path: root/language/ethiopia/ethiop/doc/ethiodoc.tex

\documentclass[a4paper]{article}
\usepackage[english]{babel}
\usepackage{ethiop}

\newcommand{\eth}{\selectlanguage{ethiop}}

\font\manual=logo10 % font used for the METAFONT logo, etc.
\newcommand*\MF{{\manual META}\-{\manual FONT}}
\newcommand*\babel{\textsf{babel}}
\newcommand*\Babel{\textsf{Babel}}
\newcommand*\EthTeX{Eth\kern-0,1em \TeX}
\newcommand*\ethioplogo{\textsf{ethiop}}

\newcommand{\servicemail}{%
  \texttt{ethiop@informatik.uni-hamburg.de}%
}

\newcount\xpos

\selectlanguage{english}
\hyphenation{Ethio-pi-an}

\begin{document}

\selectlanguage{english}

\title{\bf
  Ethiopian Language Support\\
  for the \Babel\ Package\\
  {\large Version 0.7}}
\author{
  \hfill\hbox to 0pt{\hss Berhanu Beyene\hss}\hfill
  \hfill\hbox to 0pt{\hss Manfred Kudlek\hss}\hfill\kern0pt\\
  \hfill\hbox to 0pt{\hss Olaf Kummer\hss}\hfill
  \hfill\hbox to 0pt{\hss Jochen Metzinger\hss}\hfill\kern0pt\\[\medskipamount]
  Universit\"at Hamburg,
  FB Informatik,
  AB TGI\\
  Vogt-K\"olln-Stra\ss e 30,
  D-22527 Hamburg}
\maketitle

\begin{abstract}
  The Ethiopian script differs considerably from the Latin script.
  Most important, it consists of more than 350 different letters.
  A new transcription method is presented that can be used for the
  \LaTeX\ typesetting system. It was implemented on the basis
  of the multilingual typesetting package \babel.

  In addition to a guide to the concrete usage of our system, we provide
  technical details of the implementation and sketch the
  reasons for our design decisions. Linguistic and historical
  information on the Ethiopian script is also included.
\end{abstract}
\selectlanguage{ethiop}
\begin{abstract}
  ya'iteyo.peyA .sehuf kalAtin .sehuf 'e^geg yatalaya naw:: 
  golto yamitAyaw ka 350 balAy fidalAten yayAza mahonu naw:: 
  'enazihen fidalAt 
  \selectlanguage{english}\LaTeX\selectlanguage{ethiop}
  batabAla ya.sehefat 'a.tA.tAl mAqanabAbariyA zadE ya'iteyo.peyAn 
  .sehuf balAtin
  quAnquA lama.sAf
  yamiyAs^cel 'adis yA'a.sA.sAf zaya 
  \selectlanguage{english}(Transcription)\selectlanguage{ethiop} 
  tazagA^geto qarbuAl::

  ya~nAn ya'a.sA.sAf zadE lama.taqam kamiyAs^cl 
  mamariyA gAr; ya'a.taqAqamun tEknikAl zerezernA lamen yehEn 
  ya'a.sA.sAf mangad lama.taqem 'endawasan 'asAytenAl:: sela 
  'ityo.pyA quAnquAwo^c 'a.sA.sAf zadEnA yaquAnquA tArikAwi 
  mara^gAwo^c bazih .tenAt lAy 'akelanbatAl::
\end{abstract}
\selectlanguage{english}

\clearpage

\section{Introduction}

The \ethioplogo\ package \cite{BKKM97} is a collection of fonts
and \TeX\ macros that enable you to typeset the characters
of the languages of Ethiopia.

\subsection{The origins of the Ethiopian script}

The Ethiopian script has its origin in the South Semitic alphabet
which has been used for Thamudene, Dedanite, Li\d{h}yanite,
\d{S}afaitic, Minaean, Sabaean, \d{H}imyaritic, Qatabanic, and
\d{H}a\d{d}ramautic. These are Semitic languages once spoken
in the present state of Yemen. The alphabet was without vocalization.
The South Semitic characters are known from stone monuments only,
whereas the present Ethiopian characters originate from
paper documents. The writing direction of South Semitic was from
right to left or alternating (Bustrofedon).

In the $2^{nd}$ half of the $1^{st}$ milliennium BC Semitic groups
(Ge'ez, {\eth g`ez}) from South Arabia established a kingdom at Aksum.
After 350 AD a vocalization was introduced.

Today the Ethiopian script is an official writing system in two
states, Ethio\-pia ({\eth 'ityo.peyA}) and Eri\-trea ({\eth 'ErtrA}).
There exist 86 languages (4 of them extinct)
from 4 language families with around 200 dialects in Ethiopia and
Eritrea. The writing system is used for a number of languages
in these states, see Tab.~\ref{tab:speakers} for the most
important languages (figures taken from \cite{ethno96}).
Note that Amarinya, Guraginya, Afarinya,
Hadiyyinya, Kambaatinya, Orominya (O\-ro\-miffa),
Sidaminya, Somalinya and some others are the expressions in Amharic
for the different languages whereas the names in the second column of
Tab.~\ref{tab:speakers} are from \cite{ethno96}.

\subsection{The origins of this package}

The current Ethiopian \LaTeX\ environment started
as a project assignment at the university of Hamburg in
1995. Luckily, we did not need to start from scratch,
because there were some Ethiopian fonts already.

The fonts are based on \EthTeX\ which was originally distributed
by Abass B. Alameneh. The genuine \EthTeX\ package can be
found on CTAN in the directory \texttt{language/ethiopia/ethtex/}
or in \cite{ethio97}. We changed the existing letters a little to make
them more robust at low resolutions, but the most effort went into
adding new characters that were not present in the original fonts.
Still more important is the addition of \TeX-ligatures to
the fonts in order to support our input transcription.

When we chose our transcription scheme, it became apparent that
we would have to activate some of the input characters.
Although we wrote our own set of macros to handle the
activation, we soon decided not to introduce yet another
incompatible mechanism for this task. Instead we used
the \babel\ package by Johannes L. Braams \cite{Bra97} as
framework for the implementation of the input transcription.
The \TeX\ macros are derived from the file
\texttt{language.skeleton} provided with that package, which
also allows a more well-rounded support of multiple
languages.

