[11934] in Public-Access_Computer_Systems_Forum
FYI France
daemon@ATHENA.MIT.EDU (Public-Access Computer Systems For)
Fri Jan 16 20:02:31 1998
Date: Fri, 16 Jan 1998 17:55:24 -0600
From: Public-Access Computer Systems Forum <LIBPACS@UHUPVM1.UH.EDU>
To: PACS-L@LISTSERV.UH.EDU
Reply-To: Public-Access Computer Systems Forum <PACS-L@LISTSERV.UH.EDU>
Date: Thu, 15 Jan 1998 17:48:29 -0800 (PST)
From: Jack Kessler <kessler@well.com>
Subject: FYI France: "digital libraries" in France, the BnF's Gallica
FYI France: "digital libraries" in France, the BnF's Gallica
-- "digital library" entre guillemets, because of the vagaries of this
still - very - imprecise term: does it cover just "libraries which are
digital", or "things digital which are libraries", or both, or neither?
The Bibliothe`que nationale de France / "Gallica" solution to this leading=
=20
Internet conundrum, now online via,
=09http://www.fyifrance.com/restricted/fyi2dlib.htm#GALLICA
=09http://www.bnf.fr
=09http://gallica.bnf.fr
is not the only "digital library", but it is impressive and it _is_
French. "Foreign" -- non - "anglo - saxon", non - "English / ame'ricain"=20
-- Internet solutions no longer are unique, but they are comparatively
rare, still. The online world swims now, and occasionally drowns, in
"digital libraries" mounted in the US, Canada, the UK. The BnF's approach
deserves attention, at least because it is "foreign".=20
In addition, though, the BnF has made a remarkable achievement in itself
in Gallica. The achievement certainly stands on its own -- "foreign" or
not -- against the many other projects currently labeled "digital
libraries". Gallica is the beginning of a well - planned and ambitious
effort, to document French culture by bringing a practical maximum of what
is available in digital formats to general public users via the Internet.
It is a remarkable culture, and a remarkable effort interesting both=20
for being "foreign" and for its own sake.=20
The BnF Gallica site -- quick tour
At Gallica a user now can see, online, a large and growing selection of
the digitized documents collections of the Bibliothe`que nationale de
France: "monographs, dictionaries, periodicals... from the simple page of
poetry to collections containing over a thousand pages, from 16mo to
quarto, from the popular press to bibliophiles' editions" -- it is the
BnF's intention to make Gallica, "a laboratory for the evaluation of
access to and distance consultation of digital documents".
The "collection" which may be reached online eventually is to include
100,000 digitized printed volumes -- containing 30 million pages, for now
supplemented by extracts from 250 volumes in the "Frantext" database of
the CNRS / INALF -- and 300,000 digital images. (Frantext --
http://www.ciril.fr/~mastina/FRANTEXT -- is the source for the ARTFL
database -- http://humanities.uchicago.edu/ARTFL/ARTFL.html -- the two
offer the same data but use different search command structures)
The general subject - concentration of the current Gallica effort very
interestingly -- see below, about "preservation" -- is the French 19th
century. Introductory texts are provided for that era's "history",
"politics", "law", "economics", "social science", "literature",
"philosophy", "history of science", and for each of four special
collections, "Euge'ne Atget", "Pierre Loti", "Ecole nationale des Ponts et
Chausse'es" and "livres illustre's". These introductory texts are
excellent: they are erudite, brief, and interestingly illustrated, and
they offer both internal links to materials found in Gallica itself and
external links to the many sites listed in their footnotes.=20
Multiple search criteria are available in Gallica: author, title words,
publisher, date, subject, and the fulltext of the catalog entries and
image captions. A 19th chronology has been mounted on the site: searching
may be done directly from the chronology -- author name entries in the
chronology are linked. A detailed "subject" list is available for
searching, as well, offering 43 categories from which to choose (these do
not, however, appear to correspond to the criteria used for the general
search criteria mentioned above -- "histoire du monde ancien" finds 8
entries using the latter, only 4 using the former). Searching also may be
done from online alphabetical authors' names and periodicals' names lists.=
=20
Gallica's digital images come half from the BnF's departments -- Estampes
et Photographie, Manuscrits, Arts du spectacle, Monnaies et me'dailles,
and others -- and half from other museums, libraries, and organizations
such as La Documentation Franc,aise, l'Ecole nationale des Ponts et
Chausse'es, l'Institut Pasteur, l'Observatoire de Paris, Magnum, l'Agence
France Presse.
