Path: utzoo!attcan!uunet!mcvax!cernvax!pan!jw
From: jw@pan.UUCP (Jamie Watson)
Newsgroups: comp.windows.x
Subject: O'Reilly X Books
Message-ID: <469@pan.UUCP>
Date: 11 Sep 88 20:04:17 GMT
Reply-To: jw@pan.UUCP (Jamie Watson)
Organization: Adasoft AG, Solothurn, Switzerland
Lines: 54


I've just read the announcement in comp.newproducts about the new release
of the O'Reilly books on X.  I would like to make a few comments about
these books, for the benefit of those who haven't had time to investigate
them yet.

My first complaint about these books is their binding.  The earlier
editions were loose-bound, so they were at least convenient to use
on a cluttered desk.  The new editions are "perfect bound" - glued
down the spine.  This makes it impossible to get them to stay open
to a particular page, and thus essentially useless on a cluttered
desk.

My second, and more serious, complaint, is about the books' content.
One of my colleagues bought the first edition of these books when they
were first available, at the winter Usenix meeting in Dallas.  They had
seemed fairly impressive simply from the sheer size of the two volumes.
Once we started to look at them in detail, though, we found that they
were absolutely full of mistakes and inaccuracies - to the point that
they were essentially useless.  We decided to keep our mouths shut, since
O'Reilly was already talking about a "new, improved" edition.

Well, I picked up the new edition while at Xhibition last month.  The
worst of the errors are gone, but I am still very, very disturbed by
the fact that it seems obvious to me that the authors of this book do
not understand what they are writing about at all.  The books contain
quite a lot of conceptual discussion, and the author(s) repeatedly show
a serious lack of understanding of their topics.  For example, in the
Introduction to Volume I, they discuss window managers.  In the space
of two paragraphs, they display a total lack of understanding of the
concepts involved, and manage to completely confuse the issues of
real estate management, input focus, and listeners.

This situation is repeated in their discussion of optimization and
networking; they are unable to distinguish between general performance
issues and network dependencies.

The problem with these kind of errors are that they undermine your
confidence in the entire presentation.  I find that I am not willing
to believe anything that they say with which I am not already familiar,
since I don't trust their basic knowledge, so I gain very little from
reading the books.  I do quite a bit of consulting and teaching of
the X Window System, and I tried to imagine what it would be like to
have learned X only from these manuals.  The picture of misinformation
and misdirection that I saw was not encouraging.

The second volume is essentially a reproduction of the reference
manual from the MIT documentation, but alphabetically organized, one
function to a page, in Unix man page format.  While this didn't take
any great talent, it is something that has been needed for a long
time.  It would be really wonderful - if I could just figure out how
to keep the blasted thing open to a particular page on my desk...

jw