quote: Original post by Isowave
I've got both books, so I'll take a look. I'm trying to work my way up to Real Time Strategy Game Programming by Mickey Kawick... that seems to be a ways off though.
Although I don't want to put you off buying it (you'll never know about it for sure until you read it), bear in mind the following points:
You don't get a complete game engine in the book. He presents lots of utility classes and you have to work out how to piece them together. If you prefer the LaMothe method of "Take existing engine and modify" then you may be frustrated.
The code is bugged in several places, and will not compile. I am told the author is more than willing to mail you updated, working source code if you email him.
Much of the code is hybrid C/C++ (typical of someone in the industry who started with C and has had to evolve slowly while completing projects). For instance, he makes some sort of memory management class which would probably have been done better had he overloaded the new and delete operators.
Much of the source code is duplicated! He actually reprints a class just to remind you, rather than saying "refer to page XX". Not to mention that a lot of his classes just duplicate STL functions (IIRC, his taglist is just like an STL bitset, for example.)
He doesn't adequately cover animation or the separate storage of characters and character types, hence my thread on this a few weeks back.
He oversimplifies the A* algorithm so much that the explanation misses the whole point of it (ie. the use of a heuristic added to a 'distance travelled' amount to choose which is the next best path to use).
...etc etc. It is not the holy bible of RTS or tilebased programming, sadly. In fact, the stuff I've learned from that book has proven more useful in my text-based MUD than in my tilebased games!
Edited by - Kylotan on 4/26/00 5:42:07 AM
