Cites & Insights: Crawford at Large
ISSN 1534-0937
Libraries · Policy · Technology · Media

Selection from Cites & Insights 5, Number 11: October 2005


Future Libraries: Dreams, Madness & Reality, 10 Years Later

It hasn’t really been 10 years. Future Libraries: Dreams, Madness & Reality (henceforth FLDMR, since there’s an entirely different Future Libraries) was published in January 1995. The book was written and edited during 1994, with some work in 1993. I sent the final camera-ready pages to ALA Editions in the fall of 1994. In terms of content, it’s closer to 11 years.

The book’s still in print. Although there was brief discussion a couple of years ago of doing a second edition or follow-up, that’s not likely to happen unless one author takes it on and gets permission from the other author.

To be honest, I haven’t really looked at FLDMR since 1998, when I was writing Being Analog. I rashly promised to take a hard, honest look at FLDMR a decade later. I’ve been putting it off for most of the year. Finally, I started going through the book looking for high points and flat-out errors, as well as places where I’d say it a lot differently now. I did it the easy way: one chapter a day. (Yes, the book’s an easy read. Yes, the book’s short enough to read in one long sitting. But it’s summer, I’m lazy, and old episodes of Moonlighting, Remington Steele, and Gilmore Girls—and, for that matter, The Greatest American Hero—beckon in the evening.)

Here’s what I found, trying my damnedest to be a critical reader rather than a coauthor.

Chapter by Chapter

Going through the first chapter, I found myself checking off pages where I’d say pretty much the same thing we said back then. One exception might be this paragraph (p. 10):

Looked at objectively, the relative roles of electronic communication and non-electronic communication (print, sound recordings, film/video, etc.) become clear. Electronic methods are best for “housekeeping” and for giving access to data and small, discrete packets of textual, numeric, and visual information (such as those found in many reference works). Each of the other media has areas in which it is the best. In particular, print-on-paper is and will be the preeminent medium for the communication of cumulative knowledge.

I’ll stand by the last sentence, but the digital/analog split has become fuzzy over time. Specifically, the digital realm as just-in-time distribution method for medium-length narrative in the form of journal and magazine articles has proven far more important to libraries than we could have guessed in 1994. Some would claim that most readers read those journal articles on screen. I believe many students skim through articles in electronic form to find chunks to cut and paste, and that they may glean reasonably good understanding of the sense of the articles. For all I know, maybe KTD really are different and do gain full comprehension from the screen while multitasking up a storm, although I’m still not convinced.

2. The Life of Print

Here’s a mixed case. Some magazines have fallen on hard times, although the magazine industry is still healthy. Of examples on page 17, PC Magazine still circulates more than a million copies but of a much less thick magazine and PC/Computing has vanished.

The discussion of “Appropriate technology” (18-19) is dated—but less so than you’d expect after a decade. There are readable electronic displays that don’t use transmitted light, but so far they’re not all that wonderful. Resolution has improved in some cases: Some notebooks and PDAs have displays with 150 to 200dpi resolution—but that’s an improvement from 100dpi or so a decade ago, and you only see ultra-high resolution on very small displays. At this rate, who knows when (or if) we’ll get to 300-600dpi displays at least as large as a book page? Meanwhile, almost all home laser printers now offer 600dpi printing or better, as do most inkjet printers.

An anecdote on p. 24 has been a source of misunderstanding, as my coauthor used it as an argument against the worth of Google Print. I still maintain that good nonfiction books should inform well enough that you’re better able to understand p. 154 after you’ve read pages 1-153—but that does not negate the worth of seeing a snippet of p. 154 as a clue to the usefulness of a book. If “random accumulations of opinion, disconnected data, unverified assertions, and contextless statements” was my writing, I apologize: While that’s a fair description of millions of web pages, there are also millions that provide clear worth.

My discussion of the startup costs of book printing and CD-ROM production was right then but is no longer quite true. With PoD technology, startup costs for a short-run book can be nearly zero. Similarly, short-run CD-ROM and audio CD production is now CD-R production, and there are no real startup costs other than getting the content and organization right.