Since the original \EthTeX\ package used a special editor
that is not available for all platforms, it was not easily
portable. Moreover, a special version of \LaTeX\,2.09 was
generated, thereby preventing an upgrade to the now current
\LaTeXe. By rewriting the language support completely and
by embedding the Ethiopian fonts into the framework of
the \babel\ package we got a more robust and portable
system that will be usable with future \LaTeX\ versions.

\begin{table}[t]
\begin{center}
\begin{tabular}{|l|ll|r|}
\hline
  Language     & Languages       & &
    \multicolumn{1}{|l|}{No.~of speakers} \\
  family       &                 & &
    \multicolumn{1}{|l|}{in 1000} \\
               &                 & &
    \multicolumn{1}{|l|}{(year 91--95)} \\
\hline
  Semitic      & Ge'ez           & {\eth g`ez}        & only in church \\
               & Amharic         & {\eth 'amAr~nA}    & 20\,000 \\
               & Gurage          & {\eth gurAge~nA}   &  1\,850 \\
               & Tigre           & {\eth tegra}       &       ? \\
               & Tigrinya        & {\eth tegre~nA}    &  6\,050 \\
\hline
  Cushitic     & Afar            & {\eth 'afAr~nA}    &     750 \\
               & Hadiyya         & {\eth hadiy~nA}    &  1\,000 \\
               & Kambaata        & {\eth kambAt~nA}   &  1\,000 \\
               & Oromo           & {\eth 'OrOm~nA}    & 14\,000 \\
               & Sidamo          & {\eth sidAm~nA}    &  1\,500 \\
               & Somali          & {\eth sOmAl~nA}    &  2\,050 \\
\hline
  Omotic       & Gamo/Gofa/Dawro & {\eth gAm~nA}      &     780 \\
               & Wolaytta        & {\eth walAyet~nA}  &  2\,000 \\
\hline
  East Sudanic &                 &                    &       ? \\
\hline
\end{tabular}
\end{center}
\caption{\label{tab:speakers}Languages and number of speakers}
\end{table}

Our transcription method does not provide support for the direct
entry of Ethiopian characters. Instead a natural encoding
has been developed that allows us to enter Ethiopian
text via Latin letters. This encoding is based on 
scientific transcription techniques and is closely related
to other encoding standards. One of these standards is
SERA, which is mainly intended for the recording and
transmission of Ethiopian text within an ACSII environment.
However, the SERA encoding was not realizable as an
input encoding for \TeX.

Unicode, too, is an important text format, which
provides a unified framework for \textit{all}
languages by encoding characters with 16 bits instead of~8.
The \TeX\ extension $\Omega$ can handle Unicode
input. A rudimentary support for $\Omega$ has already 
been included in this package. Note that the Ethiopian
script is not a full part of Unicode. Although it has
been approved by the Unicode committee and has passed the
ISO/JTC1/SC2 ballot, it still awaits the ISO/JTC1 ballot and
the final publication.


\section{Installation}

Installing the \ethioplogo\ package is not overly difficult,
especially if you adhere to the following instructions.

\begin{enumerate}
\item Check the prerequisites for this package.

\begin{itemize}
\item Make sure you have installed \TeX\ and \MF.
\item Make sure that the files \texttt{cmbase.mf} and \texttt{romand.mf}
from the Computer Modern fonts are accessible to \MF.
\item Make sure that you have installed \LaTeXe\ with
a release date of 1996/12/01 or later. \LaTeXe\
can be found on CTAN in the directory \texttt{macros/latex/}.
\item Make sure that you have installed the \babel\ package with
a release date of 1997/01/23 (version 3.6h) or later. \Babel\
can be found on CTAN in the directory \texttt{macros/latex/packages/babel/}.
\end{itemize}

\item If a previous version of this package is installed, remove
all the files, especially the font files that were generated 
by \MF{} after the first installation.

\item Depending on how you obtained this package it might be
necessary to unpack/uncompress an archive. Now the files mentioned
in \texttt{MANIFEST} should be present.

\item If you do not intend to use the $\Omega$ typesetting system,
you may delete the files contained in the directory \texttt{omega/}.

\item Generate the \TeX\ files from their \texttt{docstrip}
source. To do this, run
\begin{verbatim}
tex ethiop.ins
\end{verbatim}
in the directory where the files \texttt{ethiop.ins} and
\texttt{ethiop.dtx} reside. (By default this is \texttt{latex/}.)

\item The files \texttt{ethiop.ins} and \texttt{ethiop.dtx}
can be removed, or you can run
\begin{verbatim}
latex ethiop.dtx
\end{verbatim}
to generate the source code documentation. \textit{This is not
required.}

\item Move the files to their destination.

\begin{tabular}{@{}lll@{}}
The files        & are                   & and are needed by \\[1ex]
\texttt{*.tfm}   & font metrics                       & \TeX \\
\texttt{*.fd}    & font definitions                   & \TeX \\
\texttt{*.sty}   & style files                        & \TeX \\
\texttt{*.ldf}   & \babel\ language
                   definitions                        & \TeX \\
\texttt{*.tex}   & \TeX\ sources                      & \TeX \\
\texttt{*.mf}    & \MF\ sources                       & \MF \\
\texttt{*.otp}   & $\Omega$ translation processes     & otp2ocp \\
\texttt{*.ocp}   & compiled \texttt{*.otp} files      & $\Omega$ \\
\texttt{*.ovp}   & $\Omega$ virtual font descriptions & ovp2ovf \\
\texttt{*.ovf}   & $\Omega$ virtual font files        & dvi-drivers \\
\texttt{*.ofm}   & $\Omega$ font metrics              & $\Omega$ \\
\end{tabular}

The exact location where the files belong depends on your
installation. As a first approximation, install them
near other files with the same extension.

For some installations it might be possible or even necessary
to place all the files in the directory where the user
documents will be placed. \textit{This is not recommended.}

\item Remove the font files that might be left over from a
previous version of ethiop, that is all files
\texttt{eth*.gf} and \texttt{eth*.pk}.

\item Verify the installation by generating this documentation
file from its source \texttt{ethiodoc.tex}. Run \LaTeX\ twice
to get the references right.
\begin{verbatim}
latex ethiodoc.tex
latex ethiodoc.tex
\end{verbatim}
The resulting file \texttt{ethiodoc.dvi} should be identical
to this text except for the date of translation.
\end{enumerate}

It is highly probable that after doing all of the above,
you have a working Ethiopian language package set up.
But maybe you ran into trouble during the installation.
In this case try the following:

\begin{itemize}
\item  If the run on \texttt{ethiop.ins} fails, the files
    might have been corrupted during transmission or one of
    the necessary files might not be accessible to \TeX.
\item  If \LaTeX\ complains about a missing input file, check whether
    the files \texttt{*.fd} and \texttt{*.sty} are accessible and readable.
\item  If \LaTeX\ complains about a missing font, check the
    placement of the files \texttt{*.tfm}.
\item  If \LaTeX\ issues warning messages, you might have an old version
    of \LaTeX\ or \babel.
\item  If \LaTeX\ issues strange errors, you might have an old version
    of \LaTeX\ or \babel.
\item  If \LaTeX\ issues strange errors, the files might have been
    corrupted during transmission. Conversions of CR, LF, and
    so on might cause this problem.
\item  If the previewer or the printer driver complains about missing
    fonts and does not automatically call \MF\ to generate
    these font, either adapt your installation or generate
    the fonts by hand. Depending on your installation you
    must run something like
\begin{verbatim}
mf '\mode=localfont; input etha10'
\end{verbatim}
    for each of the fonts.
\item  If \MF\ is called and complains about missing source
    files, check whether these (\texttt{*.mf}) are placed
    correctly.
\item  If \MF\ is called and complains about missing source
    files, check whether the Computer Modern fonts have been
    properly installed.
\item  If \MF\ is called and complains about strange paths,
    you are probably generating the
    font at a lower resolution than 200\,dpi.
    There is little you can do except ignoring the errors or
    telling \MF\ to do so. Please report such errors only
    if they occur at resolutions above 200\,dpi.
\item  If the previewer or the printer driver complains about missing
    characters, check whether you have deleted all files from
    previous versions of this font.
\item  If the previewer or the printer driver complains about a
    checksum error, check whether you have deleted all files from
    previous versions of this font.
\end{itemize}

If that does not help, have a look at our web page located at
{\small\verb|http://www.|\discretionary{}{}{}%
       \verb|informatik.uni-hamburg.de|%
       \verb|/TGI/mitarbeiter/wimis/kummer/ethiop_eng.html|}%
\hspace{0,33333em plus 3em}%
to read about possible updates and bug reports. If this
does not provide any clue and your friendly \TeX nician is unavailable,
we will try to help you, if time permits. Send a mail to
\servicemail, hopefully we will get back to you.

