"Many of the new action games seem to be based on existing technology, such as the Quake II and Unreal engines. What drives the decision to utilize the strengths (and accept the limitations) of an existing engine, rather than custom-developing a new technology to suit the particular game"

Submitted by: Merlin, the Happy Pig

Answered by: Gabe Newell (co-founder, Valve Software)

"I can't comment on the industry as a whole, but the model we try to follow is to first understand what is the gameplay and experience you are trying to create. You then try to understand what your various options are for being able to support that - whether by building from scratch, using a pre-existing engine, or modifying an existing engine. It's also very important to look at the toolset and it's impact on the content creators, because you can have an engine that will do what you need for your game, but it may just take too long for your artists and level designers to build the content given the constraints imposed by the toolset.

"In the case of Half-Life, we had plenty of work to do in the areas we thought were important to the game (animation, AI, the interactivity of the levels, ...), and John Carmack's existing work on BSP rendering, his networking model, and so on were more than adequate for our needs. The licensing terms were reasonable, so that we could focus on what makes Half-Life interesting technologically and as a game rather than trying to play catch-up to all of the work John had done."

Read more questions and answers!