| |||||||||||||||||
![]() |
![]() |
![]() |
|||||||||||||||
![]() ![]() ![]() ![]() ![]() ![]()
|
Evil Dead Developer Diary: Week 2
![]()
The breeze is hot, tepid. Not the cool ocean air we're supposed to have here on the West Side. Sweat drops from my brow to the keyboard and I'm wondering "why does the old man keep saying the same thing over and over?" Is he trying to tell me something? Something about instructions I've forgotten, a letter misplaced, a number reversed... Ash is here with me now, in the room chanting some crazy incantation that we both know isn't going to work. It never has worked. Suddenly I realize - I'm not in hell, the HVAC just shut off two hours ago. News flash from the front. Evil Dead: Hail to the King is some number of days out from Beta on three platforms: Playstation, Dreamcast and PC. There used to be a sign on the door that was counting down. Someone or some thing ripped it down. Can't say I blame it. The testers are beating up on the game like a gang of Deadites hopped up on caffeine and candy bars. But just like Ash, the code keeps coming back for more. Finishing video games is one of those cathartic struggles that brings people together. There's a bond that develops within the team that isn't found many other places. We all know that without each other we can't make the next build, and we know that if each of us doesn't give it our all the game won't make it's ship date. This is the best team I've ever worked with some of us have been on the same projects for a decade or more, and the newer additions only serve to up the ante. Reset. So now everyone wants to know... is this game going to suck? We've all gotten so jaded over the years, and so many products don't live up their potential. After many years working in this business and in other parts of the entertainment industry, I know one way to tell for sure whether a product I'm working on is going to be good or not. If after a year (or more!) of working on the same project, the whole team is still so jazzed on making the product great, and you can see people in hallways sitting around PLAYING THE GAME and not just going through the motions to test the latest features then the game is good. I won't say great, you'll have to play it for yourselves and decide but the game is good. I like this game and by now I should be thoroughly sick of it.
My job has been to pull together the best and brightest in each area of this Renaissance profession we call the interactive entertainment industry. (How appropriate that Sam Raimi, Rob Tapert and Bruce Campbell are the founders of Renaissance Pictures our partner in this project.) At the same time, I'm still slaving over a hot keyboard building my parts of the game. Over the next few diary entries, you'll be hearing from other members of the programming, design and art teams. They'll give you more specific information on how we pulled this game together. The Code The game engine is a cross-platform engine that already has hooks in it to support future next generation platforms. We've built a system of level and character editors that store their data in an SQL database (using Microsoft's SQL Server and ODBC interfaces). This data is combined with information extracted from Alias' Maya 3D system to build platform specific "streams" that contain all the data needed to run levels in the game. Each "stream" is actually a memory image customized for the platform to reduce load times to an absolute minimum. Our decision about what languages and development environments to use was 100% based off of what we thought would be the most stable (i.e. had been in use the longest at either the hardware manufacture or at early developers):
One of the more unique components of our game engine is an AI scripting system that is based on the Java language. We use Microsoft's VJ++ development environment, but have a post-processor for the compiled programs and our own execution environment that is integrated into the game engine. This allows us to take advantage of all the many advanced features of both Java and of VJ++ while eliminating the vast majority of the CPU and memory intensive parts of the off-the-shelf Java binaries and execution environments. The Art Pipeline From an art standpoint, we are a feature film visual effects company that has adapted itself completely to the production of real-time products. That's where most of our crew comes from and it's where my background in computer graphics is. It is my firm belief that the last 20 years of experience that the feature film industry has with creating digital worlds and characters is 100% applicable to the interactive industry. Those companies that have decided they can re-invent the wheel better have all failed to produce even remotely cost or quality effective proof of those "new" ideas most of which are just rehashed ideas from five, ten or even fifteen years ago that these would be pioneers are simply unaware of. We have an art team that is segmented into to modeling, animation, color & lighting, painting and compositing as separate disciplines. Obviously, there is some cross-over but we tend to compartmentalize and pipeline that art process fairly heavily. We primarily use the following tools: I've got to get back at it. I see a huge deadite kicking the living $%# out of Ash. The player should have some chance after all.
Peace Out. Steve Gray
![]()
|
|