Please do \textbf{not} report bugs to Johannes L. Braams (the
maintainer of \babel) or to the \LaTeX\ team. They are \textbf{not}
responsible for our bugs and they are \textbf{very} busy already.


\section{Usage}

An important fact about the Ethiopian script is that
it uses more than 350 different characters. Hence the text
entry itself is a big problem. But we must also deal with
other topics like punctuation and spacing that arise when
using this package.

We will describe here the use of the \ethioplogo\ package
without the $\Omega$ system. The $\Omega$ support can be found
in section~\ref{omegasec}.

\subsection{Including the package}

The \babel\ language support is selected by adding the line
\begin{verbatim}
\usepackage[english]{babel}
\end{verbatim}
to the preamble of a document. Of course a different language
might also be selected. But since the \ethioplogo\ package is not yet
a part of the \babel\ package, you must select it with
\begin{verbatim}
\usepackage{ethiop}
\end{verbatim}
in the preamble of a document. There is no need to include
\babel\ explicitly unless we want to use two or more languages.
If both \ethioplogo\ and \babel\ are included, we can use the ordinary
language switching mechanism of \babel\ to take us
from one language to the other. For example
\begin{verbatim}
\selectlanguage{english}'adis 'ababA
\selectlanguage{ethiop}'adis 'ababA
\selectlanguage{english}'adis 'ababA.
\end{verbatim}
will give you:
\selectlanguage{english}'adis 'ababA
\selectlanguage{ethiop}'adis 'ababA
\selectlanguage{english}'adis 'ababA.


\subsection{Typing the text}

In Tab.~\ref{tab:ethiopchar} you can see the Ethiopian
characters that are accessible using this package.

All characters with {\tt .d} are only used for {\it Orominya (Oromiffa)},
all with {\tt 'q}, {\tt 'k}, {\tt 'h}, {\tt 'g} and
{\tt mua}, {\tt mui}, {\tt muE}, {\tt mue} only for
{\it Guraginya (Chaha)}, all with {\tt .q} only for {\it Tigrinya},
and all with {\tt fu}, {\tt pu} only for {\it Wolayttinya}.

\begin{table}[p]
  \begin{center}
    \selectlanguage{ethiop}
    \def\sci|#1|{\selectlanguage{english}%
      \mdseries #1%
      \selectlanguage{ethiop}}
    \begin{tabular}{|c|r|ccccccc|ccccc|}
      \cline{3-14}
      \multicolumn{2}{r|}{}&
        \sci|a|&\sci|u|&\sci|i|&\sci|\=a|&\sci|\=e|&\sci|e|&\sci|o|&
        \sci|wa|&\sci|wi|&\sci|w\=a|&\sci|w\=e|&\sci|we|\\
      \cline{3-14}
      \multicolumn{2}{r|}{}&
        \verb|a|&\verb|u|&\verb|i|&\verb|A|&\verb|E|&\verb|e|&\verb|o|&
        \verb|ua|&\verb|ui|&\verb|uA|&\verb|uE|&\verb|ue|\\
      \multicolumn{2}{r|}{}&
        &\verb|U|&\verb|I|&&&&\verb|O|&&\verb|uI|&
        \multicolumn{3}{r|}{\textrm{\small\itshape\mdseries if preferred}}\\
      \hline
      \sci|h|   &\verb| h|& ha&hU&hI&hA&hE&he&hO&&&huA&&\\
      \sci|l|   &\verb| l|& la&lU&lI&lA&lE&le&lO&&&luA&&\\
      \sci|\d h|&\verb|.h|& .ha&.hU&.hI&.hA&.hE&.he&.hO&&&.huA&&\\
      \sci|m|   &\verb| m|& ma&mU&mI&mA&mE&me&mO&mua&muI&muA&muE&mue\\
      \sci|\'s| &\verb|'s|& 'sa&'sU&'sI&'sA&'sE&'se&'sO&&&'suA&&\\
      \sci|r|   &\verb| r|& ra&rU&rI&rA&rE&re&rO&&&ruA&&\\
      \sci|s|   &\verb| s|& sa&sU&sI&sA&sE&se&sO&&&suA&&\\
      \sci|\v s|&\verb|^s|& ^sa&^sU&^sI&^sA&^sE&^se&^sO&&&^suA&&\\
      \sci|q|   &\verb| q|& qa&qU&qI&qA&qE&qe&qO&qua&quI&quA&quE&que\\
      \sci|\d q|&\verb|.q|& .qa&.qU&.qI&.qA&.qE&.qe&.qO&
                            .qua&.quI&.quA&.quE&.que\\
      \sci|b|   &\verb| b|& ba&bU&bI&bA&bE&be&bO&bua&buI&buA&buE&bue\\
      \sci|v|   &\verb| v|& va&vU&vI&vA&vE&ve&vO&&&vuA&&\\
      \sci|t|   &\verb| t|& ta&tU&tI&tA&tE&te&tO&&&tuA&&\\
      \sci|\v c|&\verb|^c|& ^ca&^cU&^cI&^cA&^cE&^ce&^cO&&&^cuA&&\\
      \sci|\b h|&\verb|_h|& _ha&_hU&_hI&_hA&_hE&_he&_hO&
                            _hua&_huI&_huA&_huE&_hue\\
      \sci|n|   &\verb| n|& na&nU&nI&nA&nE&ne&nO&&&nuA&&\\
      \sci|\~n| &\verb|~n|& ~na&~nU&~nI&~nA&~nE&~ne&~nO&&&~nuA&&\\
      \sci|'|   &\verb|'| & 'a&'U&'I&'A&'E&'e&'O&'ua&&&&\\
      \sci|k|   &\verb| k|& ka&kU&kI&kA&kE&ke&kO&
                            kua&kuI&kuA&kuE&kue\\
      \sci|\b k|&\verb|_k|& _ka&_kU&_kI&_kA&_kE&_ke&_kO&
                            _kua&_kuI&_kuA&_kuE&_kue\\
      \sci|w|   &\verb| w|& wa&wU&wI&wA&wE&we&wO&&&&&\\
      \sci|`|   &\verb| `|& `a&`U&`I&`A&`E&`e&`O&&&&&\\
      \sci|z|   &\verb| z|& za&zU&zI&zA&zE&ze&zO&&&zuA&&\\
      \sci|\v z|&\verb|^z|& ^za&^zU&^zI&^zA&^zE&^ze&^zO&&&^zuA&&\\
      \sci|y|   &\verb| y|& ya&yU&yI&yA&yE&ye&yO&yua&&&&\\
      \sci|d|   &\verb| d|& da&dU&dI&dA&dE&de&dO&&&duA&&\\
      \sci|\d d|&\verb|.d|& .da&.dU&.dI&.dA&.dE&.de&.dO&&&.duA&&\\
      \sci|\v g|&\verb|^g|& ^ga&^gU&^gI&^gA&^gE&^ge&^gO&&&^guA&&\\
      \sci|g|   &\verb| g|& ga&gU&gI&gA&gE&ge&gO&
                            gua&guI&guA&guE&gue\\
      \sci|\d g|&\verb|.g|& .ga&.gU&.gI&.gA&.gE&.ge&.gO&
                            .gua&.guI&.guA&.guE&.gue\\
      \sci|\d t|&\verb|.t|& .ta&.tU&.tI&.tA&.tE&.te&.tO&&&.tuA&&\\
      \sci|$\check{\textrm{\d c}}$|&\verb|^C|&
                            ^Ca&^CU&^CI&^CA&^CE&^Ce&^CO&&&^CuA&&\\
      \sci|\d p|&\verb|.p|& .pa&.pU&.pI&.pA&.pE&.pe&.pO&&&.puA&&\\
      \sci|\d s|&\verb|.s|& .sa&.sU&.sI&.sA&.sE&.se&.sO&&&.suA&&\\
      \sci|\d c|&\verb|.c|& .ca&.cU&.cI&.cA&.cE&.ce&.cO&&&&&\\
      \sci|f|   &\verb| f|& fa&fU&fI&fA&fE&fe&fO&fua&fuI&fuA&fuE&fue\\
      \sci|p|   &\verb| p|& pa&pU&pI&pA&pE&pe&pO&pua&puI&puA&puE&pue\\
      \cline{10-14}
      \sci|\'q| &\verb|'q|& 'qa&'qU&'qI&'qA&'qE&'qe&'qO\\
      \cline{12-14}
      \sci|\'k| &\verb|'k|& 'ka&'kU&'kI&'kA&'kE&'ke&'kO&&&
      \multicolumn{3}{|l|}{{\selectlanguage{english}\texttt{\char126 mA}}\hfill ~mA}\\
      \sci|\'h| &\verb|'h|& 'ha&'hU&'hI&'hA&'hE&'he&'hO&&&
      \multicolumn{3}{|l|}{{\selectlanguage{english}\texttt{\char126 ri}}\hfill ~ri}\\
      \sci|\'g| &\verb|'g|& 'ga&'gU&'gI&'gA&'gE&'ge&'gO&&&
      \multicolumn{3}{|l|}{{\selectlanguage{english}\texttt{\char126 fi}}\hfill ~fi}\\
      \cline{1-9}\cline{12-14}
    \end{tabular}
    \selectlanguage{english}%
    \caption{The Ethiopian characters}
    \label{tab:ethiopchar}
  \end{center}
\end{table}