We discussed PC Magazine and its CD-ROM version and some other examples of magazine economics. As already noted, PC Magazine issues don’t “average 500 pages” these days (150 pages is closer to the mark)—and the CD-ROM version has disappeared, I believe. PC Week seems to have disappeared into the digital mists.

On later pages, discussing the Daily Me (a concept that still disturbs me), we overestimated electronic distribution costs. For one reason or another, much of this stuff has stayed “free” (well, $20 to $50 a month…) although various business pundits still see most professional content disappearing behind subscription or pay-per-article walls.

3. The Madness of Technolust

What did we get wrong here? In a footnote on p.37, we thought that Apple’s Newton might “succeed quite nicely” in some niche markets even after failing as a mass-market technology. Oops.

We were correct in believing that CD-I and CD-V wouldn’t explode onto the marketplace. The “2.2 computers in every household by 1999” projection is still an absurd claim in 2005. “Digital convergence” hasn’t happened—but we did miss the “explosion” of DVD (faster than most new media, but still a slow explosion).

Looking at dollar figures on p. 40, it’s worth noting that LCDs are finally cheap enough to drive CRTs off the market (slowly, ever so slowly): The 19" Sony LCD at which I’m writing this cost me $500, and a cheaper unit would have been $350, where in 1995 you’d pay $6,000 for an 11" LCD color display. I’ll stand by the last sentence in that comparison: “Only someone with a terminal case of technolust would trade in a 17-inch CRT, or even a 15-inch CRT, for a much more expensive 11-inch LCD screen.” The gap between CRT and thin-screen devices finally shrank, to the point that LCDs are better for most computer purposes other than gaming and high-end color work. (But then, we said that on p. 47: “Is an LCD screen better than a CRT? All else being equal it may be…”—and now “all else” is pretty much equal.)

We were too optimistic about LaserVision’s survival as a long-term consumer success; the discs were too big and too expensive to produce, making it easy for DVD to kill off the medium. We were right about that “dead duck” hard disk storage and its continued price advantage over RAM. Back then, the big news was that engineers got RAM down under $100 a megabyte—but hard disks were down to $0.60 to $1.00 a megabyte. Look at what 11 years has wrought: A 512MB flash RAM USB drive goes for $50 (about $100 per gigabyte, one-one-thousandth of the 1994 price)—but hard disks go for $0.60 to $1 per gigabyte, also one-one-thousandth of the 1994 price.

A couple of reviewers suggested that we were engaged in polemics. That’s true. This chapter may be the most polemic of them all. By and large, it stands up pretty well.

4. Electronic Publishing and Distribution

Remember Project Xanadu? It’s now been two decades of “coming to town any day now.” Meanwhile, Project Gutenberg’s Michael Hart continues to damage a useful project through his crazed mathematics and other publicity-hound quirks. New forms of “electronic broadside” have emerged and they’ve moved beyond the elite, what with blogs and wikis as well as tens of millions of personal websites. E-journals continue to emerge, but they’ve hardly served as “savior for libraries concerned with STM fields.”

The good news is that, so far, we were wrong about “metering the internet”—most of it continues to be available for a flat monthly access fee. Otherwise, this chapter is just fine. We missed the rise of journal aggregators—but we were on the money about $20/month “virtual libraries” not catching on within the average household.

5. Coping with Electronic Information

Almost this entire chapter reads well today, in my opinion. The section “Will full-text searching in very large databases ever work?” has been partly answered, but only partly. Yes, people do it. Yes, it can be made reasonably fast—but I question whether such full-text searching produces results anywhere near as good as those yielded through professional indexing and abstracting. We suggested that computers were ill suited to determine meaning. I’ll argue that this continues to be true.

6. Deconstructing Dreams of the All-Electronic Future

Another polemic, at least in part. The claimed projections on p. 86-87 were not straw men then, and aren’t now. They may be simplifications, but they represent real statements by real, and seemingly knowledgeable, writers and speakers. Some of these statements still arise from time to time, although the tide has shifted, possibly due in part to FLDMR.

