP Harrison
ballot the biographies and p ersonal statements
N E Glisan
that would have app eared on the ballot are
Portland OR
presented here to introduce these individuals to
U S A
the membership
Internet mimi etp com
The new Board members are listed in
Biography
alphab etical order Their terms of o ce all extend
President and CEO of ETP Harrison a
to the annual meeting in the year
full service electronic publishing rm lo cated in
Barbara Beeton
Portland Oregon We b egan the company as a
For the Elections Committee
provider of tro typesetting services but moved
into T X in at the request of textb o ok
E
Donna Burnette
publishers With the help and supp ort of the
SCRI
T X Users Group we found the resources necessary
E
Florida State University
to b ecome a serious provider of comp osition
Tallahassee FL
and related services Prior to starting this
U S A
business I b egan several other businesses including
Internet donna scri fsu edu
Oregon Serigraphics The Renaissance Press
The Resource and Marketing
Biography
Communications all related to print or
My job over the past years at the
advertising ETP Harrison employs p eople and
Sup ercomputer Computations Research Institute
freelancers full time I have served on numerous
SCRI has evolved from T Xnical typing to T X
E E
b oards including the TUG b oard in
software installation and supp ort to full supp ort
of VMS systems and their software We supp ort
Personal statement
a large community of research scientists university
TUG has b een a vital contributor to our success
faculty and administrative sta I have written
and I am eternally grateful The b ene ts of
and deciphered many T X programs during this
membership have b een questioned as long as I
E
p erio d and b ecome quite knowledgeable in the
can remember and I would like to contribute the
interrelationships between T X and its utilities
time and resources to help de ne and enhance
E
dvips xdvi etc I designed and implemented
the value of the organization Only through a
the T X programs required to maintain our T X
thriving p opulation will the group continue to grow
E E
publications database and provide information on a
and contribute to the success of others needing the
variety of disciplines in many di erent formats I
supp ort of the T X Users Group
E
designed the program that generates Florida State
I feel that my exp erience in management and
University masters and do ctoral theses compliant
leadership qualify me to serve as a member of
with their graduation requirements Word of
the Board of Directors and wish to serve in that
this spread quickly so I routinely get requests
capacity
from other departments to copy it to their
systems I have continually provided my T X
Patricia Monohon
E
exp ertise to those wishing to design and distribute
not delivered for publication
their own do cuments and have assisted with the
installation upgrade of Unix T X systems prior to
E
the teTeX distributions and the creation of the
TUGb oat Pro duction Group work on installation
generation here lo cally
preliminary draft Apr preliminary draft Apr
preliminary draft Apr TUGb oat Volume No
Petr So jka Arthur Ogawa
Faculty of Informatics Cherokee Oaks Drive
Masaryk University Brno Three Rivers CA
Bure sova U S A
Brno Czech Republic Internet ogawa teleport com
Internet sojka informatics muni cz
Biography
Biography PhD Physics UC Berkeley I b egan working
I studied Math Computer Science at Masaryk with T X in my job at the Stanford Linear
E
University Brno where I work now as an Assistant Accelerator Center in ab out After a two
decade long career as an exp erimental physicist I Professor I am currently presidentof C TUG the
S
decided in to work as a publishing consultant zech and Slovak T X Users Group
E
and have run T X Consultants since that time I
E
Personal statement
presently live and work in Three Rivers California
My p otential activities for TUG could b e based
USA
on my exp erience gained in
I have assisted numerous p eople world
revitalization of C TUG
S
wide in using T X and have worked with
E
managing consulting leading various publication
a number of commercial rms in setting up
and other pro jects based on T X
E
and managing T X based publishing enterprises
E
maintenance of T X related services
E
For more p ersonal information please see
participation in organizing scienti c conferences
http www teleport com ogawa
I exp ect to help for b etter communication
Personal statement exchange between LUGs in Central East Europ e
I decided to o er my services on the TUG b oard and the rest of the world
b ecause I b elieve I can make a p ositive contribution
to TUG s future TUG is an organization worth
turning around and my intention is to do just
that with your help
The most serious indicator of TUG s troubles is
its declining membership so my goal is to address
that issue squarely I rmly b elieve that TUG can
accomplish this task by realigning its services and
b ene ts with its current and future membership
Isay that TUG can do this not that I can do
this simply b ecause no p erson can singlehandedly
achieve this goal So I will be relying on your
participation not just as a member of TUG but as
a TUG activist and volunteer
TUG needs your help to identify what member
b ene ts TUG should o er in order to be truly
relevant to the needs of T X users worldwide Then
E
your assistance will be needed in providing those
b ene ts whether through the vehicle of TUGb oat
training classes the annual conference our website
or other activities I b elievewe can make TUG once
again the vehicle for volunteers to b etter the lot of
T X users everywhere
E
If you share my goal of revitalizing TUG
through its own membership please commit your
energies to doing what is needed to make TUG the
success it once was Iinvite your input my email
address is mailto ogawa teleport com
TUG s meaning is only what we give it and its
very justi cation lies in what it gives its members
preliminary draft Apr preliminary draft Apr