We will now explain how the characters are entered.
Every character represents a syllable that consists of a
consonant followed by a vowel. If possible, every character
is encoded in a way that matches its pronounciation or its
scientific transcription as closely as possible.

As an example we choose the letter {\eth da} which is pronounced
\textit{da} and entered as \verb:da:. You will find the
character in the character table in row \texttt{d} and
column \texttt{a}.

The letter {\eth de} might represent the syllable \textit{de},
but it might also denote the consonant
\textit{d} without an accompanying vowel. To reflect this,
it is possible to enter either \verb:de: or \verb:d: at
the users choice.

If accented characters are used in the
scientific transcription of an Ethiopian syllable, they are
entered without the accent, but prefixed with an appropriate
special character. {\eth ^ca} has \textit{\v ca} as its transcription,
hence we will enter it as \verb:^ca: which is as close
to the proper transcription as we can get.

Long vowels are usually indicated by a bar,
\textit{d\=e} can serve as an example. But since long vowels
are fairly frequent, we do not want to use two letters for their
representation. Instead we will employ the uppercase letters
for this purpose, which leads us to \verb:dE: as our transcription
of the aforementioned syllable {\eth dE}.

When a vowel does not have both a short and a long form, like
the \textit{i} in \textit{di}, there is no need to insist on the
proper case for the vowel. Hence we might enter either \texttt{di}
or \texttt{dI} and get {\eth dI} in both cases.

Some consonants may be followed by a diphtong,
i.e.\ a combination of two vowels. A diphtong occurs e.g.\ in
{\eth duA} which is best transcribed \textit{dw\=a}.
We will code it as \verb:duA:, thereby slightly deviating
from the standard transcription. But this is unavoidable,
because if we enter \verb:dwA:, we will get an ambiguity
with \texttt{d\,wA} which we want to result in {\eth dwA}.
The SERA encoding, which is developed by
Daniel Yacob, Yitna Firdyiwek, and Yonas Fisseha,
suggests \verb:dWA:, which has been considered as an alternative
to the present encoding. It did not show any
significant advantages, however.

For the consonant series\quad{\eth
^Ca ^CU ^CI ^CA ^CE ^Ce ^CO ^CuA}\quad there exists a variant
form that looks like\quad{\eth\ethvariantCtrue
^Ca ^CU ^CI ^CA ^CE ^Ce ^CO ^CuA}\quad and
denotes the same syllables. The variant series can be activated by 
issuing the command \verb|\ethvariantCtrue| once.
The original letterforms can then be restored by typing
\verb|\ethvariantCfalse|. This option will only
be needed in comparative studies, usually the standard
series should be preferred.

In the lower right corner of Tab.~\ref{tab:ethiopchar} you can 
see an inlay with the three characters {\eth ~mA}, {\eth ~ri}, and
{\eth ~fi}. These characters are probably the remnants of
three complete series of seven syllables that were once used.

We added these three in order to
completely cover the character set of the proposed Unicode
standard for code positions 1200h to 137Fh. The encoding of
these characters is not fixed yet and may change at any
time in the future. At the moment the encoding is based on
the characters' appearance, but we are aware that the
pronounciation is different today.


\subsection{Punctuation}

Although the punctuation characters look different from
the punctuation of the Latin script, they have essentially the
same meaning. We made some compromises between
visual similarity and similar interpretation when
we chose the encoding of the punctuation characters.
In Tab.~\ref{tab:ethioppunc}--\ref{tab:ethiopspecial}
we have collected the appropriate inputs for each of the
characters.

