Ninja Gold!

Nothing much sucks worse than when Hollywood tries to mix it up with new kid on the block, Teh Gamez. I went to a nice talk at last years AGC titled "We want your money!" or "How Hollywood pwns games"..or was it "We can make your business just like ours!".

No, No that's not it. Ah yes, "Hollywood and Videogames: A Marriage Made in Hell". Even better. See Hollywood types like to say things like "What if we, like, put controls in all the seats and then people, could like click and junk and stuff would happen on screen!" If it weren't for Bill and Ted's Excellent Adventure, I, II, and III I'd think these people were off their rocker. They obviously know how to churn out oodles of crap and spend 300 million on the abortion that is Spidey III: MJ is the New DP!... wait, wait, my title logic is definitely skewed today. Well anyway...

I was wrong! Yeah not the first time I hear you moan. Well, listen up. John Woo, who knows more about making a good action movie than Chuck Norris times Steven Segal is teaming up with a guy you might have heard of by the name of Warren Spector to create a movie and a game at the same time called Ninja Gold!

I'm not too sure about this Spector chap but he gets mad props in the biz so I'll defer. Sounds like a stage name to me, kinda like Stone Phillips or Rock Hudson or Jennifer Aniston. Anyway, my bets in. Sure. Fired. Winner.

Link

I Need This!!1

Lately we've been looking for some middle ware on which to base our next terrain editor. We've got some pretty steep requirements. Need to build a half scale map of Europe with several thousand towns with roads and rivers and satellite digital elevation data as the foundation.

In the past we've looked at many editors but nothing really designed to do what we need. I want it to be automated as much as possible. I'd like a cross between Sim City and Combat Missions. I don't need what most MMO engines have and shoe box shooters can't make stuff on the scale I need. Take a look at Blueberry or Terrex for something close for sim applications. Trouble is I don't want to use satellite pictures and I really need to have playable spaces out of all this.

Those guys have some cool stuff. I think we can work something out of it that might give us quick production and good results. Then I see this:

1 - Good screens of progressive generated interiors with objects.
http://forums.introversion.co.uk/introversion/viewtopic.php?t=566

2 - Mostly chat about the game history:
http://forums.introversion.co.uk/introversion/viewtopic.php?t=571

3 - Procedural cities. Code generated in a few weeks into the project....AMAZING! http://forums.introversion.co.uk/introversion/viewtopic.php?t=586

4 - Using the same basic algoritrhms from street and river and building placement you can do interiors as well.
http://forums.introversion.co.uk/introversion/viewtopic.php?t=600

5 - Chicken and the egg, need a dbase for the buildings before you can define the rules to build them...DOOM FTW!
http://forums.introversion.co.uk/introversion/viewtopic.php?t=683

6 - The buildings and streets...
http://forums.introversion.co.uk/introversion/viewtopic.php?t=733

To turn a phrase...F. Me.

1.27 Features Vlog

Doc on the Vblog

Wow...Snellman

So I've been installing Picasa from Google on my rig today and it has this amazing feature to search through all of the images on my rig very easily. This one caught my attention. BTW Lum and SirBruce, you are both free to hit me with that cane at GDC this year is you so feel.

.report goodness

A little something from the report log today:

2007-04-30 15:38:43 arland French 43e Régiment d'Infanterie volkie mission Charleville
"PLZ GIVE STAT HE TO AA ENOUGH TO SEE STUKA STAYING ALIVE AFTER 5 HITS"

Actually statHE would be terrible in that application. Statistical HE determines targets and then shoots shrapnel at those targets only, from the impact point to the center of the target with a conical distribution based on the rough size of the target.

Problems arise when the impact point or detonation is inside the vehicle because the stat HE model will pick the center point of the vehicle at which to throw shrapnel. It doesn't understand individual hit components because that would defeat the computational savings of using a statistical model over a radial shrapnel generation model so instead of hitting lots of things it would just send a small amount of shrapnel directly at the center of the vehicle and might not do any real damage.