Pages 90-92 may be wrong: Costs for conversion have come down considerably and the cost of storage has come down even further. That section wasn’t wrong in 1995, but times have changed. Still, despite the promise of Google Print, we don’t know what it would cost to do truly large-scale conversion to true preservation quality, or whether it’s feasible to do so on a worldwide scale. In 1995, three terabytes was a lot of storage. Now? Six inexpensive PC disk drives will do it, most likely for less than $3,000.

Page 98: While it’s still true that “one cannot buy a computer today for $47 that is as useful as the $3,000 computer was nine years ago,” $1,000 is no longer “the lowest viable price for a complete computer system if a minimal level of support and construction quality is to be achieved.” That price now appears to be $400 to $500 (including display), and as little as $600 for a notebook computer—but that’s partly because “minimal level of support” means something different in 2005, something closer to “it’s on the web; go find it yourself.”

One suggested price hasn’t changed all that much: the estimate of $0.025 per page (for paper and supplies) for “individual printing.” Except that nowadays the individual is more likely to use an inkjet printer—and it’s tough to print text pages on an inkjet for much less than $0.04 per page including paper.

The discussion of distributed printing costs continues to be relevant (and ignored) today. Truly “free electronic distribution” continues to be at least partly a myth. Much of the polemic still stands.

7. Enemies of the Library

My sole note after reading this chapter again, ten years later: “All jes’ fine.”

8. The Diversity of Libraries

On the downside, there’s possibly too much repetition here—both within the chapter and from previous chapters. It’s no longer true that 200 million Americans lack access to the internet, but still true that “tens of millions of Americans may have no need for, or interest in, such access”—to the chagrin of those who believe you can’t live without being online.

9. Economics of Collection and Access

Here again, we missed aggregators. That affects some of the discussion on p. 136 and pp. 147-148. Otherwise, this chapter continues to be pertinent.

10. Survival Guide to the Serials Crisis

I look at “Protecting Intellectual Property,” pp. 160-163, and take some shared pride, as I do in the rest of the chapter. There was no Open Access Movement in 1995. I have no reason to believe that any of OA’s founders read FLDMR. But I think we said important things about the serials crisis, and I think they’re still relevant.

11-12. Future Libraries: Beyond the Walls; Successful Libraries Make Their Own Luck

Both of these chapters still work.


I was looking for problems as objectively as I could. But I’m still the coauthor. I didn’t concern myself with excessive rhetoric. In 1994, there were far too many people projecting the death of print and the virtualization of the library, justifying some extreme rhetoric in response.

I’m not thrilled about the look of the book. For some reason, possibly the paper used, the desktop typography didn’t come out as crisp and clean as I would have liked. If you compare Being Analog, prepared with the same typefaces on the same printer using the same software and only slightly tweaked styles, you’ll see the difference. (First Have Something to Say is even nicer looking but I didn’t prepare camera-ready copy—and, as some of you may know, FHSTS resulted in my purchase of Berkeley Book, the type now used for the words you’re reading [if you’re reading the PDF version, that is].)

We didn’t prophesy every new technological development in libraries, but that wasn’t our aim. I, for one, would have done very badly in such an attempt—partly because I’m too much of an optimist. See our comment on Apple’s Newton as an example.

If I had it to do over again, given 1994’s situation, would I tone it down? Not really. I believe FLDMR changed the discussion among librarians. I’d like to think it helped move common library wisdom to recognition of complex libraries as the most likely and best future.

We did good. Sorry to disappoint those of you hoping I’d trash my own earlier work: Not this time.

Cites & Insights: Crawford at Large, Volume 5, Number 11, Whole Issue 67, ISSN 1534-0937, a journal of libraries, policy, technology and media, is written and produced by Walt Crawford, a senior analyst at RLG.

Cites & Insights is sponsored by YBP Library Services,

Hosting provided by Boise State University Libraries.

Opinions herein may not represent those of RLG, YBP Library Services, or Boise State University Libraries.

Send comments to Cites & Insights: Crawford at Large is copyright © 2005 by Walt Crawford: Some rights reserved.

All original material in this work is licensed under the Creative Commons Attribution-NonCommercial License. To view a copy of this license, visit or send a letter to Creative Commons, 559 Nathan Abbott Way, Stanford, California 94305, USA.