\selectlanguage{ethiop}:-\selectlanguage{english}

\begin{table}[htb]
  \begin{center}
    \begin{tabular}{l|ccccc|cc|cccc|c}
      input&
      \verb|:=|&
      \verb|:-|&
      \verb|::|&
      \verb:,:&
      \verb:;:&
      \verb:|:&
      \verb;:|:;&
      \verb:?:&
      \verb:'?:&
      \verb:!:&
      \verb:'!:&
      \verb:...:\\
    \hline
      output&
      \selectlanguage{ethiop}:=\selectlanguage{english}&
      \selectlanguage{ethiop}:-\selectlanguage{english}&
      \selectlanguage{ethiop}::\selectlanguage{english}&
      \selectlanguage{ethiop},\selectlanguage{english}&
      \selectlanguage{ethiop};\selectlanguage{english}&
      \selectlanguage{ethiop}|\selectlanguage{english}&
      \selectlanguage{ethiop}:|:\selectlanguage{english}&
      \selectlanguage{ethiop}?\selectlanguage{english}&
      \selectlanguage{ethiop}'?\selectlanguage{english}&
      \selectlanguage{ethiop}!\selectlanguage{english}&
      \selectlanguage{ethiop}'!\selectlanguage{english}&
      \selectlanguage{ethiop}...\selectlanguage{english}\\
    \end{tabular}
  \end{center}
  \caption{The Ethiopian punctuation characters}
  \label{tab:ethioppunc}
  \vspace{\floatsep}
  \begin{center}
    {\selectlanguage{ethiop}
    \begin{tabular}{l|cccc|cccc}
      \selectlanguage{english}input&
      \verb:<:&\verb:<<:&\verb:>:&\verb:>>:&
      \verb:':&\verb:'':&\verb:`:&\verb:``:\\
    \hline
      \selectlanguage{english}output&
      <&<<&>&>>&'&''&`&``\\
    \end{tabular}
    }
  \end{center}
  \caption{The Ethiopian quotation characters}
  \label{tab:ethiopquot}
  \vspace{\floatsep}
  \begin{center}
    {\selectlanguage{ethiop}
    \begin{tabular}{l|cc|cc|cc|c}
      \selectlanguage{english}input&
      \verb:(:&\verb:):&\verb:[:&\verb:]:&\verb:\{:&\verb:\}:&
      \verb:\$:\\
    \hline
      \selectlanguage{english}output&
      (&)&[&]&\{&\}&\$\\
    \end{tabular}
    }
  \end{center}
  \caption{Special characters}
  \label{tab:ethiopspecial}
\end{table}

The punctuation characters match their SERA equivalents closely.


\subsection{Spaces}

When the Ethiopian script is printed today, an interword gap
is signalled by a white space, in the same way as it is
done for the Latin script. This kind of spacing can
be used simply as in ordinary \LaTeX\ documents.

But in former times word breaks used to be denoted by the
character {\eth :} and even today this method is used for
handwriting. To get the appropriate effect look at the following
\LaTeX\ source
\begin{verbatim}
'abAs : 'alamenahe: 'abAs :'alamenahe:
'abAs:'alamenahe:'abAs:'alamenahe:
'abAs:'alamenahe:'abAs:'alamenahe:
'abAs:'alamenahe:'abAs:'alamenahe::
\end{verbatim}
which results in
\begin{quote}
\eth
'abAs : 'alamenahe: 'abAs :'alamenahe:
'abAs:'alamenahe:'abAs:'alamenahe:
'abAs:'alamenahe:'abAs:'alamenahe:
'abAs:'alamenahe:'abAs:'alamenahe::
\end{quote}
in the output. As you can see, line breaks are allowed after
the {\eth :} even if there is no space character in the
source. Space characters immediately following or preceeding
a \texttt{:} in the input are ignored as we can see from
the first line. Therefore newlines in the input
will not cause any problem either.


\subsection{Line breaking}

The ordinary spaces as well as the white space surrounding
an~{\eth :}~can be stretched a little,
so that it is possible to achieve proper justification.
Nevertheless, the lack of hyphenation in the Ethiopian language
makes itself felt from time to time, when \TeX\ cannot find
suitable breakpoints for a paragraph.

There are a few standard solutions to this problem, the easiest
is to use a \texttt{sloppypar} enviroment which allows \TeX\ to
stretch the interword spaces more than usual. But this does not
work when some words are simply too long. In that case
one may want to rewrite the sentence that causes the bad
break, maybe only by changing a few words.

However, the text might not be easily changable, e.g.~because
it is a quote from some other source, or the author insists
on that very phrase. In that case you an insert a \verb|\-|
into a suitable breakpoint where the text will be split
between two lines. Unlike the usual \TeX\ behaviour,
no hyphen will be added at the breakpoint. This kind of
line breaking is especially well suited when the
character~{\eth :}~is used for interword spaces.

But maybe even the insertion of break points is impossible.
In this case, some explicit
\verb|\hspace| must be added in a suitable position or
a raggedright layout must be selected during the whole document or
part thereof.


\subsection{Numbers}

Since today Arabic numbers are more frequently used than
the original Ethio\-pian numbers, the \ethioplogo\ package
outputs the Arabic numbers when the letters \texttt{0}
up to \texttt{9} occur in the source code.

But Ethiopian numbers can be typeset, too, by
using the command \verb|\ethnum|. If we enter \verb|\ethnum{1}|
we get {\eth\ethnum{1}} as the result. But the macro
\verb|\ethnum| can do more than that. In fact it can convert all
numbers up to 999\,999 to their Ethiopian equivalents:
\verb|\ethnum{999999}| gives {\eth\ethnum{999999}}.
We can see that the program knows quite a lot about
the representation of large numbers.

\begin{table}[thp]
  \begin{center}
    \begin{tabular}{l|ccccccccc}
      Arabic&
      1&2&3&4&5&6&7&8&9\\
    \hline
      Ethiopian&
      {\eth\ethnum{1}}&{\eth\ethnum{2}}&{\eth\ethnum{3}}&
      {\eth\ethnum{4}}&{\eth\ethnum{5}}&{\eth\ethnum{6}}&
      {\eth\ethnum{7}}&{\eth\ethnum{8}}&{\eth\ethnum{9}}\\
    \end{tabular}

    \bigskip

    \begin{tabular}{l|ccccccccc|cc}
      Arabic&
      10&20&30&40&50&60&70&80&90&100&10000\\
    \hline
      Ethiopian&
      {\eth\ethnum{10}}&{\eth\ethnum{20}}&{\eth\ethnum{30}}&
      {\eth\ethnum{40}}&{\eth\ethnum{50}}&{\eth\ethnum{60}}&
      {\eth\ethnum{70}}&{\eth\ethnum{80}}&{\eth\ethnum{90}}&
      {\eth\ethnum{100}}&{\eth\ethnum{10000}}\\
    \end{tabular}
  \end{center}
  \caption{The Ethiopian numbers}
  \label{tab:ethiopnum}
\end{table}

In \LaTeX\ we must output the contents of a counter
from time to time. This can be accomplished using the
macro \verb|\ethiop|. Saying \verb|\ethiop{subsection}| will
cause \TeX\ to print {\eth\ethiop{subsection}}, since this
is subsection \arabic{subsection}.


\subsection{Math mode}

One of \TeX's most important features is its math mode.
We can use math within Ethiopian text, but
by default all letters in mathematical formulas will
be taken from the ordinary \TeX\ fonts.

Ethiopian letters can be used in a formula, although this
will require slightly more work. When the macro
\verb|\ethmath| appears in math mode while the
Ethiopian language is selected, its single argument will
appear in the proper size typeset with the Ethiopian fonts.

For example
\begin{verbatim}
$$b+\frac{d}{f^g}\iff
  \ethmath{ba}+\frac{\ethmath{da}}{\ethmath{fa}^{\ethmath{ga}}}$$
\end{verbatim}
will result in
\selectlanguage{ethiop}
$$b+\frac{d}{f^g}\iff
  \ethmath{ba}+\frac{\ethmath{da}}{\ethmath{fa}^{\ethmath{ga}}}$$
\selectlanguage{english}%
where you should note the varying fonts in the first and second
subformula. Actually all the work is done by the macro
\verb|\ethmath| that outputs the Ethiopian characters.


