Customer Reviews

The Linux Kernel Book

5 star:		(0)
4 star:		(0)
3 star:		(1)
2 star:		(5)
1 star:		(1)

 

 

-

Computer-related books are a vast wasteland. Most of my quibbles about the book would be irrelevant if it were titled "The Linux Kernel for Cretins." In that case I'd have had to give it two stars. But, it appears to be targeted at serious kernel programmers.

In most places, the previously mentioned "poor translation from the French" wasn't perceptible to me. However, I _was_ disappointed with the technical content. Conditioned by things like the BSD book by McKusick et al, I expect discussion about interesting technical points instead of descriptions of kernel entry points. If I wanted to know what functions were available for signal management, I guess I'd read the man pages.

However, all the user-level information does bulk the book up nicely, obscuring the fact that it doesn't have much fiber. For example, I wanted to know something about how _procfs_ was implemented. After eliminating the description of various files in /proc (which are documented better elsewhere) there are about two pages of description--mostly of the "Jeshua begat Shemtup" variety. I can use tags and glimpse as well as the next guy--what I'd like is a higher-level description.

In summary, it has some interesting information, but any kernel book I can skim through in less than a day isn't much of a kernel book, Senator.

Before buying this book, I have bought another book - Linux Kernel Internals which is good but not up-to-date. So I buy this - The Linux Kernel Book for getting more up-to-date materials. I cannot get what I want in it. One of my disappointment is the index, the index is so messy, I cannot follow the structure of the index, so I can't find the materials quickly. Also, the content is organized in a mess also. I regret that I have bought this book, I waste my money and waste my time.

This is a very difficult book to read, which is a pity as I think the French version must be good. The content is well structured given its highly technical nature. The presentation is superior to similar books on Linux (which seem to think it's OK use the source code to explain the source code.) The authors clearly know their stuff. All the right ingredients.

Unfortunately the translation to English is dreadful. Another Amazon reviewer suggested it was done using a computer. I can't really tell, but the effect is that I was unable to concentrate on the topic at hand. Instead I was trying to understand the sometimes nonsensical grammar, or grind my way through sentences full of thoughtless reuse of a particular word. (The processes chapter has an great example, five instances of the word 'process' in two lines of prose.) Even a translator program might try a synonym once in a while!

Though the presentation is better than other Linux technical publications, it still has the feel of a 'techie' book. Classic books on similar subjects, for example those by Bach or Tanenbaum, manage to avoid printing endless tables full of useless data, yet they still educate the reader. It's a matter of personal taste of course, but in my opinion the editor could have saved a good deal of paper simply by tactful application of the delete key.

I guess it must be tough to find someone who has enough understanding of both written English and programming to translate such a book, but it's a shame they didn't do a better job. I have seen it done well with equally tough topics.

I hope there's a new edition for Linux 2.2, but I wouldn't recommend buying it unless they get someone else to do the English version.

Awkward english usage that appears to have been translated by a computer program and then glossed over by an editor not familiar with English usage and grammar keep this book off my list.

I am returning it after reading three chapters. I can't deal with the many obvious misprintings, and the almost unreadable sentences. Sentence fragments and unclear statements abound.

The content would otherwise be indispensible to me. Each topic is introduced conceptually, and then the implementation is explained as an overview, and then in detail. I appreciated the space taken for explanations of basic Unix architecture concepts. There are footnotes at the bottom of many of the pages, listing other books to read, and places to go on the web for more information.

I recently purchased this book, and if one can get past the poor French-English translation, you will find that the book's content is very informative. It is not the best book on the subject; however, it is certainly not the worst, nor is it as poor as some people say it is. Most of the errors in the English can and should be corrected by the publisher, and presented in a new updated edition. Or perhaps they should GPL this book,so we can edit it, and improve it for the authors, as well as add new features to cover the latest Linux Kernel. What I am getting at? The book's structure is not bad, and I don't think that the authors' were wrong for adding various tables, and introductory information in the book, I think people are upset because it appears as if it is written in poor English,which is a shame, there really is good information in it. Well, if your like me, buy it anyways, read it, correct the English mistakes, edit any other obvious errors, and e-mail the authors with what you find. Despite its shortcommings, I still think it is a good introduction to the Linux kernel. The point is to learn about kernels, Linux in particular, and get past all obstacles in the process.

Apart from the poor translation quality (some variable names and file names are actually in French), the authors seem quite undecided about the target audience; each topic starts from the scratch, from basic user level stuff before going on to the kernel and implementation. This can be quite frustrating for users who bought the book for what its title would convey. It must neverthless be said that there is no dearth of technical content or competency - provided the reader is ready to wade through what should probably belong to another book for a different audience.

Apart from the poor translation quality (some variable names and file names are actually in French), the author's seem undecided about the target audience - each topic starts from the start, from basic user level stuff and then goes on to the kernel and implementation. This can be quite frustrating for users who bought the book for what its title would convey. It must neverthless be said that there is no dearth of technical content or competency - provided the reader is ready to wade through what should probably belong to another book for a different audience.