WAIS-discussion digest #37: More on Public Libraries
Brewster Kahle
Forum On Wide Area Information Servers and Electronic Publishing
Brewster Kahle
Contents:
CNN Short piece on Thinking Machines and WAIS (Brewster Kahle)
Re: Electronic Publishing and Public Libraries (Steve Cisler)
Re: Steve Cisler's review (Brewster Kahle)
[There have been a couple of very good pieces spurred by the public library
essay. Very informative. This issue has Steve Cisler's response. More in
the next issue.
-brewster ]
----------------------------------------------------------------------
Date: Fri, 3 Jan 92 14:08:32 PST
From: Brewster Kahle <brew...@Think.COM>
Subject: CNN Short piece on Thinking Machines and WAIS
The official CNN broadcast of it's Thinking Machines segment will air this
weekend, January 4th and 5th: Saturday 11:00am - 11:30am and Sunday 12:30pm
- 1:00pm as part of its Headline News program. It should run about 5
minutes.
It has a very short clip on WAIS. I have not seen it yet.
-brewster
Brewster Kahle Thinking Machines Corporation
Brew...@Think.com 1010 El Camino Real
Project Leader Menlo Park, CA 94025
Wide Area Information Servers 415-329-9300x228
------------------------------
Date: Fri, 3 Jan 92 16:35:20 -0800
From: Steve Cisler <s...@apple.com>
Subject: Re: Electronic Publishing and Public Libraries
My background is in public libraries, though I work in a corporate
library and my job, like Brewster Kahle's, is to promote the use of our
technology in all sorts of libraries (and similar institutions). This is a
reply to his posting about Reading Rooms in public libraries.
About five years ago a consultant named Keith Henson called me up
with the idea he called "Gutenberg Express" and he wanted to know
what libraries could afford to participate in an on-demand printing
operation where the information was shipped weekly on digital platters
the size of laserdiscs in order to build an archive of electronic works for
library users to access. There would have been printing and binding
machines (similar to ones used for phone directories). This project did
not achieve lift-off. Brewster is working on a launching pad for
Connection Machines, but his target, the institution of the public
library, is somewhat different than he describes in his provocative
essay.
>Public libraries have served two clashing goals, to allow open access to
>published information and to archive written history for future scholars
>and posterity. This uneasy combination has been joined into a single
>institution --public libraries, both centralized and branch-- due to the
>technology of distribution of information, paper.
These goals are not clashing; they are just different. And they
certainly are not the only ones that define a public library. The
library is a place for other activities such as story hour for kids,
informative programs for teen-agers and adults, getting videos
and light reading matter for entertainment, meeting to work on
school reports, and meeting with your literacy tutor. Other people
are defining public libraries in ways that are stressing the system
and the staff. Working parents are dumping their kids off because it's
safe, warm, and a lot cheaper than day-care centers. This is more
common in the affluent areas--the very kind of suburb that might be
expected to support an idea like an electronic reading room. Homeless
are contending for space and resources with people whose main
goal is reading, browsing, or research. Some homeless share these
goals; others naturally want to keep warm, doze, wash in the bathrooms.
So the library is serving as far more than a cheap and friendly
information spigot for the community. This means to me that the
best that technological innovators can do is to fit their wonderful
tools and systems into the insitution as it exists, as it is used by
the populace that supports it. Proposing a print-on-demand service
as an overlay on exisiting services might make sense in some areas,
but the state and local budgets are so tight during this recession-
depression, that a project will only get off the ground with seed money,
equipment, and support from those proposing it, and probably not
from the state and local governments.
Books and magazines are
>both inexpensive and difficult to reproduce which encourages the wealthy to
>buy their own copy, if available, tends to protect publishers from the free
>use of libraries from seriously decreasing their sales. The archiving goal
>is not catastrophically sacrificed by risking loaning out books since the
>limited public use does not destroy the books. As electronic distribution
>(so called, electronic publishing) becomes common for different segments of
>written material, access and archiving become quite different, since making
>a copy is easy and inexpensive, and delivering a copy can be done without
>requiring a person to come into the library.
Books by mail is a service being offered in a number of places. While it
is only a fraction of the walk-in circulation, it is an important service for
the home-bound, people in rural areas (Ohio Valley Area Libraries has
a very successful service), and those not served by bookmobiles--
which are really sort of like mobile reading rooms and some are
outfitted with packet radio connections to central catalogs.
I do agree that many of the public library services are for the patient,
not the person in a hurry (the businessperson, the teenager who
waited until the last minute to do a report, the city official heading for
a meeting in the afternoon), and a speedier service would be welcome.
>The result of this technology change can be an exciting one where the
>public library system can refine its charter and serve the public in a
>widespread way that was not possible with paper. The unique aspects of
>libraries --service oriented staff, lack of profit motive, prevalent
>locations, and the role in schools-- can give them a more important role in
>the future than they ever had in the paper era. This essay will suggest a
>new model for the access goal of the public library, the Reading Room, that
>protects and promotes publishers, while serving the public in the tradition
>of the branch library.
I agree that the unique aspects of public libraries can serve them
well in the future when electronic networks are more widespread.