\subsection{Ethiopian dates}

The Ethiopian calendar is based on the Julian calendar with
twelve months of 30~days and one month of 5~days. Every fourth
year is a leap year, which means that the last month will have
6~days. The calendar system is implemented in our package,
so it is possible to type \verb|\today| and get {\eth\today}.
(This is the date on which this document has
been translated. Compare it to the date on the title page!)


\subsection{Two examples}

We provide the first sample text to illustrate the appearance of
our font. The \LaTeX\ source of the text begins with
\begin{verbatim}
\subsubsection*{`amala~nAytun mArq}

pAduwA bametbAl web ya'i.tAliyA katamA si~nor bAptisetA 
yatasa~nu 'and tegu_h `sarAta~nA yenofu nabar:: 'enih sawe 
hulat qon^go sEto^c le^g^c nabaruA^caw::
\end{verbatim}
and gives the following result:

\selectlanguage{ethiop}
\subsubsection*{`amala~nAytun mArq}

pAduwA bametbAl web ya'i.tAliyA katamA si~nor bAptisetA 
yatasa~nu 'and tegu_h 'sarAta~nA yenoru nabar:: 'enih sawe 
hulat qon^go sEto^c le^g^c nabaruA^caw::
kuAduA.helinA yatabAla^cew webatuAnA
.tabAyua yatamuAlA naw:: puafEluA 
webatuA 'enda 'e_hetuA 'amuAlto selAlesa.tAt ba`amala~nAnatuA 
tetAwaqAla^c::

puafEluA lAbAtuA yabakuir le^g bamahonuA bazih 'alabelAbit 
melAsuA sAbiyA menAlbAt bAl 'a.tetA qumo-qar 'enedAthon 'abAtuA 
segAt 'aderobA^caw nabar::

yafarut 'alqaram laqon^gowA kavEreno 'eska quidA_hualA sigorf 
la`amal~nAwA 'a.tagabuA yamiders saw .tafA:: 
\selectlanguage{english}

The next example will clarify the usage of bold and
slanted Ethiopian fonts. Italic characters are mapped to
slanted characters. The font selection works just as in
ordinary \LaTeX\ with NFSS. The source text
\begin{verbatim}
'adis 'ababA \textbf{'adis 'ababA}
\textsl{'adis 'ababA \textbf{'adis 'ababA}}
\end{verbatim}
gives us
\selectlanguage{ethiop}
'adis 'ababA \textbf{'adis 'ababA}
\textsl{'adis 'ababA \textbf{'adis 'ababA}}
\selectlanguage{english}
as the output.


\subsection{Using \ethioplogo\ with Arab\TeX}

Starting with version 3.6i, \babel\ is now compatible with
Arab\TeX. But still there are some problems with the many active 
characters which are used by \ethioplogo, so that the two packages
do not work right away.

But it is possible to use Arab\TeX\ with \ethioplogo\ by
including the special style \texttt{etharab.sty} after 
Arab\TeX\ has been loaded. Some internal macros of Arab\TeX\ are
redefined, so this is not guaranteed to work with every version
of Arab\TeX, but it has been successfully used with version
3.06g3 of Arab\TeX.

\begin{verbatim}
\documentclass{article}
\usepackage{arabtex}
\usepackage[english]{babel}
\usepackage{ethiop}
\usepackage{etharab}
\begin{document}
  \selectlanguage{english}
  The Arabic (<al.har_t>) and the Ethiopian script 
  (\selectlanguage{ethiop}sel.tAnE\selectlanguage{english})
  may occur within one sentence.
\end{document}
\end{verbatim}

We do not provide the output of the example, because this document is
intended to be translatable even in the absence of Arab\TeX.
In fact, everything works as usual, Arabic text can be inserted
using the \texttt{arabtext} environment or using
\verb:<>: pairs. However, these commands must not be used in 
arguments to other commands or in command definitions. If that
is desired the complete commands or command definitions must be
enclosed in a \texttt{noethiop} environment.
\begin{verbatim}
\begin{noethiop}
\section{<al.har_t> -- cultivation}
\end{noethiop}
\end{verbatim}
provides an example. If Ethiopian characters are needed, too,
then a little trick is in order.
\begin{verbatim}
\def\temptext{%
  \selectlanguage{ethiop}sel.tAnE\selectlanguage{english}}
\begin{noethiop}
\section{<al.har_t> -- cultivation -- \temptext}
\end{noethiop}
\end{verbatim}
But there should be really, really few occasions for 
such ugly code.


\section{Advanced topics}

Although the usage of the \ethioplogo\ package is not really difficult
once one gets used to it, there are a few points to be
aware of. We will highlight the internal structure
of the package first, to make it more plausible why some
problems just cannot be easily patched away.


\subsection{Implementation notes}

Because we have to deal with so many characters, we
placed them in two separate fonts. The two codetables
are shown in Tab.~\ref{codetab:etha} and
Tab.~\ref{codetab:ethb}, together with the recommended
input strings. (Remember that there might be different
ways to achieve the same result.)

\begingroup
  \catcode`\~=12\relax
  \catcode`\^=12\relax
  \catcode`\_=12\relax
  \def\x{\endgroup
    \let\plaintilde=~
    \let\plainhatacc=^
    \let\plainbaracc=_
  }
\x

\newcommand{\codetable}[2]{%
  \begin{table}[p]
    \begin{center}
      \global\xpos=0
      ~%
      \kern-2cm
      \begin{tabular}{@{}r|l|l|l|l|l|l|l|l|}%
        \multicolumn{1}{r}{}&
        \multicolumn{1}{l}{~\texttt{0}}&
        \multicolumn{1}{l}{~\texttt{1}}&
        \multicolumn{1}{l}{~\texttt{2}}&
        \multicolumn{1}{l}{~\texttt{3}}&
        \multicolumn{1}{l}{~\texttt{4}}&
        \multicolumn{1}{l}{~\texttt{5}}&
        \multicolumn{1}{l}{~\texttt{6}}&
        \multicolumn{1}{l}{~\texttt{7}}\\
        \cline{2-9}%
        \input{#1}%
      \end{tabular}%
      \kern-2cm
      ~%
    \end{center}
    \caption{\label{codetab:eth#2}The Ethiopian codetable ETH\uppercase{#2}}
  \end{table}
}

\newcommand{\outputcode}[1]{%
  \ifx\relax#1\relax
    %\rule{1em}{1ex}%
  \else
    \rule[-4pt]{0pt}{14pt}%
    {\def~{\plaintilde}%
     \def^{\plainhatacc}%
     \def_{\plainbaracc}%
     \texttt{#1}}%
    ~\hfill
    {\foreignlanguage{ethiop}{#1}}%
  \fi
}

\newcommand{\docode}[2]{%
  \ifnum\xpos=0 \relax
    \texttt{#1}~&%
  \fi
  \outputcode{#2}%
  \global\advance\xpos by 1
  \ifnum\xpos=8 \relax
    \global\xpos=0
    \\\cline{2-9}%
  \else
    &%
  \fi
}

\codetable{codeetha.tex}{a}

\renewcommand{\docode}[2]{%
  \ifnum\xpos=0 \relax
    \texttt{#1}~&%
  \fi
  \ifnum#1>48 \relax
    \ifnum#1<69 \relax
      \hfill
      \foreignlanguage{ethiop}{\fontfamily{ethb}\selectfont\char#1}%
    \else
      \ifnum#1>191 \relax
        \ifnum#1<200 \relax
          \hfill
          \foreignlanguage{ethiop}{\fontfamily{ethb}\selectfont\char#1}%
        \else
          \outputcode{#2}%
        \fi
      \else
        \outputcode{#2}%
      \fi
    \fi
  \else
    \outputcode{#2}%
  \fi
  \global\advance\xpos by 1
  \ifnum\xpos=8 \relax
    \global\xpos=0
    \\\cline{2-9}%
  \else
    &%
  \fi
}

\codetable{codeethb.tex}{b}

In the first font we preferred to place characters that
result from a \TeX\ ligature (which must not be confused
with a ligature from ordinary printing) in the positions
0--31 and 128--255. These characters usually cannot be
entered from a keyboard and hence it is safe to assume that
they resulted from a ligature.

