Game Programming Gems Volume 2
Author: Various Contributors (Edited by Mark Deloura)
Publisher: Charles Rivers Media
Published: 2001
ISBN: 1-58450-049-2
Purchasing: [Amazon.Com] [Fatbrain.Com] - RRP US$69.95
Reviewed: 3rd September 2001

Front Cover Shot:

Overview of whats on offer

Picking up this book for the first time it looks and feels almost identical to the previous volume in the series (reviewed here), but that is to be expected. Looking inside the book we find much of the same, however this isn't a bad thing at all - if you liked the first volume you'll like the second volume just as much. This is one situation where more of the same is more than welcome...

The section layout:
Section 1 - General Programming (22 gems) - general programming techniques and tips, debugging, optimising, structures etc...
Section 2 - Mathematics (8 gems) - Clever maths algorithms and techniques for games.
Section 3 - Artificial Intelligence (14 gems) - Articles on making lifelike enemies and NPCs in your games.
Section 4 - Geometry Management (11 gems) - How to manage the now huge amounts of geometry in games (and other geometry tricks)
Section 5 - Graphic Display (8 gems) - Mostly 2D related stuff - shadows, light maps, cartoons, procedural textures
Section 6 - Audio Programming (7 gems) - Sound effects and music programming.

In total we have 70 gems to read, 7 more than in the previous volume, which is a good start - although some of the articles do give the impression of space filling really, whilst I dont hold anything against the "Using Web Cameras in Video Games" (it's a good article), it is a bit of a strange 'gem' to find...

More More More !!

As I already said, in this case more is better. I can see Mark Deloura (editor) lining up at least another couple of these books - we may now have a total of 133 gems between the two volumes, but I can list several things he's not covered, and with the success of this series so far I can see the money-men are going to be more than happy to produce another volume or two (not a bad thing in my opinion).

Apart from the all-new content, there are some nice refinements to the book - Mark Deloura really didn't need to change the format of the book much, but what he has done is add numerous little changes here and there that refine the quality of the book in general. None of them are particularly significant on their own, but they add up. New things such as Author biographies at the beginning of the book (we never got told much about the authors in the first volume), section introductions by the section-editor, "On The CD" icons in the margin marking what is to be found for that article on the CD...

The content has also been refined, quite a few of the gems refer back to gems in volume 1 (not necessarily by the same author); a good example of this is the neural networking gem. Andre LaMothe did a great (and complicated) gem in the first volume, whilst I didn't understand much of it at the time I got the impression that with the gem alone you could do that much work with neural networks, in this volume we get John Manslow producing a gem that demonstrates a concrete example of neural networking in games, whilst I still dont understand it a huge amount it is certainly much easier to take the code/idea and form your own neural networking AI engine.

On the other hand, the content has also got more technical, and a bit "odd". Take the gem 3.13 "Imploding Combinatorial Explosion in a Fuzzy System" - now that just sounds hard! looking at the article it isn't really that complicated an idea, but it's based on Fuzzy Logic AI - which is required as a foundation before reading this gem. Then again you get the "Using Web Cameras in Video Games" gem, which just sounds odd to me - since when have people wanted to (or attempted to) use web-cams in a computer game?? Maybe this is a sign of things to come, but still... :-)

Another aspect of the increased content is the overlapping, many articles reference to other articles in the previous volume, OR, and more importantly, techniques you have learnt in other gems open the doorway to new techniques in this volume, or allow you to expand your current understanding to a new (higher) level. The first volume had a lot of gems on 3D graphics and geometry, as does this book (but not quite so much) - If you own the two volumes you can quickly find two or three articles on a very general subject - collision detection, view culling, landscape generation, and yet more tricks for speeding up your AI engine. The bottom line is that having both volumes makes for a very good resource - which has the knock on (and negative) result of having to spend twice as much buying two books (if you dont own volume 1 already).

Surely it cant be this great?

Not everything is as good as it seems - even this book, which so far is proving to be top-quality.

The biggest problem that I see, and one that I mentioned in the review of the first book (here) is the coverage of different areas. We now have a section on audio programming, whilst it's a welcome feature, it could be bigger and better, and we still dont have anything about networking/multiplayer games - which in this day and age is very strange (people are always spouting about how internet/multiplayer gaming is the future...), and less importantly we still dont have a section on input programming. I can see whats going on here, there are already two volumes - I'll put money on this getting upto volume four within the next two years; there is more than enough content to warrant more volumes (and the content is always evolving), but I really hope that they change the balance for the next volume - a little less on 3D graphics and a bit more on audio/music/multiplayer programming please!

The second downside is that it fits too well as a series - what? thats no negative thing! true, true, but if you dont have enough money for both volumes (and you dont own either already) then you will be missing out on a large part of the 'experience'. You can quite easily get a lot out of either this volume or the first volume, but you get a lot lot more out of having both volumes.

Thirdly, the language aspects. This is something I mentioned in the first review, and something I've mentioned in all but 1 book review so far - it's all in C or C++. This is to be expected, as I'm aware of only 1 VB game programming book, yet there are probably 10-20 recent C/C++ game programming books. Either this is an instant turn-off for you or it wont bother you too much. The generally advanced nature of the content requires you to have a good head for game programming - hence learning a bit of C/C++ so you can understand the code isn't going to prove too hard... I dont think you will ever find an equivelent resource in plain VB.

Finishing things off

In summary then, we have a good book here - no questions there. If you were to shell out the cash for this book you would not be making a mistake - I think you would be hard pressed to find anyone who was unhappy with their purchase, not everyone is amazingly impressed (for a variety of reasons), but the overall general opinion is thumbs-up. The biggest question is whether the content in this book is worth the $60/£50/however much it costs in your local store/webshop. Here is my summary of key points: