Newsgroups: sci.classics,alt.etext
From: perseant@stein2.u.washington.edu (Konrad Schr{\o}der)
Subject: project Libellus [followup announcement]
Message-ID: <25edmd$eap@news.u.washington.edu>
Followup-To: sci.classics
Date: 25 Aug 1993 01:02:05 GMT
Organization: University of Washington
Summary: Project Libellus announces itself again; corrections & additions.
Keywords: Libellus Vergil Latin Classics FTP
Lines: 168

[N.b.: this was cross-posted to alt.etext from sci.classics.  Followups
are directed to sci.classics.  --ks]


                           =PROJECT=LIBELLUS=
                 Classical Latin texts on the Internet

As you may remember, a few months ago an archive of classical texts
was announced at the University of Washington, Seattle.  The project has
since acquired a name; moved its files; and generally been busy correcting
errors and working on new offerings.  This document is intended to bring
knowledge of our project up to date, as well as answer any questions that
one might have about it.  

=Purpose=
Whereas most vendors of classical materials place restrictions on
their use, either by placing them under copyright or by the application of
a licensing agreement, or both, the purpose of project Libellus is to make
available, *for*any*use*whatever,* classical texts and associated
materials in electronic form.  We are not attempting to create new works,
but only to make available those which already exist.  
    It should be made clear, however, that since we cannot (yet) guarantee
that our e-texts are accurate reflections of the original printed editions
they should not (yet) be used for purposes in which adherence to a strict
standard is required.  (See the section on =format=, below.) 

=Location=
Since things have been moved around, the location of the files needs
to be clarified.  The host name is 

                          ftp.u.washington.edu

    If the archie server (or anyone else, for that matter) tells you that
the hostname is ``milton.u.washington.edu'', or simply
``u.washington.edu'', the information is out of date<1>.  The directory is 

                            /public/libellus
or 
                      /pub/user-supported/libellus

    Log in as ``anonymous'' or ``ftp'', and send your e-address as
password, as you normally would.  
    You may be able to get the libellus materials from a site nearer
yourself; if you do so, be aware that those sites may not have the
``freshest'' version of those materials.  If you are running such a mirror
site, please send us a note (see below as to where to send it) so that we
can inform you should significant changes be made to the archive.  

=Texts=
We now have the entire _Aeneid_, _Georgics_, and _Eclogues_ of
Vergil; Caesar's _De_Bello_Gallico_, book I (with commentary); Livy's
_Ab_Urbe_Condita_, book I; and a few selections from Catullus.  Work is
being done here on the rest of the _De_Bello_Gallico_, and Allen &
Greenough's _New_Latin_Grammar_; we have been offerred Cicero's orations
_In_Catilinam_ and a Latin--English dictionary, and hope to have those
reasonably soon.  

=Greek=
Until very recently, our scanning software was not able to recognize
Greek fonts; therefore, there are no Greek texts in our archive.  We are
in the process of testing out some new scanning software that purports to
be able to learn any font.  If that works out, Greek texts should begin to
appear shortly.  

=Format=
A word about orthography:  since we are attempting to reproduce
individual printed editions, the orthography of the edition may not be
what one is used to; we do not impose `i' over `j', for example, or `u'
over `v'.  As a more interesting example, in the Vergil texts, the letters
`o' and `u' are often confused, yielding such words as `volnus' for
`uulnus' and `divom' for `diuum'.  These constructions have been
preserved, not because they are preferable, but because they agree with
the printed original.  
    Because of this, it should be obvious that *this*particular*text* is
not fit for purposes requiring adherence to strict critical
standards---regardless of the multitude or paucity of errors in the
e-text.  Other texts, e.g., the _Ab_Urbe_Condita_, were taken from more
reputable sources, in this case the (old) Oxford Classical Text; in this
case, if the text has come through to our archive with no errors, it
should be fit for almost any purpose.  It should be worth noting, too,
that the more people use the texts, the more errors are likely to be
found; since we will only modify the text in accordance with the printed
original, this along with careful proofing to begin with should in a
reasonably short time result in a pure e-text.  
    Alternatively, the possession of a less-than-perfect e-text, such as
our Vergil text, should make proofreading a more reputable text much
easier, once preliminary garbage collection has been done---simply diff
the two texts, and follow the printed copy on all differences.  
    All materials that we provide are in TeX format;<2> this means that
they are almost as readable as plain ASCII documents, but can be fed to
the TeX computer typesetting system to provide professional-quality
output.  

