Michael Abrash's Graphics Programming Black Book (http://www.amazon.com/Michael-Abrashs-Graphics-Programming-S...) includes a good sized chunk written about Quake as it was being developed. Even though the book is now about fifteen years old I still count it among my favorite programming books because it is the only one I've found that decently narrates the mindset of optimization, of turning a problem on its head in order to find a better solution.
Because the book is essentially a series of collected magazine articles the section on Quake development will have one chapter that reads "hey, this is what we're going to do!" followed by "well, that didn't work, but this does!". One important aspect of post-mortems and development journals that I find is frequently missing is a detailed look at what didn't work technically and why, which it has in spades.
A lot of the stuff is still in use - one of the best selling video games recently (and many others) are still using this tech to some degree (brushes, bsp, and csg).
The vast majority of the book is dedicated to optimizing Pentium-and-earlier processors as well as EGA/VGA programming modes, with a smaller but still significant chunk on software polygon rasterization. There is value in the bits on BSP and visibility, but it's probably only a third of the last ten (of 70) chapters. That said, there's probably more relevant worth in those bits than any Game Programming Gems books :).
Abrash is a great writer and, even if you skip the deep technical dives and code listings, the book is still worth reading.
Comments
Michael Abrash's Graphics Programming Black Book (http://www.amazon.com/Michael-Abrashs-Graphics-Programming-S...) includes a good sized chunk written about Quake as it was being developed. Even though the book is now about fifteen years old I still count it among my favorite programming books because it is the only one I've found that decently narrates the mindset of optimization, of turning a problem on its head in order to find a better solution.
Because the book is essentially a series of collected magazine articles the section on Quake development will have one chapter that reads "hey, this is what we're going to do!" followed by "well, that didn't work, but this does!". One important aspect of post-mortems and development journals that I find is frequently missing is a detailed look at what didn't work technically and why, which it has in spades.
edit: The book is also available as a free download: http://www.drdobbs.com/high-performance-computing/184404919
A lot of the stuff is still in use - one of the best selling video games recently (and many others) are still using this tech to some degree (brushes, bsp, and csg).
The vast majority of the book is dedicated to optimizing Pentium-and-earlier processors as well as EGA/VGA programming modes, with a smaller but still significant chunk on software polygon rasterization. There is value in the bits on BSP and visibility, but it's probably only a third of the last ten (of 70) chapters. That said, there's probably more relevant worth in those bits than any Game Programming Gems books :).
Abrash is a great writer and, even if you skip the deep technical dives and code listings, the book is still worth reading.