The technical spec's for Gallica are impressive, as they are for the BnF
generally. About 2500 non - OCR / image - only monograph and periodicals
texts currently are loaded: these were digitized at either 300 or 400 dpi
in TIFF and CCIT group 4 compressed. 300 Frantext texts are loaded, which
can be searched using "Frantext" commands. 7000 photographs, at either
2000x3000 or 1000x1500 in video resolution -- these must have been slides?
-- and JPEG 10 compression, now are online.=20
Gallica's server, supplied by Sequent, is the latter's new "NUMA-Q 2000"=20
model ("Non-Uniform Memory Access" -- US $250,000, but the very latest
thing in architectures -- parallel processing!, vs. "SMP" -- the approach
which Sequent showed at Oracle OpenWorld in 1996), using a "System
NTX2000" running 200 mhz Pentium Pro Quadra processors, Windows NT Server
4.0 and Information Server 3.0, and Oracle, with over 60 gig storage.=20
This is all part of a general Sequent program to supply the BnF with 12
servers -- 2 of which are these high - end "NUMA-Q 2000's" -- for a total
of over 100 processors and 1.4 terabytes of storage! (Someone please
correct me if I have any of this wrong -- we're translating techspeak from
French into English, here, and it's hard enough just in English -- but
these are primarily the numbers which Sequent itself gives at the Gallica
site and on its own site.)
Gallica as a "digital library" -- general questions
Several general questions can be asked of Gallica, then, as they might
be asked of any project calling itself a "digital library":=20
1) compromises. What compromises, of quality and quantity and selection,
have been made -- as they inevitably must -- to achieve a "practical"
solution, and how are criteria to be reviewed and changed periodically as
techniques and circumstances evolve?
For example, the BnF has scanned in text here at low resolutions, without
OCR, and already as been criticized heavily in the French library press
for doing so. They also have faced the terrible questions of "selection",
which to critics like Umberto Eco are so central to the digital decade and
this century -- see FYI France ejournal issues of February and June 1993
-- and these BnF librarians have made their choices. And nevertheless --
unlike most of their critics -- they have produced a product and mounted a
useful and attractive service. Gallica could serve at least as an object
lesson in "the art of the possible" for digital libraries.
2) teaching. Digital information just has entered the era of the "un -
interested user": the user who is not stupid but does not have time to, or
just does not, care about computers and software and systems. It is a
giant learning situation now: for inventors, developers, promoters and
users, certainly, but also increasingly now for everyone else. Does this
particular "digital library" effort help? Is it designed deliberately to
be a teaching tool, and does it do this well?
Remarkably, for a culture which has a reputation -- abroad certainly, but
also among its own citizens -- for callousness with respect to the general
public, this French "digital library" project greets users with an
elaborate "feedback" page. Few other "digital library" projects elsewhere=
=20
have thought to bother.
3) scaling up -- statics. Technical criteria for "digital library" success
seem easily satisfied by the BnF, if not entirely yet by Gallica.
* Multilingual access is a general BnF priority: its main Website screens
are available in English. It would be good to see non - French language
access to Gallica someday -- in the meantime critics might consider how
many other "digital library" projects of comparable size provide
multilingual screens and searching?=20
* The politics of the BnF's development of their system with US and other
non - French vendors, such as Sequent, would make a fascinating study. One
wonders how much of the assembly and operating manuals are available in
the customer's French?
* Technical standards at the BnF generally and in Gallica in particular
would be interesting to analyze. Will Sequent's "NUMA" scale up, for
example -- or perhaps, rather, without it would Gallica or any other
similar project ever be able to scale up at all -- ditto for "Oracle" and
of course "NT" -- and are we gaining flexiblity and loosening things up,
or are we simply building the next generation's dinosaurs?=20
And what about the data, and the metadata?: are BnF cataloging norms
keeping up -- all those digitized texts and images -- and are things being
described so they may be identified, easily, several hardware / software /
systems generations from now?=20
Is the cataloging and indexing in fact being done? Is this a digital
library "ad hoc", or "in futuro"? Much of the historical record which we
have today of the French Middle Ages consists of nothing more than lists of
books, the inventories of medieval collections, the works which they
describe long ago having been damaged by water, eaten by bugs, burned,
borrowed by a French king or Pre'sident and never returned, or "pillaged
by the English" (the earliest Royal Library, according to the French)...=20
* The role of the commercial world, finally, needs examination in France
as it does anywhere else in the "digital libraries" world now. The cost of
what Gallica and the BnF have achieved so far has been enormous. If
Gallica is a project which can be financed by the French state, as an
important part of the national "patrimoine culturel", how many other
similarly - expensive projects might have to go to commercial sources --
and postures -- for their funds?=20
A good cost - accounting, and hard thinking about financial resources and
commercialism, seem to be necessary but much - ignored preludes to any
"digital library" project -- certainly if the idea on the laptop ever is
to scale up to something of the quality and size of Gallica.=20
(see my "Internet Digital Libraries: the International Dimension", Boston
& London : Artech House, 1996, ISBN 0-89006-875-5 -- Part III develops
these "digital library" criteria)
4) scaling up -- dynamics. Gallica does not appear to have faced, yet,
several of the most difficult problems which the rapid rate of change in
the digital revolution will cause for it, however -- nor have other
"digital libraries" projects, really, although a few of the best thinkers
in the US NSF projects have worried greatly over them (Stanford's Terry
Winograd -- see ref. supra ch. 14).=20
* How to "institutionalize change", in a project like this, for a
technology which replaces itself now every four months? Must the BnF
undertake a complete revision of all of its hardware, software, systems
and personnel on a quarterly basis, just to keep up? Non - hi - tech
commercial firms barely manage this sort of broad organizational review
every decade; every century, if they are lucky, for government departments
and institutions like universities -- Yale's Bart Giamatti used to muse
about his "20th century institution run on 15th century principles" --
and still longer for most archives and libraries.