>
>Briefly, a "Reading Room" would offer patrons convenient access all
>published information in printed form or for screen display in many places
>in a town with the help of librarians, but I think the best way to explain
>the idea is to use the analogy with the american banking system transition
>from bank buildings to automatic teller machines.
>
>Old Banks Automatic Teller Machines
>--------- -------------------------
>Centrally located Everywhere
>Limited hours Available all the time
>Local information Global information
>Physical savings books Access cards
>Used rarely Constantly used
>
>Libraries Reading Rooms
>--------- -------------
>Centrally located Everywhere
>Limited hours Available all the time
>Local information Global information
>Physical library cards Access cards
> --but still similar--
>Friendly librarians Friendly librarians
>Comfortable reading chairs Comfortable reading chairs
>Current fiction Current fiction
>Easy browsing Easy browsing
>
>Reading rooms can become an active part of people's lives much as ATM's
>have transformed people's relationships with banks.
I spoke with a veteran bank teller the other day and asked her about
the effects of ATM's. She said it had really cut down on in-bank
traffic. She enjoyed the face-to-face contact (and was really helpful
to me), but she said there are a few new jobs as a result of the
ATMs: a couple of women are working alone in a room just
handling all the deposits brought in from a local network of ATMs.
She said it was really boring work. Let's hope libraries don't
duplicate this.
>A minimal design for a Reading Room would be a one room store-front that
>had a few comfortable chairs, a bookshelf with current fiction, several
>computers for browsing and reading, a printer and binder for printing
>copies for people, and a librarian during regular hours. Interlibrary loan
>would be used when a particular volume is needed. Therefore, these Reading
>Rooms would be used by the vast majority of the population, while the
>central libraries would be dominantly used by specialist and archivists.
>The rest of this essay will outline how this might work.
>... The widespread preference for online card
>catalogs has been shown by the large scale conversion from physical cards
>despite the costs. Furthermore, good card catalogs that are accessible
>remotely have become very popular, such as the University of California's
>Melvyl system that gets 1/3 of its requests from people not in the library
>and often from around the globe [Lynch 91].
Much of the conversion has been for cost reasons. Convincing the public
that it is better or superior to the paper catalog relies on skilled p.r. staff
who are not always very successful in making the public discard their old
ways. I still hear young and old alike wishing for the familiar card catalog.
There is still a long way to go before the catalogs are even accessible.
The concentration of online library catalogs in Silicon Valley is not repeated
around the country. Therefore, most librarians would want to get
their catalogs automated before they venture into an electronic reading
room project.
I spoke with an automation vendor a couple of days ago, and he said
that the market is very slow because of the recession/depression. Libraries
are really hurting, folks, except in a few lucky instances.
>
>Catalogs, however, do not replace the stacks for browsing. Computers have
>been making gains in browsing and serendipity as the screens get bigger and
>sharper, as full text becomes available, and as better searching systems
>are developed. Bellcorp found that serendipitous learning of journal
>information was enhanced by using a computer over paper [Egan 89].
>Experiments in Japan using computers to replicate the experience of looking
>at bookshelves has had encouraging results [Sato 90]. We still have a ways
>to go to improve browsing, but great strides are being made.
I've seen a video of the experiments from Japan (as have people at Library of
Congress), and it is very crude once you get past the image of a book
opening, or a Japanese scroll unfurling. No great stride in browsing there!
But yes, there are others over here that hold great promise.
...
>What about the economic issues of royalties? Since the Reading Room
>monitors all the books printed and viewed, it is easy to keep a record of
>this and reimburse the publishers. Bear in mind, that the current
>copyright laws are intended to ensure that authors, editors, and publishers
>get a fair stake from the use of their work. Reading Rooms can easily help
>in this process. In fact, Reading Rooms help a publisher greatly since it
>handles the distribution and retail handling and the printing of the works,
>so the publishers can lower their costs of putting out a new volume. This
>cost savings can cause a flourishing of works that would have a more
>specialized audience.
John Garrett of the Corp. for National Research Inititiatives will be
publishing a paper on copyright issue in a large digital library. This
should be available this month, and probably for ftp. I'll post info when
I find out.
>
>How about covering the costs of the library? The total cost of a public
>library are about $3-7 per book per year, if all the books in the library
>are totalled and divided by the total costs.
Cataloging a new book can cost as much as $110 (average amount at
Library of Congress) way on down to a buck or so for those who use
copy cataloging, but the whole processing costs are closed to $20.
Add the cost of the book to that fee. What Brewster may be saying is
that maintaining a collection will cost that. I don't have statistics to
verify that.
>Conservatively, public
>libraries cost a suburban family about $60 per year [check this]. Thus, if
>we were to give each family in a community a debit card worth with $60
>worth that can be used in any Reading Room, then every person would have
>every incentive to use that card to read, check out, and enjoy their
>library. If someone needed another card, then this can be awarded based on
>the librarians assessment of how the last one was used. Since not all
>families would use their full amount each year, this would guarantee that
>costs could be contained.
I looked up some recent California statistics on materials costs:
Sacramento City/Co Library 90-91 income: $16 million serving 1 million
So that's about $16 per person or about $50 for a three person family.
