Book review: Fedora 7 Unleashed
by Neil Horman
Have you ever had deja vu? I re-read books on occasion, because I like them, and every once in a while I’ll re-read a book that I think I’m reading for the first time. Then I’ll sit there with this twisted-up look on my face, wondering why all the words seem so familiar. Then I remember when and where I saw them last.
I’ve been reading the new Fedora™ 7 Unleashed book by Andrew and Paul Hudson, and I’ve had that feeling several times. So I’ve made my face and wracked my brain, trying to figure out how I’ve read this before. The answer? I read Fedora Core 6 Unleashed and Fedora Core 5 Unleashed before that.
Its unfortunate. I think these ‘distro tomes,’ so to speak, provide a valuable service to the Linux uninitiated, and can be useful to more seasoned sysadmins interested in the latest technology a new release of a given distribution has to offer. The unfortunate part is that both of these groups have to suffer through repetitive rehashing of methods, processes, and utilities that have been throughly documented in a multitude of locations and media.
To be fair, this book does some things quite well, but as I’ve looked at more and more works in its genre, I’ve come to realize that the good ones are characterized by a bias toward documenting whatever new technology is available, while the less useful ones focus more on maximizing the width of their spine with well-worn stock material. I’m sorry to say that this book leans more toward the latter.
But lets not dwell on the negative (or at least let’s not start there). Theres good in everyone, and there is good in this book. Some of you may remember that I wrote a review of Tammy Fox’s book, Red Hat® Enterprise Linux® 5 Administration Unleashed. This book, like that one, does many of the same things rather well. Specifically, it handles new technologies with aplomb.
Xen, while not new technology per se, is sufficiently in its infancy. The online documentation is still somewhat fragmented, and a concise summary of how xen virtualization operates is a perfect example of what this book excels at. SELinux also gets very fair treatment in this book, as does the Mono programming language. While not strictly bleeding edge, these topics are still sufficiently new that consolidated documentation is a benefit that books like this provide very well.
Most of the sections in this book have their usefulness. Everyone needs to know how the useradd or cp commands work. But I just can’t get past (and yes, I’m back to dwelling on the negative) how re-hashed so many of these chapters are. As part of my review of this book, I went and took a quick look at the table of contents for Fedora Core 6 Unleashed. Written by the same authors, its an amazing example of documentation re-use. Several of the chapters look sufficiently similar as to be identical.
I don’t want to insinuate that the information in these sections isn’t useful, it really is. It contains information everyone needs to know. You can’t administer a system without being able to change passwords, set up remote access, preform backups, so on and so forth. But good gosh! Lets just take one example: Chapter 17, Apache Web Server Administration. Its almost 40 pages of documentation about how to manage, configure, and tune the Apache web server to provide web pages to interested parties. Hmm, I wonder what would happen if I typed ‘Apache howto’ into Google. Dum de dum….hey, 2.550,000 hits! I wonder if Apache has a web site…..hey what do you know? They look like they might have some comprehensive documentation. And Amazon indicates there are 1,285 books on sale at their site exclusively on the subject of the Apache web server.
The same is true of the sections on FTP, database, SSH and network file access. Note to the authors: We’re covered. Your material is useful and factually accurate, but I just can’t get past the fact that it’s not really necessary. You could have written a book that was half as big, was consequently less expensive, and yet still provided all the same useful content.
I have a feeling it’s not the authors fault. You can pick up any number of comprehensive high-level OS documentation books, and see this. There is some material that is just covered again….and again….and again. My inner conspiracy theorist is convinced it’s more than coincidence. I firmly believe that (a) extraterrestrials are out there, and (b) publishers of these books will encourage their authors to add content in an effort to claim more shelf space. I think a better technical bookshelf could be constructed with a series of more targeted topics, but such an approach lowers the revenue generated by this type of books.
So that only leaves the final question: Should you buy this book or not? I’ve been pretty hard on it–in my view, with good reason. Despite that, I still think there is a demographic for whom this book would be fairly useful.