* How to integrate an effort like Gallica with other similar projects, and
with resources which are not so similar? Sure, you can put a link to it on
your Website, but does that really exhaust the store of the world's
inter - woven knowledge on that "digital library's" subject?=20
Technical standards -- mentioned above -- will help a "digital library"=20
project like Gallica to work together with other digital resources.=20
Wouldn't it be nice, though, for a student in Australia one day to be able
to reach all sorts of printed and photographic and cinematic and even
_sound_ evidence about Atget or Loti, physically located in all sorts of
places, from -- to her -- a single source? Gallica's own immediate
"context" are the magnificent surroundings of one of the world's greatest
research libraries, and one of the Internet's better general "digital
library" Websites -- http://www.bnf.fr -- but not all "digital library"
projects are or will be so - blessed.=20
* Preservation. Most interestingly, for the Gallica project, the era which
they have chosen to work on is precisely the era most threatened currently
by the "acid paper" preservation problem.=20
Anything on paper from the 1830s to the 1970s -- books, periodicals,
business records, personal letters, anything -- now is turning yellow and
brittle and crumbling. Cotton rag paper from prior periods is not so
threatened, and from the 1980s on we will have digital information on a
variety of media.=20
But for the intervening century - and - a - half "Umberto Eco's nightmare"
is being realized: the 22nd century may have no evidence to examine of the
19th and the early 20th. Before the cynics say "just as well", consider
that Gallica's choice of the 19th century as the era on which to
concentrate -- a choice made for several reasons -- may prove=20
ultimately to have been one of the best things about the project.=20
5) Books and bytes, bricks and digits -- does a "digital library" still
need a "building"?
Gallica probably has been expensive. The BnF's new Bibliothe`que Franc,ois
Mitterrand building at Tolbiac certainly has been. Do we need both? If a
"digital library" like Gallica can exist out in the aether, in Cyberspace,
and eventually can satisfy most of our "information" needs, do we still nee=
d
to construct and maintain expensive buildings, like the BFM?
Predictions have been made of the death of books, of paper, of reading,
and of libraries -- also of education, of culture, of intelligence -- all
at the hands of the revolution in digital information: ask any parent, in
France or the US or elsewhere, who worries today about online pornography
and video games. The new techniques make many threats.
Such threats are not taken seriously by many in the US. But France --
unlike the US -- is a place which has known, and suffered, through many
"transitions in media" and cultural and political upheavals, throughout
its longer and more violent history, transitions and upheavals which have
destroyed "documents". For all the riches in the current collections of
the BnF and other repositories of the national "patrimoine culturel",
there is much more that is missing from the French historical record.=20
This marks a fundamental difference between the perception of the risks
involved in "transitions in media" as viewed in Omaha versus Lyon -- or
foor that matter in any US location compared with Sarajevo, or Hiroshima,
or Hanoi or Berlin or Moscow or anywhere else in the world which has both
recent and longstanding memories of severe cultural loss -- outside the
US, there is real and justifiable fear involved. The question of cultural
loss from an accidental or purposeful loss of the BnF books is a very
serious one in France: budgetary amounts for buildings or "digital
library" projects there must be examined with this difference in mind.=20
Now leave aside for the moment, though, the question of whether "digital
libraries" which replace the buildings can happen or should, and assume
that they will -- would that even be a "Good Thing"?=20
There is a growing literature -- Arlie Hochschild et als. -- criticising
or at least implying criticism of the "workaholic / yuppie" generation
which has tried to mask neglect, of children and family life and to some
degree sanity, behind labels such as "flextime" and "telecommuting" and
"telework". Putting in longer hours "at the office" does not resolve the
conflicts at home. Bringing work home -- via "telecommuting" and
"telework" -- does no better, crowding "family life" nearly as much as
"being away at the office" does.
