5 comments about Linux 3D Graphics Programming.

1. =

   This is a great book.

   1) It actually goes into the detail of X11 pixel formats and xshm. As an xlib programmer the generic class to determine pixel formats is great. It is superior is a number of ways to the evil O'Reily books in its dealing with XImage and graphics under linux.

   2) It ALSO goes over the basic foundations of 3d programming, and software rendering; as well as introducing the concepts of interfaces using c++ classes (factories, etc).

   For anyone who has not done this before, this is an excellent book. For anyone who has: It is probably still worth a read. Unless you are truely a master at linux graphics programming, there's probably something to be gleaned from this book. >However, fair warning: This is not a trivial topic. Reading the book end to end will achieve nothing. <

   Anyhow: I enjoyed the book. I thought it was worth while. My regards to the author!




2. To start with, let me just say; I'm a second year electrical engineering / computer science uni student. In Australia. =P

   This is a great book.

   1) It actually goes into the detail of X11 pixel formats and xshm. As an xlib programmer the generic class to determine pixel formats is great. It is superior is a number of ways to the evil O'Reily books in its dealing with XImage and graphics under linux.

   2) It ALSO goes over the basic foundations of 3d programming, and software rendering; as well as introducing the concepts of interfaces using c++ classes (factories, etc).

   For anyone who has not done this before, this is an excellent book. For anyone who has: It is probably still worth a read. Unless you are truely a master at linux graphics programming, there's probably something to be gleaned from this book. And if you are a master, why would you buy it?

   However, fair warning: This is not a trivial topic. Reading the book end to end will achieve nothing. I have highlighted (in bright yellow marker ;) about 1/2 the book. And spent a lot of time drawing pictures and trying understand what it said. This book requires a fair bit of study. I took me what? About 3 weeks to work through the whole thing.

   Having said so, it wasn't hard. I just haven't done much of this before, so I spent a lot of time programming examples to prove that the actually worked. =P

   Anyhow: I enjoyed the book. I thought it was worth while. :) My regards to the author!




3. I have to give the author enormous kudos for tackling this project. Over the course of two volumes (I'm referring to this book and his companion "Advanced" Linux 3D Graphics Programming), he builds, from the ground up, a complete working 3D engine, complete with texture mapping, collision detection and world editing capability.

   His writing and explanations on the topics at hand are very well written, and whenever the reader needs a diagram to make sense of something, there's invariably just the right one at hand to help him through it.

   It's really amazing, when you step back, to see just what is covered in these books. Not only are all major 3D graphics programming issues dealt with, but the author also takes time to discuss programming topics such as design patterns and why he uses them in his code.

   Having said all this, I did find some problems with the books, problems which kept me from giving 5 stars:

   1. Some of the code is difficult to follow. Really difficult. The polygon engine created by the author uses double-pointers indexed by arrays, and folks, that gets tough to read after a bit. In particular, I couldn't follow his polygon clipping code very well at all, nor his screen creation code, which involved a lot of bit-shifting, none of which was really explained all that well in the text. A bit of a lapse from the author I thought, very atypical.

   2. The author wanted his code to work with fixed and floating point math, and for that purpose he created macros for doing things like multiplication and division. All well and good, but again it hurts readablity to have all of those macros in the code when all you're trying to do is multiply two numbers together. Almost any CPU made in the past five years can handle floating point math very well, and so I don't see the need for fixed-point adjustments in the code. Just an opinion, others can disagree.

   3. The sample programs seem a bit lacking. After 300 pages, your reward is to see a program with a few flat-shaded polygons spinning around. It's hard to work up enthusiasm and bull through the book when that's your reward. Sorry if that sounds too harsh!

   4. The world editor was written with Perl, and... well, it adds another layer of complexity that maybe didn't need to be there. The syntax gets very scary very quickly.

   I could also say something about the fact that the author seems hot to trot for free development tools (Blender for modeling, xxgdb compiler, etc.), when maybe using some low-priced commercial products would have allowed for quicker progress and better results. I'm talking about expensive... tools here, not 3D Studio Max. Again, only one man's opinion here.