Here are a few other figures to show the varied spectrum of support that
libraries
have. Sort of a library version of J. Kozol's 'savage inequalities'.
Expenditure per capita: 1988 figures
L.A. $16.77
Santa Clara Co $30.00
Butte Co. (rural north) $4.93
Beverly Hills $87.46
Menlo Park $27.63
San Francisco $25.52
Shasta Co. $2.51
>A Reading Room could be very inexpensive to build and operate assuming a
>generally thrifty approach. Two computers, a laser printer, comfortable
>chairs, and bookshelves could be $25k in investment, a well paid librarian
>would be another $50k per year, $4k for rent, leading to well under a $100k
>outlay every year. Since many medium city libraries have a $3-7 million
>dollar budget, this money could lead to 30 to 70 Reading Rooms distributed
>in communities for the same costs. This is not to say that this is an
>exact financial statement, but rather an estimate showing that this is
>feasible.
In another part of the essay the Reading Room is said to be available
anytime. That means you'd have to more than triple the staffing
because few people are going to want to work graveyard shift at the same
pay as the 8-5 person. If you limit access to 16 hours a day, you will
still have to pay extra for the swing shift librarian. $50K per year is more
than you'd probably have to pay for one librarian in many towns
I don't think your hardware costs are realistic for a print-on-demand
setup that will pump out a five hundred volumes a day. I pulled that
figure from my hat by assuming that a small branch circulates about
30,000 items a month (that varies tremendously) and that half
those items would be online and available electronically.
(we all know that's a looong way off). Xerox Docutech systems are
very large, very expensive but could meet the demands of a queue
of disparate users at the computer catalogs and browsing stations.
Current pricing relegates them to big institutions (Cornell, Harvard,
Apple). Because high speed printer/binders are expensive they
are centralized, and their products would have to be distributed in
standard ways: mail, delivery van, or picked up by the requestor.
I guess you can say that in X years there will be affordable printers,
but right now there are not.
The oddest part is that the costs of the network to support this are
not mentioned. Perhaps this will be absorbed by the publishers and
entrepreneurs who sell the sytem to the library. If Mitch Kapor is
right, we could all have basic ISDN at no extra cost on our phone bills,
and that would certainly be a boon for library connectivity to the
electronic information infrastructure that Brewster and others are
helping to build. At present having dedicated lines is getting cheaper,
but it is still a signicant cost.
There would also have to be additional support staff for the hardware,
and this might be library employees, the information utility, or a
combination of both.
>Might it happen? I dont know. Reassessing our basic institutions in the
>light of new technology is not something governments do very well.
I don't think companies, the institutions themselves, or most
intellectuals do this very well either.
I highly recommend the current issue of WHOLE EARTH REVIEW on
questioning technology as a theme as well as Robert Bellah's new
book, THE GOOD SOCIETY, wherein he and his co-authors look
at the role of institutions in our society.
Steve Cisler
Apple Library
s...@apple.com
------------------------------
Date: Fri, 3 Jan 92 17:43:46 PST
From: Brewster Kahle <brew...@Think.COM>
Subject: Re: Steve Cisler's review
Outstanding, Steve, thank you. You have a calm and informative style even
when you disagree.
Your statistics are very helpful and discouraging for the idea, especially
the circulation stat. 30k books/month or 300k books/year. Say there are
10 reading rooms for that branch library, makes 30kbooks/year. At 150
pages/book and 4 pages/8.5x11page at $.05/page yeilds: $1.85/book or $55k
per year printing costs. Not cheap. Also 100books/day or 4000pages/day is
about 10pages/minute all day. I assumed you could use multiple reinforced
laser printers rather than docutext systems.
I wonder, if you saved the books that you printed and redistributed them
(this would mean that you would have to print the whole thing the first
time, as well as administer them when they come back...), how many books
you would actually print. In computers, caches work extremely well; I
wonder about it in this case. Are 90% of the active books, 10% of the
library? In which case, keeping the 10% printed might be a way to go.
I really dont know the answers on this type of thing, but that was the
assumption of having a few bookshelves of popular material.
I hear that popular fiction is a hunk of libraries business, so there might
be alot of reuse of printed material.
Another approach is to only print the part of the books you need. What
percentage of the books taken out are actually read? If you could count on
getting it at any time, then hoarding, even on a small scale, would not be
helpful. Would this cut down the number of pages taken out?
Another approach is a more hybrid scheme with large scale inter-library
loan (ILL). I have read studies that say the full costs of an ILL
transaction is about $10. This is too high. If it can be made less than
printing a new copy ($1.85), while keeping the delay to a reasonable
time such as one day, then maybe this can be used to get the works to
people.
You did pick up one of my bugs: the available all the time issue. The only
way this could work after hours would be via dialup from user's homes and
picking up the books the next day if any were printed. This has obvious
problems.
Network costs, I suggest, will be low. Unlimited use 56kb internet feeds
these days are about $1k/month and the price will probably drop.
May I reprint this in the next wais-di...@think.com?
Thank you again.
-brewster
------------------------------
End of WAIS-discussion Digest 37
************************
-------