Some sort of "neutral place" is needed -- for work, for the voracious
appetite for hours exhibited by all of these new "digital information"
techniques: not the old office, for that increasingly is too far away from
home, in the Paris and San Francisco of the 2 - hour - commute - 1990s --
but not the home either, if sanity and family life are to be held together
in the Age of the Internet.
So perhaps the local library? Perhaps a neutral, congenial, comfortable
place, not at home but physically near it, a place in which to take up
telework and telecommuting and Internet surfing of sites like Gallica, all
without intruding on family time?=20
And why should that space not, on occasion, be magnificent and monumental,
like the new BFM? Was monumentality any more necessary for printed books
than it might be sometimes now for the Internet?=20
Perhaps the growing need for "neutral space" should be separated from
questions of the monumental design of that space: we may or may not need
the latter, but an information society of "digital libraries" like Gallica
rapidly may be developing a tremendous and widespread need for the former.=
=20
-- something else to think about, anyway, as you tour the Gallica site.
(For more on this topic, see Michel Melot's introduction to "Nouvelles
Alexandries : les grands chantiers de bibliothe`ques dans le monde", Paris
: Editions du Cercle de la Librairie, 1996, ISBN 2765406197; also my "The
Biblioth=E8que Nationale de France", in "Rational Space: Library Buildings
for the 21st Century", T.D. Webb, editor, Jefferson, North Carolina :
McFarland & Co., [1998 forthcoming -- book title is provisional]).=20
Samuel Beckett was a hopeful poet; and he wrote in French, in France. The
future which he feared is portrayed in his plays: not too many libraries,
or books, or information of any kind at all -- only a few lost and lonely
people, with nothing to do and nowhere to go to do it.=20
Beckett's nightmares were filled with loss: the loss of definition, the
loss of meaning, the loss of any "sense of place". His characters Hamm and
Clove spend their lives, together and apart, with little else to do,
looking and waiting for "Godot".=20
The French have recent and poignant memories of their own which match
Beckett's nightmares, and a long history punctuated with the very
disasters which he feared: "bouleverser" is a term which has much meaning
in French political history, but is nearly un - translateable in the
American context.=20
The BnF's sometimes - maligned new Bibliothe`que Franc,ois Mitterrand
building, and now its even newer Gallica online "digital library" exhibit
of French culture, perhaps are two of the latest and best attempts to
provide France and the rest of the 21st century with something very
worthwhile to do, and an interesting and impressive "place" in Paris in
which to do it.=20
Note1: "official" -- these are from Annexe 6 of the new "Internet: Enjeux
juridiques / Rapport au ministre de'le'gue' a` la Poste, aux
Te'le'communications et a` l'Espace et au ministre de la Culture / Mission
Interministe'rielle sur l'Internet pre'side'e par Isabelle
Falque-Pierrotin" (Paris : La documentation Franc,aise, 1997) Collection
des rapports officiels. ISBN 2-11-003756-3 ISSN 0981-3764.
Note2: So don't blame me -- I just report these things -- and anyway this
list is "officiel", so presumably you _have_ to know about it...
Bonne anne'e...
=09=09=09=09--oOo--
FYI France (sm)(tm) e-journal ISSN 1071 - 5916 =20
*
| FYI France (sm)(tm) is a monthly electronic journal,
| published since 1992 as a small - scale, personal,
| experiment, in the creation of large - scale
| "information overload", by Jack Kessler. Any material
/ \ written by me which appears in FYI France may be
----- copied and used by anyone for any good purpose, so
// \\ long as, a) they give me credit and show my e - mail
--------- address and, b) it isn't going to make them money: if
// \\ if it is going to make them money, they must get my
permission in advance, and share some of the money which
they get with me. Use of material written by others requires their
permission. FYI France archives are at http://infolib.berkeley.edu (search
fyirance), or http://www.cru.fr/listes/biblio-fr@cru.fr/ (BIBLIO-FR
econference archive), or at http://www.fyifrance.com , or at
http://listserv.uh.edu/archives/pacs-l.html . Suggestions, reactions,
criticisms, praise, and poison-pen letters all will be gratefully received
at kessler@well.sf.ca.us .
Copyright 1992- by Jack Kessler, all rights reserved. =
=20