   Overall, it's a fine two-volume set, and if you want to see a 3D engine built from scratch, take a gander at this and prepare to learn how the magic works. Just don't expect to skate through it with no effort on your part.




4. This book is pretty good to get a novice up and running. Its definately not an advanced book so if you already know thsi stuff skip it.

   But for beginners its a good place tro get started




5. Norman Lin's Linux 3D Graphics Programming book is an excellent book for those with a programming background who wish to dive into the world of real time 3D graphics programming. The book is very well written and gives step by step explanations of all the concepts introduced as well as the most of the source code. 3D graphics is a complex subject and the author realizes that (lots of) hand-holding is very helpful in understanding the concepts introduced. This book covers the basics, i.e. rasterization, transformations, projections, and some modelling. In teaching all of these subjects, the author provides what is constantly missing from other books and what I feel is very important in the learning process - rationale for why things are done a certain way. The author clearly realizes this, and this is among the strongest points of the book. Other strong points of this book include:

   1) Some books simply discuss the concepts behind 3D graphics without providing incite into actually implementing them, i.e. accessing the screen, creating offscreen buffers, which data structures are best suited for a task. This books, aimed toward novices, does both.

   2) The author starts out by giving an introduction to graphics under Linux, with background on software as well as hardware rendering (using OpenGL/Mesa as well as the aging Glide API). This is great for those who aren't yet familiar with graphics under the Linux operating system.

   3) The code is object oriented - the author dedicates a section to explain the benefits of an object oriented approach in creating a graphics library.

   4) Software rendering is stressed, althrough hardware rendering is covered as well. Very important in creating a 3d engine - using OpenGL without understanding how it does the things it does will only get you so far.

   5) The book is NOT just Linux specific. Linux is the author's choice of implementation but all the concepts can be applied to other operating system. The object oriented approach allows for the graphics library to be implemented in other OS's - the appendix even discusses compilation under Windows.

   6) All tools to get started are available on the companion cd-rom as well as free on the internet. The source code used in the book, the Debian distribution of Linux, the compiler tools and libraries, and the modelling software Blender (for use in the chapter that discusses modelling) are all included.

   7) The author is very knowledgable in Linux and is clearly a advocate for the operation system :).

   The author assumes you are sufficient in C/C++ programming. This is a fair assumption as you really have to be in order to do graphics programming. Be warned however, that the source code may not compile right away under other Linux distributions besides Debian and newer versions of gcc. Some modifications in the source code (very minor) may be required in order for the code to compile correctly - something you should be able to do. If not, perhaps you should hold off on graphics programming and get some more programming experience first (harsh but true). All in all, Linux 3D Graphics Programming is an excellent introductory book.

5 comments about C++ Programmer's Guide to the Standard Template Library.

1. This book is very helpful for somebody familiar with C++ and wanting to start using the STL. It is clear and complete if you're willing to read each chapter end to end. But if you're reasonably familiar with STL's concepts, or comfortable with abstract container classes in general, this book will frustrate you to no end.

   The book could be very good if the "reference" section really was a reference, and not just a rehash of the description. Instead of just having lists of methods and descriptions of what they do, the reference is written in lengthy prose, and must be read sequentially in order to be understood. Rather than each entry standing on its own, method descriptions say "hardly any change from the similar method in the *blah* class." This makes looking up information extremely difficult.

   It was great to get me on my feet. Now I'm really sick of it.




2. Unlike some authors, Nelson doesn't oversell the STL or try to defend its most ill-advised features and peculiar terminology. He addresses the readers as respected colleagues with whom we wants to share some exciting information.

   The 21 chapters are orgazized into three main parts:

   I. Introducing the STL

   II. The essentials: containers, iterators, algorithms, functions.

   III. The public interface: reference information If you're a experienced C++ programmer, the first two parts are must reading. The third part is a bit too detailed and repetitive.