=Utilities=
If you do not have TeX on your machine, or are interested in using
the Libellus materials as electronic rather than printed documents, the
TeX directives can be stripped off, by hand or by automaton, to provide a
straight ASCII document.  Up to now, doing this ``by hand'' or with a
wordprocessor's search-and-replace function has been relatively easy,
since the ancient texts are particularly simple in their format; with the
arrival of more complicated texts, however, this may not be so easy.  
    With this in mind, we provide a conversion program, tex2asc, which
will attempt to convert TeX documents into straight ASCII, ANSI, or RTF
format.  The original version of this program (0.1) would only convert to
straight ASCII, and did not handle any but a very few constructions.  I
have re-written it to handle slightly more sophisticated constructions; in
particular, emphasis and footnotes are supported; multinational characters
are rendered as well as possible; diacritical and some critical notation
(a la _Textkritik_) is supported, and automatic line- and
section-numbering are available approximately as they would appear in the
TeX output.  We have binaries for DOS and VMS machines---the program seems
to crash Macintoshes---and of course the source code is provided for
compilation on other platforms (notably Unix) or for modification if you
prefer.  If you are reading this announcement in electronic form, it was
generated by tex2asc.  
    In the course of correcting the materials, we have found certain other
tools to be useful; in particular we have rigged up a spell-checking
companion, lexp, that attempts to deduce the conjugation/declension of
words in a word-list, thereby greatly increasing the number of words that
the spellchecker has access to.  I recommend =not= attempting to use this
program on a desktop PC; but it is effective on a mainframe.  If you want
to add specific words to a dictionary, lexp will also act as a quick
declension/conjugation engine.  Currently this program is offerred as a
beta-test version; but wordlists generated by it are available in the
utils directory.  
    Any utilities held in our archive will be redistributable at no
cost---whether they be Public Domain, Freeware (may not be sold), GPL, or
what have you:  if you got it here, you can give it away.  The tex2asc and
lexp programs are under the GNU General Public License; see the file
COPYING in libellus/utils for more details.  

=What=can=I=do?=
If you are using our archive, please send a short note to
perseant@u.washington.edu, saying what the materials are being used for,
what you would like to see, etc.; this will give us a better idea of what
sorts of things we should put up in the future.  
    If you find errors in our texts (or commentaries or what have you),
please tell us! If what you have found is an error, rather than a variant
reading, it will be fixed within the week, if not within the day.  
    If you would like to help prepare documents, contact
perseant@u.washington.edu and volunteer for any part of document
preparation.  You do not need a scanner or above-average knowledge of
computers to do most of what we do; to proofread a text, for example, you
do not really need a computer at all.  
    If you have a text that you are going to contribute, please see the
README.copyright file in libellus/info before you begin working on it.  

=Where=can=I=get=more=information?=
There are various README files in the archive, in directory
libellus/info.  If what you want to know is not there, try sending e-mail
to perseant@u.washington.edu.  If you are reading this message on the
mailing list classics@uwavm.bitnet, you can also send mail to the list,
and we'll get it, so long as one of ``ftp'' or ``libellus'' is in the
subject line.  

                                                       Konrad Schroder


Footnotes:

    <1>The canonical hostname is ``stein2.u.washington.edu''; so if your
nameserver does not recognize ``ftp...'', try ``stein2...'' instead.  If,
however, the sysadmins decide to move the anonymous FTP space from stein2,
as they did from milton, the name ``ftp'' will move with it, and so in
that sense is preferable.

    <2>TeX, properly pronounced, rhymes with ``blecchhh''.  (It is not
t-e-x, but rather <tau>-<epsilon>-<chi>.)