Only the characters in the primary font are accessible
by entering ordinary characters and forming ligatures.
For the other characters it is necessary to explicitly
select the secondary font within the \TeX\ code.
But this requires the execution of \TeX\ macros, hence
the activation of some characters was required. The
activated characters can inspect the following
characters and request the necessary font change.
In fact, this method of enlarging the number of
available characters is quite general and might be used for
other languages, too, e.g.\ to provide a unified
input mechanism for all Latin characters.

For a complete documentation of the input parsing mechanism
we refer the interested reader to the commented source code
in \texttt{ethiop.dtx}, which can also be typeset by
\LaTeX\ to get a more readable version.

A long calculation is required to convert the
Gregorian date provided by the \TeX\ primitives
\verb|\year|, \verb|\month|, and \verb|\day| to the
Ethiopian date. The implementation in \TeX\ is straightforward,
but hardly readable, because \TeX's expressiveness for
formulas is very weak. As calendar routines are provided
for all the other \babel\ language definitions, it was obvious
that this problem had to be addressed.


\subsection{Common pitfalls}

After discussing some of the internals of the
\ethioplogo\ package, we are now prepared to examine
some of the problems that result from our implementation.

First of all the characters \verb|~|,
\verb|^|, \verb|'|, \verb|_|, and \verb|.| are
made active. This is unavoidable, but there are some
drawbacks.
\begin{itemize}
\item We cannot use~\verb|^|\verb|^| for entering special characters.
  Usually this is done in package files only, so we do not
  get into real trouble, since \babel\ activates the
  characters only at the beginning of the document.
\item We cannot use a~\verb|.| in numbers and \TeX\ dimensions
  while Ethiopian text is being typeset. We can circumvent
  this problem by using a~\verb|,| instead of the~\verb|.| when
  entering numbers for \TeX.
  Note, that we can use the~\verb|.| without
  problems when we have temporarily switched to a language
  other than Ethiopian, e.g.\ English.
\end{itemize}
We could expect that the activation of~\verb|^| and~\verb|_|
spoils \TeX's math mode, but this is not the case.
In fact math mode behaves just like before, with subscripts
and superscripts in their proper position.

The complex calendarical calculation require the allocation
of a large number of counters to hold the intermediate
results.
\begin{itemize}
\item In connection with other counter intensive packages
\LaTeX\ may run out of counters when using the
\ethioplogo\ package. Maybe we can get rid of two or three of the
counters in the next version, but it will still remain
a problem.
\end{itemize}
This problem is actually due to the lack of temporary
counters in \LaTeX, which are not provided, even though this
is done for all the other types of registers.

Since our package is not yet an integral part of the
\babel\ system, we have some other difficulties to overcome.
\begin{itemize}
\item The \ethioplogo\ package cannot be loaded
via an option to \babel. Instead an explicit
\verb|\usepackage{ethiop}| has to be used.

\item We do not know what the future will bring. While this
package \textit{might} work with future versions of
\babel, there is no guarantee that it will. So you should keep
your old version of \babel\ until you are sure it works
with the \ethioplogo\ package or until an updated version of
\ethioplogo\ is issued.
\end{itemize}

Typing errors will usually not result in
an error message. Instead a black rectangle will appear
in the output, if some illegal character is encountered.
\begin{itemize}
\item Because it is allowed to enter consonants without a
trailing vowel, there are plenty of typos that
simply result in the wrong letters being printed.
\end{itemize}
So look at your finished document carefully.


\subsection{Support of the $\Omega$ typesetting system}
\label{omegasec}

The advantage of using $\Omega$ for typesetting the
Ethiopian language is that it can handle text
files encoded in Unicode. Therefore the complex,
timeconsuming and error-prone conversion process
needed our transliteration within \TeX\ can be
skipped.

$\Omega$ can still profit from \babel's support for
captions, dates etc.
In $\Omega$ you can simply enter the Ethiopian charaters
as their Unicode equivalents. They will be converted to
our fonts by means of virtual fonts.

You will have to experiment when you want to use $\Omega$
with \ethioplogo. Tests have been only rudimentary so far.
There are some currently unused files that may help you.
\texttt{ethotlit.otp} simulates the old transliteration process
in $\Omega$. \texttt{ethohyph.otp} allows word breaking between every
syllable.

Let us note a few differences between our package and
Unicode.
\begin{itemize}
\item The letter {\eth huA} (\verb|huA|) is not present in Unicode. 
It is mentioned in \cite{Wedekind95}. 
It is used in the language of Agew ({\eth 'agaw},
also known as Awngi). \cite{Wedekind95} also uses {\eth _kuA}
(\verb|_kuA|) as an alternative representation for the same
sound.

An example word would be {\eth sOhuA}, which means in Agew
\emph{to eat}.

\item The letter {\eth mui} (\verb|mui|) is equipped with a additional tail
at the lower left in Unicode.

\item The four series {\eth 'q} (\verb|'q|),
{\eth 'k} (\verb|'k|),
{\eth 'h} (\verb|'h|), and
{\eth 'g} (\verb|'g|) are not present in Unicode.
They are suggested as possible extensions, but will
not occur in Unicode in the near future, as it seems.
\end{itemize}



\section{Changes}

Version 0.2 was the first version to be publicly released,
but we will shortly list the changes that this version made
to \EthTeX.

\begin{itemize}
\item Several letters were added, most notably the Leslau extensions
\verb|'q|, \verb|'k|, \verb|'h|, and \verb|'g|. Some diphtongs
were added, too.

\item The multi-letter encoding was chosen and implemented
using \babel.

\item The calendar algorithms were programmed.
\end{itemize}


\subsection*{Version 0.3}

\begin{itemize}
\item The letter {\eth huA} (\verb|huA|) was added.

\item Some bugs regarding subscripts and superscripts 
in ordinary math mode were fixed.

\item The command \verb|\ethmath| was added to allow 
Ethiopian characters in math mode.

\item Fonts in 5 point size were added to allow
Ethiopian characters in subscripts and superscripts.
\end{itemize}


\subsection*{Version 0.4}

\begin{itemize}
\item The punctuation characters {\eth |} and
{\eth :|:} were added.

\item The syllables {\eth ~mA} (\verb|~mA|), {\eth ~ri} (\verb|~ri|), and
{\eth ~fi} (\verb|~fi|) were added. Now all characters of the proposed
Unicode standard for the code positions U+0x1200h to U+0x137F are 
included.