   The index is comprehensive, and definitions are provided for important terms (exception: "algorithm complexity" and O(n) notation, concepts well known to Computer Science majors but not to every practicing programmer).




3. The book is kinda old and slighty out of sync with the latest achievements of the C++ standard-makers, and yet it is the best tutorial book on the core STL there is: no hype, lot of smart and relevant sample code, good explanations (really detailed and indepth). I keep using it and I'm very happy.
   
   Not one from among the latest flood of the stl books (Josuttis, etc.) is even close to the usefulness of Mark Nelson's book -- one can only wish he issued an update.
   
   I've revisited this review on 02/19/2007 -- and everything I'd said originally still holds: whenever I need to really understand how STL things come together, I got to Nelson's book, Josuttis doesn't cut it, he's too superficial, it's a headers dump rather than a good explanation. Not a bad book (and it's more up-to-date), but not in the same league. I'm amazed at the fact that there's a bunch of used copies offered above for buck-fifty. Whoever works with C++ and uses STL must have this book. I really mean it!



4. If you have started using STL heavily and need to know how it works, this book is a must have.(albeit outdated). I have not come across such an easy to read C++ book in years.Somehow the experts save Scott Myers,tend to hide the crucial details ,may be because its too obvious for them, not Mark Nelson.
   The code used in the book is not exception safe, but the basic ideas still hold. I really benefited by keeping a copy of the latest STL headers open , and read the text.
   Wonder if the author will ever redo his work to adjust to the latest STL. Nevertheless its indispensable for the intermediate C++ programmer trying to think like an expert.



5. True, a great deal of the material contained within these pages is now available online. But, having this volume at your desk for handy quick reference to the basics of using the standard template library (STL), is a necessity for anyone hoping to utilize the STL potential. The book is organized in a way that makes sense to experienced software engineers - with container template introductions in one section followed by algorithm introductions in another. The final section contains the STL specification. It is a reference I would not be without and find that I am loaning it often to others who have also noted the benefit of having so much STL information in one volume. A must have for every C++ professional.

5 comments about Practical Algorithms for Programmers.

1. I bought this book 3 years ago. I learn a lot from it. I still use it. It's worth the money.



2. This book has the singular benefit of doing complete implementations of algorithms, not just code snippets nor pseudo-code. (The code is in C.)

   The book also covers lots of the algorithms other books forget: searching for multiple strings at once, searching for regular expressions, complete calendar routines, arbitrary-precision arithmetic, phonetic searches, most data-compression algorithms (including the one used by Microsoft in its product distribution), etc. Plus all the traditional sorts and data structures that appear in other books. As to data structures, it's one of the few books that implements an ISAM routine using your choice of red-black or B-trees.

   (By the way, although there is a coupon at the back of the book for a source code disk, the code is made available here for free download: http://www.pacificdataworks.com/bookdownload.html The authors update the codebase intermittently.)

   A good book, which has no direct counterpart.




3. "Practical Algorithms for Programmers" fails to deliver what is mentioned by authors as "The purpose of this book is providing a practical compendium of algorithms for use in applications" and "Most algorithm books today are either academic text books or rehash of the same tired set of algorithms". This book goes over that tired set of algorithms over again (B-Trees, bubble sort, shell sort etc) and is filled with long code listings with little comments and faux paus code practices (if column > 61).
   
   Almost all of the algorithms discussed in the book have already been very well described in various books of the same genre, notably "Introduction to Algorithms, Second Edition by Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest, Clifford Stein" which has better explanation and examples. I see that "Practical Algorithms for Programmers" has good description of algorithmic efficiency, B-trees, AVL trees, phonetic comparisons, soundEx and metaphone along with excercises, but as an overall study for a CS grad and/or skilled developer, it's repetitive and meaningless. If authors' idea was to provide a cookbook for algorithm implementation, due to lack of component oriented thinking, this book lack this prospect as well. As mentioned, it's not a text book and therefore I don't see a reason of having about hundred pages of printed source code in the book? If the intended audience are software developers, why not highlight the important code segments and let the rest available via CD/FTP and use remaining pages for practical industry implementation discussions like the title suggests.
   
