[
Lists Home |
Date Index |
Thread Index
]
As I think many of the people on this list know, my day job is working
as an editor at O'Reilly & Associates. For the past two plus years,
I've focused on building an XML book list, with some significant
attention to Web Services and occasional work on Web books.
Lately, however, I'm spending less and less time on XML and Web Services
books and more time on books in other computing fields. Given how much
easier it is for me to edit material in fields I know well, this is a
serious change on my part. It's necessary, though, because as Tim Bray
summarized my employer:
"Oh, and by the way, that XML book you were thinking of writing?
Forget it." [1]
XML book sales have dropped substantially, even relative to the overall
decline in technology books. A few books dominate the broad (typically
though not necessarily beginner) end of the market, while more focused
books struggle to achieve the numbers needed to justify their
publication.
It's not merely a matter of the pie being divided into too many
overpublished pieces, but rather a matter of the space itself shrinking.
There aren't a lot of publishers still enthusiastic about publishing
into the XML space as a result. (A few, including O'Reilly, are
continuing to publish, though at a slower pace.)
There are a lot of things going on here, but it's interesting because
the shrinkage is more complex than the predictable collapse of a hype
wave. There's been very little sign - in my hindsight anyway - that
developers ever moved from beginner books on XML to more detailed books
focusing on more advanced technologies. XSLT appears to be the furthest
frontier most readers reach, and many stay within a range of books
covering XML and their particular (Java, .NET, Perl) programming
environment. A few XML Schema books have done adequately, but a lot
have disappeared. Web Services books have fared especially badly. Even
books on hiding XML behind other technologies, like O'Reilly's _Java &
XML Data Binding_, haven't really caught on.
My current guesses as to why this decline is happening include:
The hype wave, of course -
I don't think XML was ever quite what it was promised to be
to programmers, and there's a good deal of annoyance from that.
(Publishing folks of various kinds still seem to be expanding their
use of XML, however, maybe even getting happier with it.)
Standards confusion -
I think W3C XML Schema convinced a lot of people that true
understanding of XML and vocabulary creation was way beyond their
skills. The cost/benefit ratio is way out of whack at the
beginning. The pile-on of additional specs and the lack of a clear
processing model for combining them didn't help either - too many
optional parts. The amount of XML most programmers need is in the
intro books, maybe supplemented with something environment-specific
like Java & XML. (Web Services has massive general standards
confusion, an even tougher problem, not to mention even more "WS"
specs than there are "XML" specs.)
Big books for small problems -
While some of us do have complex problems we need to solve with XML,
it seems like 80% of the XML out there is produced by people who
know about elements and attributes and maybe the built-in entities.
XML is ubiquitous, but in forms which don't require enormous study.
These folks don't really need multiple books - they can find what
they need online easily enough.
Now that the hype wave has retreated, the world seems full of people who
are getting by with a remarkably smaller set of pieces than the XML
community - and its supporting publishers - have been selling. In some
ways, I think Common XML [2] is triumphing. At the same time, I worry
that "XML" in the broader sense may have armed itself for
inter-continental thermonuclear combat when a broom for cleaning things
up was all many people wanted, and pretty close to the proposition of
the original project.
I'm not entirely sure what I'm hoping to accomplish with this message in
this forum. I suspect that I'd like the standards makers to know that
their publications are generating less excitement - sales are down,
difficulty of producing books is up. Perhaps more important, I'd like
people who use just a relatively small subset of all of these XML
technologies to know that they're definitely not alone.
[1] - http://tbray.org/ongoing/When/200x/2003/10/11/FooNotes - Tim
O'Reilly and Mike Hendrickson presented on trends in publishing using
both Bookscan data and Amazon rank data. Sadly, that information is not
itself public (Bookscan doesn't want its participating stores to
see the data from other stores, etc.), but does include multiple
publishers. I'm not just discussing O'Reilly results here, and as my
employer has discussed this much in public, I don't think I'm treading
on dangerous confidentiality grounds.
[2] - http://www.simonstl.com/articles/cxmlspec.txt
|
