From: Bennett, Chip
To: ask_tim@oreilly.com
Subject: UNIX Side Appears to be Dwindling
Now, my response to Mr. O'Reilly:
- >>> Chip,
- >>>Well, there are two issues here: new topics, and updates.
Agreed ...
- >>> Frankly, many of the new topics for UNIX actually go under the guise of
other subject areas. For example, Perl has become a category unto
itself. Ditto network admin. So many things that used to appear as
"UNIX" books now fall into other categories. This is a sign of the
success of UNIX, not of its decline.
- >>>> Perl, Java, the Web--these are all children of UNIX.
OK, I see your point. In my mind these topics are "liberated" from UNIX, and I don't count them as UNIX topics anymore. But you're point is valid.
- >>> And while there is some change in the core UNIX topics, once we've
worked through an area, we don't do new books just for the heck of it.
Apart from the free variants (Linux and FreeBSD), there isn't as much
innovation in UNIX as there used to be. That's part of why we've been
banging the drum about "open source software." We've seen a lot less
innovation in UNIX since AT&T "took it commercial," and Sun, OSF and
AT&T started fighting over its future direction. The focus for
innovation has moved on to new areas that are still driven by open
source, such as the Internet.
My perspective is specific: Solaris (SunOS), which has been enhanced a lot of late. I understand that Solaris is only one flavor of UNIX, but it's market share is (or at least used to be) substantial. I remember a pie chart where Sun had half the UNIX workstation market share. But I suppose with the growth of the PC, MS Windows and free versions of UNIX have severely cut into that market share, making a Solaris driven publication strategy not profitable. At the same time there are many good books available from SunSoft Press that address Solaris directly.
- >>> The reason that there are still books for us to do on proprietary areas
like Microsoft or Oracle products isn't because there's been a lot of
innovation there, but just because we haven't done books in that area in
the past.
>>> I'd also like to note that even in our Windows publishing program, we're still spreading the same religion. For example, our NT administration books show how you can automate tasks with Perl rather than being stuck with the limitations of the Microsoft GUI.
>>> What's more, our books focus on cross-platform administration. We try not to get caught in religious battles about what software people *should* be using. We try to solve the problems they have with the software they *are* using. And cross platform issues with UNIX and NT are high on the list of problems our users face.
But system administration is very specific. UNIX administrators can't get much use out of an NT administration book! In system admin, there are more differences between UNIX and NT than there are similarities. (IMHO)
- >>> As to updates: we're often limited by the availability of the authors!
We've been trying to get Hal to update NFS and NIS for about 4 years!
I'm sure Mr. Stern is very busy. And Sun is probably not going to pay him to work on his own books! :-)
- >>> System Performance Tuning is being updated right now. You'll also see
updates coming soon for DNS and Bind, Learning Vi, and TCP/IP Network
Administration.
Oops... I just ordered TCP/IP Network Admin. You never know how long to wait. :-)
- >>> You're right about X Window System Administration. We should perhaps
take a look at that.
Thanks!
- >>> If you look at our latest catalog, you'll notice that we're also coming
out with a whole lot of "pocket references" on UNIX-related topics,
including GNU emacs, Perl/Tk, Tcl/Tk, Vi, and Python. (These join the
existing pocket refs on Perl and Sendmail.)
Thanks for taking the time to respond.
--Chip Bennett