   As mentioned by other reviewers, this book might be a short & quick review or refresher course but I believe that it doesn't add any valuable reference to existing set of books available in this niche. Especially in the current development era when underlying software architectures and programming languages provides the built-in complex datatypes and memory management, algorithm world now belongs to exploration of binomial option pricing, naïve Bayes filtering and normal distribution approximation style studies.



4. We have this book in our lab and many times I've been able to refactor my own implementation after reading the code in this book. The chapter that deals with dates is very well presented and doubles as a history lesson since it covers all the adjustments that have been made to our calendar over the centuries. The source is writen in plain C so it should be easy to port to other languages.



5. There are numerous well written books on algorithms. Those by Sedgewick and Knuth come to mind, for example. But some students find these too hard. Binstock and Rex aimed their work at this need. This book has very little in the way of fancy maths. It emphasises examples with complete code listings. Not unlike "Numerical Recipes". Hence, you can also treat the text as a cookbook for your needs.
   
   The code is in C. If you are using another language, you can get some practise in seeing how well and accurately you understand the book's code.

No comments about The Dictionary of Standard C.

No comments about Introduction to C Programming with a little OpenGL Graphics for fun.

No comments about C++ Basics.

3 comments about Obfuscated C and Other Mysteries.

1. This book is not only about obfuscated C, as seen in the Obfuscted C Code Contest, although that is certainly a big part of the book. Libes also covers a lot of other ground, that will be a big help in your day-to-day programming.

   But that isn't all: Libes actually makes all this hardcore C stuff fun! I burst out laughing many times while reading the book, which (you may have discovered) doesn't happen often, when reading books on programming.

   This book is for anyone who thinks that C-programming will one day be recognized as the artform it truly is!




2. If you liked the ``butt ugly fish book'' from Peter van der Linden, you'll love Libes! Like _Expert C Programming_, this is an immensely readable book complete with humor and occasional drama. Don't let the ``Obfuscated'' part of the title fool you. The book doesn't dwell excessively on the IOCC nor is it oblique. Libes does an excellent job of turning obfuscation (and mystery) into clarity. It's great _recreational_ reading for C linguists...with recreational meant in the ``re-create'' sense of the word. Hey Don -- how 'bout an update so you can merit a 10 rating?!?



3. The OCC code snippets and Don Libes commentary should be
   a required reading in any programming class. Unfortunately
   It is hard to find.
   
   This book was an relevation, the gems in it
   proved that language (C/C++/Java) has nothing to do
   with beatiful ideas and the genius of programmers like
   Larry Wall (past winners of OCC) and the algorithms
   that make a program.
   
   Read this book, and you will never have to join a
   'C vs C++ vs Java' discussions, you will be too busy writing
   beautiful program, in the tradition of Stallman, Knuth and Wall.
   Programs should be beautiful like Poetry.

No comments about Quick Recipes on Symbian OS: Mastering C++ Smartphone Development (Symbian Press).

No comments about Financial Modeling Using C++ (Wiley Finance).

2 comments about STL for C++ Programmers.

1. This text was a great introduction to the use of STL, but it lacked depth in that it only covered how to use STL not how STL works, or how a programmer may derive their own STL classes from the base. It would have been very informative to get a look at a set of overridden classes such as a special iterator, a variant of an allocator class where a constructor with parameters is called instead of the base constructor, etc.. I haven't been able to find such an in-depth book yet. This book never claimed to explain these sort of questions, but it would have been extremely useful, say as the last chapter.



2. Mr Ammeraal writes in a clear, easily understood style. He light on theory but very clear on application. However the book does not have the depth to be a refernce manual. In fact, he doesn't even cover the entire STL. The String class isn't even mentioned. The index is minimalistic. If you need to get a basic understanding of the STL and do not plan on manipulating character strings, this will get you started. If you need to understand the string class or are looking for a useful reference manual, stay away from this book.