Roger T. Stevens, Object-Oriented Graphics Programming in C++ (AP Professional, 1994)
[originally posted 12Feb2002]
Whatever other faults this book may have, most of it is (quite thankfully) out of date. Remember the days when 640x480x16.7 million colors was something only the high-end computers could manage? Well, that’s when this was written. Also, raytracing as a way to display graphics seems to have gone the way of the great auk for most programmers, largely due to speed concerns. (This is unfortunate, because raytracing produces some of the highest-quality graphics you’re likely to find on a computer screen; I have a suspicion that if someone were to sit down and really work with newer implementations of C++ and today’s high-powered graphics cards, raytracing could be done with enough speed to allow it to rival newer methods of graphic representation.)
Aside from the out-of-dateness, which makes this valuable only in an archival sense at this late date, the title is a bit on the misleading side. While Stevens does take some of the methods and tactics of object oriented programming into account, large portions of the code therein look suspiciously like C that’s still in the process of conversion. This is especially true eight years after the writing of the book, when so much more of the C++ backbone has been standardized by the ANSI committee. Creating an
updated edition of the book would require almost all the code to be rewritten.
Given the first paragraph of this review, that might not be a bad thing. There is a marked dearth of books available right now for game and graphics programmers that don’t deal with DirectX. For now, however, the book’s worn out its welcome. **