It's all in your head, Part 6

[BrianBlessed]

I've hoping it's like an contemporary urban Populous, whereby you vie for Mafia like control using drive by attacks and so on. The game comes to a climax when the two tribes battle it out in a massive citywide street fight.

[Kadayi]

But if it will a strategic game, why IV is creating their Office generator? Indoors don't count in RTS'.

You've never worked in an office apparently....

[martin]

Looking cool, I'm beginning to wish that I lived in London now

Anyway, all this procedural stuff is something I'm looking into for a top secret game me and Tom are working on, it's like a crime/Heist game thingy but requires either 1E6 hours of map building or some procedural stuff

Oh, and for those who can't view that format for no good reason, here is a google video mirror, looking cool Chris

http://video.google.co.uk/videoplay?doc ... 5885926465

[KingAl]

Looking good. It's fair to assume then that the game takes place in floating environments rather than fixed (why build a generator otherwise)

Not necessarily. It saves time over hand-crafting every single level, lovingly shaping every single building in the tens of square kilometres of city in each level or area of the game.

so it's definitely going to be strategic in nature, Corporate subterfuge I hope

Again, not necessarily - Uplink had many generated aspects without being what I'd call a 'strategy' game.

[aiusepsi]

I hope you guys are presenting again next year. It was killing me that I was sitting mere tens of metres away in a lecture in the Blackett lab (Physics building at Imperial, for those who don't know) when I could have registered for the event and gone to see your talk.

It was heartbreaking, it really was.

[Wasgood]

Urge to give intro money rising.
Will it involve nukes?

[d37312m1n3d]

But if it will a strategic game, why IV is creating their Office generator? Indoors don't count in RTS'.

You've never worked in an office apparently....

Because i'm not in that age

I keep trying to tell you people that it's a Cube inspired game with the indoors being a certain level/s and the outdoors another level/set of levels.

Wherefrom do you know that?

[Lowkay]

Howdy,

The city generator is awesome... I was at the GAME event and i have to say that it was one of the more inspirational presentations -- a really excellent live demo on a very interesting subject. I think the motivation behind it all is the key, it's often that games are bogged down by all the spangly graphics and artwork, but the seriously great games don't need any of it, graphical design is one thing, game play is in a whole other league... More importantly is that it still looks good with minimum effort (perhaps not minimum effort in that you have to create the tools to do it first) but once those tools are in place the shear amount of unique generation you can get from them makes them far more resourceful than texturing everything specifically for individual games with hours upon hours of artistic development.

I salute your genius, without developments like this games would just be dull, we'd be living in a world of textures not gameplay...

And remember content *is* bad!

[Bad Intel]

This is some fascinating stuff. I'm getting the feeling IV has yet to write its Revolver.

This generator business seems too compeling to only be used to make levels for whatever Subversion will end up being. I'm ready to bet that it's going to be key to gameplay somehow.

How about loading up actual city data into the viewer and running some comparisons with what you guys are doing? That would make a nice kind of Turing test analog for your algos.

[KingAl]

This is some fascinating stuff. I'm getting the feeling IV has yet to write its Revolver.

This generator business seems too compeling to only be used to make levels for whatever Subversion will end up being. I'm ready to bet that it's going to be key to gameplay somehow.

Well, in the past IV has found the greatest challenge to be creating content, and this considerably speeds up the process. So while it is possible that they have another application in mind, it would be entirely reasonable to have created it for this purpose alone (Can you imagine how time consuming manually creating every single building and road to create a convincing city would be?)

That would make a nice kind of Turing test analog for your algos.

"We only wanted to generate cities, but now it can think!"

[Nutter]

It looks interesting, it almost got a slight crackdown meets uplink look to it (dark blue and slightly cell shaded).

It could be nice if once you've generated a level you'd get a generation code for it. Entering that code into the generator would create that exact level. That way people could make cities and post them to each other ("hey I just created L.A., here is the code for it" sorta thing). I have no idea how much of a impact it would have on the actual gameplay or how hard it would be to do, but it might expand the time people play around with the editor and something to do none the less.

[Lowkay]

I wouldn't be surprised if this generator were to be simply that, a generator. It's kind of a template for production, a tool for content, it has no meaning beyond its purpose to generate and it doesn't need to have anything more... As it stands it doesn't look to be developing into a generation and play game such as sim-city, it's been done before, no, it is far more likely that what you see here is a fantastic reason not to get pent-up on graphics, to take your time and energy away from production and into conceptual design. It's in the unique and compelling ideas that make the games, not in the environment design.

Plus this little beauty of a tool is worth its weight in gold (probably more considering it's basically code on a computer which doesn't weigh much), as it can be used again and again in different scenarios (perhaps with a little modification). Once you have something great enough to produce a city, it'll be a vital stepping stone in producing more fine grained detail procedurally - and that can be used in any game concept that you care to wish for!

[KingAl]

Exactly what I've been trying to say. Just because they've made a generator does not mean any generating will be done on the end-user's machine.

[martin]

Just because they've made a generator does not mean any generating will be done on the end-user's machine.

True, but if you could make the campaign scripting dynamic enough it would offer a different campaign for every user, and an infinite amount of slightly varied campaigns for every user. Somewhat like the LANs in uplink.
And of course it's far more efficient to store a generator seed than an entire map.

[KingAl]

Just because they've made a generator does not mean any generating will be done on the end-user's machine.

True, but if you could make the campaign scripting dynamic enough it would offer a different campaign for every user, and an infinite amount of slightly varied campaigns for every user. Somewhat like the LANs in uplink.
And of course it's far more efficient to store a generator seed than an entire map.

My main objection was to suggestions that people would be able to make their own maps using the generator, which seems highly unlikely.
The fact that Chris specifically designed the generator to allow intervention at any point suggests against your former suggestion, in that it makes it more likely to be an aid to creation rather than a self-creating mechanism, and while procedurally generated content is certainly more efficient in terms of space used, unless it loads on the fly it can be prohibitive in terms of time and processing power used - Uplink's LANs are considerably simpler in this respect as they do not have to generate a 3D representation, nor are they of a comparable scale. This is not to say that IV is not going to take this route - I'm just irritated by the way people have jumped to conclusions regarding the nature of Subversion's content.