If you have any significant experience with a Linux distribution of any origin, I would expect that you have the skill (both technical and research-oriented) to find the information contained in this book on the web (or elsewhere). Likewise, if you’re a sysadmin, and you have experience (in a Linux or non-Linux environment) I would expect that your technical bookshelf has much greater depth, and much more targeted breadth, specific to your needs. So, clearly, this one’s not for those with experience.
However, if you are new to Linux and are interested in getting your feet wet, you probably don’t want to invest a fortune on reference material, or spend too many hours poking about on mailing lists or web sites putting together a reasonable set of information to get started. If this is you, then yes, this book is a good choice. It offers a comprehensive outline of what you need to know to work with Fedora 7, both as an end user and as an admin. And thats quite likely why these books tend to be a perennial favorite.
The Linux community has a consistent influx of new users and curious onlookers who have the desire to know more. A consumer-oriented, one-stop shop might get them there faster than the community would, with its plethora of websites and arcane knowledge passed down though stories on cryptic mailing lists. So to keep current, these books keep coming. And if you’re new, it could be your first step into a bigger world.
November 1st, 2007 at 6:10 pm
One of the things that come up when writing a ‘tome’ book on technology is that editors want the book to be packed for important reasons.
1) Technology books have a store-shelf life of 6 months to 1 year. Most people buying such books do so on size of book and if they know what is in the book how much that book might help them after the OS is wiped out with FC-8/9/10.
2) A good portion of these books are sold to non-technical savy people who came home and found the home computer has some new OS on it when Billy goes abck to college. And Billy’s instructions of ‘look it up on Google’ means zip to you. You market to this crowd with size. They could go for the 100 page NO BS Guide To Fedora’s OR you could supersize it with the 1200 page Fedora BIBLE New, Old, and Apocryphal Edition. Most people will buy the bigger version because thats how the human brain works.
3) Fedora and Ubuntu do releases at extremely high rates so the authors do not have time to update/change things as much as they might like. So the publisher must make as much of the book stable so that you can basically reuse the chapters over and over again because they may only get one book sell in 3-4 releases (eg that guy who just bought Fedora Core 6 unleashed is going to not buy 7,8,9 unless something really major changed).
4) People selling books are selling them to people who do not like to read long times at computers or want to read them when ‘cogitating’. There is a reason that many of these books are printed on ‘toilet paper’ stock. I find that these times of deep thinking are usually when I want a nice big book, and I want them to have lots of stuff.
So anyway.. my take on big books.
November 3rd, 2007 at 6:59 am
The very first RedHat book I owned was an unleashed edition of RedHat Linux 5. I was thrusted into a SysAdmin position as my company transition from HPUX to Gnu/Linux because I was the only one who ever heard of Linux, and I happend to run it on my desktop.
I can’t remember ever reading the book, in the sense one reads a novel. I do however remember feeling a slight sense of security with a 1,000 page reference guide on my shelf overtop my workstation.
I have been a Gnu/Linux user now for 10 years, and as my career grew so did my collection of specific books on topics such as Bash, Perl, etc … but I still have that first unleashed (can’t seem to throw it away even though it is outdated). I notice as I go into other folks offices a different version of some comprehensive OS book adorning their shelf as well, be it O’Reily, Unleashed, etc … but it seems at some point everyone owned/owns one of these for al the derision we give them.
I would appreciate some reviews of more low-level books, like a kernel internals book and a side by side comparison of what ones you find most useful because face it, googling things are great for your immediate sysadmin/troubleshooting duties…but when it comes time to write modules I still don’t fully trust google for the deep understanding of what you are doing. It is a great source for commands and how-tos, but the WHY is often missing…which is where these books fill out.
November 8th, 2007 at 11:31 pm
Tonight I went to my favorite book store and carefully looked over 10 or so books for two hours only to be disappointed as usual. This was one of the books I looked over. I do this about once a year and I am always disappointed. I would love to see a book that leverages new tools and distros to build a best practices network of Linux systems and describes the best way to manage them. All these books focused on one or two systems at a time. I suppose that might be useful it I had one or two Linux systems, but it doesn’t help me determine the best way to manage many systems. I’m left wondering what kool stuff is going on behind the closed doors of other medium to large Linux networks.