\item \verb|\ethvariantCtrue| and \verb|\ethvariantCfalse| were
introduced. At the same time the shorthand \verb|_C| was removed.
The variant forms of the series \verb|^C| can only be reached via
the aforementioned commands from now on.
\end{itemize}


\subsection*{Version 0.5}

\begin{itemize}
\item Problems with activating \verb|'| in math mode have been
solved.
\end{itemize}

\subsection*{Version 0.6}

\begin{itemize}
\item \texttt{etharab.sty} has been added to allow cooperation
with Arab\TeX.
\end{itemize}

\subsection*{Version 0.7}

\begin{itemize}
\item As suggested by Donald~E.~Knuth, some
faulty parameters in the files \texttt{etha8.mf} and
\texttt{ethb8.mf} were corrected.

\item As suggested by Donald~E.~Knuth, individual glyphs
for the characters {\eth :|:} and {\eth :-} were added.
In previous versions these characters were combined from
other glyphs.

\item The \ethioplogo\ package may now be loaded
before or after \babel, as desired. Previous versions
required that \ethioplogo\ had to be loaded after
\babel.

\item Limited $\Omega$ support has been added.
\end{itemize}


\section{Copyright and Liability Notice}

This software is available under the GNU General Public License,
which you can find in the \texttt{COPYING} distributed with
\ethioplogo.

We distribute \ethioplogo\ in the hope that it will be useful,
but \emph{without any warranty}; without even the implied warranty of
merchantability or fitness for a particular purpose.

The authors of \ethioplogo\ are \emph{not liable} for any direct, 
indirect, incidential or consequential damage including, 
but not limited to, loss of data, loss of profits, or
system failure, which arises out of use or inability to
use \ethioplogo.
This clause does not apply to gross negligence or premeditation.

Although we do not require this as a part of our license,
we would be very happy if you send us any changes you make.


\section{Closing Remarks}

Much work has been done during the development of
this package, but much work had been done before and
there are still open points.

\subsection{To Do}

What remains to be done?

\begin{itemize}
\item Bugs must be removed. We know that there are
  some, but we do not know which. Please report bugs to
  \servicemail, we will be happy to make some updates.
\item More languages must be added. To do this, we must
  get to know the names for `bibliography', `chapter',
  `index' and so on for as many of the languages of
  Ethiopia as possible. Suggestions for these and other
  non-technical improvements can also be sent to
  \servicemail.
\item It must be possible to typeset whole books in the
  Ethiopian script. Although this is possible at the moment,
  there are several problems that occur with the
  different \LaTeX\ structures like \verb|\part|,
  \verb|\section|, or even \verb|\item|.
\item We need to find out more about the conventions for
  typesetting in the Ethiopian script.
\end{itemize}

\subsection{Thanks}

We would like to thank Abass B. Alameneh, Johannes L.
Braams, Donald E. Knuth, and Leslie Lamport for their
efforts. The programs provided by them made our
package possible.

Paul Seelig from the Debian team provided the neccessary 
motivation for finally releasing \ethioplogo\ under the GNU GPL.

Daniel Yacob gave encouraging and constructive
feedback on our project.
Michal Jerabek did some extensive field testing of our package.

While we are on the subject, here is a list containing
some of the names that appeared in this text.
You can already guess it, they are written in
the Ethiopian script. But which name is which?

\selectlanguage{ethiop}
\begin{center}
\begin{tabular}{ccc}
berhAnu bayana   & yOhAnes brAms   & mAnfrEd kudlEk \\
'abAs 'alamenahe & 'OlAf kumar     & lasli lampOrt \\
yO_ken me.sngar  & dOnAld knut     & dan'El yA`eqob \\
\end{tabular}
\end{center}
\selectlanguage{english}

\begin{thebibliography}{99}
\bibitem{Ale94} Alamneh, Abass: \EthTeX.\newline
\verb|ftp://ftp.dante.de/tex-archive/languages/ethiopia/ethtex/|

\bibitem{BeBoCoFe77} Bender, M.L.; Bowen, J.D.; Cooper,  R.L.;
Ferguson, C.A.: Language in Ethiopia. Oxford University Press, London 1997.

\bibitem{BKKM97} Beyene, Berhanu; Kudlek, Manfred; Kummer, Olaf;
Metzinger, Jochen: The \ethioplogo\ package. Fachbereich Informatik,
Universit\"at Hamburg, 1997.\newline
\verb|ftp://ftp.dante.de/tex-archive/languages/ethiopia/ethiop/|

\bibitem{Bra97} Braams, Johannes L.: The \babel\ package.\newline
\verb|ftp://ftp.dante.de/tex-archive/macros/latex/packages/babel/|

\bibitem{ethio97} EthiO Systems: \EthTeX.\newline
\verb|http://www.neosoft.com/~ethiosys/ethtex/ethtex.html|

\bibitem{ethno96}  Summer Institute of Linguistics: Ethnologue.
Dallas, Texas, 1996.\newline
\verb|http://www.sil.org/ethnologue/|

\bibitem{Gui01} Guidi, Ignazio: Vocabolario Amaharico-Italiano.
Casa Editrice Italiana, Roma, 1901.

\bibitem{Ham73} Hammerschmidt, Ernst: \"Athiopische Handschriften
vom \d T\=an\=asee 1. Franz Steiner Verlag, Wiesbaden, 1973.

\bibitem{Knu86a} Knuth, Donald E.: The \TeX book. Addison Wesley,
Reading, Massachusetts, 1986.

\bibitem{Knu86b} Knuth, Donald E.: The \MF book. Addison Wesley,
Reading, Massachusetts, 1986.

\bibitem{Lam86} Lamport, Leslie: \LaTeX~--
A Document Preparation System. Addison Wesley,
Reading, Massachusetts, 1986.

\bibitem{Les65} Leslau, Wolf: Ethiopians Speak, Studies in 
Cultural Background, Vol. 2. University of California Press, Berkeley 1965.

\bibitem{Les87} Leslau, Wolf: Comparative Dictionary of Ge'ez
(Classical Ethiopic). Otto Harrassowitz, Wiesbaden, 1987.

\bibitem{LiHo62} Littmann, Enno; H\"ofner, Maria: W\"orterbuch der
Tigr\=e-Sprache. Franz Steiner Verlag, Wiesbaden, 1962.

\bibitem{SERA93} The Unicode Technical Report \#1~--
Draft Proposal on Ethiopian Script. Houston, TX, August 1993.

\bibitem{Wedekind95} Wedekind Ch.; Wedekind K.:
A Survey of Awngi. In:  Survey of Little-known Languages of Ethiopia,
Linguistic Report No.~28. Institute of Ethiopian Studies,
Addis Ababa University. May 1995.

\bibitem{woldeki66} \selectlanguage{ethiop}'aklila berhan walda
kirkos:= sela g`eznA 'amAr~nA quAnquA tArik:: negd mAtemiyA bet,
'adis 'ababA 1958::\selectlanguage{english}

\bibitem{TWol70}\selectlanguage{ethiop}takla wald dastA:= 
ya'amAr~nA mazgaba qAlAt:: 'artistik mAtamiyA bEt, 'adis 'ababA 1970::
\selectlanguage{english}

\bibitem{sera97} Yaqob, Daniel: SERA FAQ.\newline
\verb|http://www.cs.indiana.edu/hyplan/dmulholl/fidel/sera.html|
\end{thebibliography}

\end{document}