Not saying that it isn't, but I'd rather have this.


Regarding Carmacks approach, I've read about it and find it interesting as well. The one 'flaw' as I see it is the inability to create space. E.g. a cliff leaning out over the ground. Or is there some way to overcome that?

Notch, the guy behind Minecraft, also roasted Unlimited Detail. His takeaway point was "be excited about voxels, but not about UD." Then he pointed to somebody doing the voxel thing much cooler than the UD guys:

http://atomontage.com/

He's got procedural generation of the landscape, and it looks awesome. He's actually more interested in physical simulation of his atoms than on the convenient side effect of it looking cool. His stuff isn't copy-paste repeats. It's all deformable/destroyable. And if you find the right demo videos, you can see that he can do very large areas -- possibly not unlimited, but "past the horizon" at least. He's already got it interactive where he can drive around in his landscapes leaving ruts.

His smallest voxels are larger than the UD guys, but he could always procedurally fill in around them in the necessarily small viewing area you would have if you got down close enough to see. But if you're walking along 5-6 feet above ground level, you're not going to notice anything but awesome ground.

Now, everything he's doing so far as I've seen is different kinds of rock and dirt, and "concrete" buildings. Gotta get vegetation and water in there somehow. But I still think his approach is much cooler than UD.

Just to be clear, Notch is not the developer of Atomontage -- he just likes it.



Also to clarify my wording: Atomontage takes the same basic technological approach as UD (and anybody else doing serious large scale voxel rendering, for that matter -- it's a well known basic technique). But Atomontage does it better, larger, with more real features that you would need in a real game.

Imagine laying siege to a walled city where you have the option of breaking through the walls or tunnelling under them, at your discretion. Imagine trying to defend against those real threats.

Such things are all well and good but Daggerfall is a game created back in 90's. Besides detail, graphics, and exploding walls and such are all pointless and useless if the core gameplay sucks and is boring.

Rather play Baldurs Gate then Crysis, yeah Crysis has DX10 and fancy glowing graphics and certain things blow up and physics etc. But Baldurs Gate has more fun and in depth gameplay and even with it being an older game it still has a charm that will last for ages to come compared to Crysis which was equivalent to a Hollywood summer Block Buster. Made to make money and not a long lasting impact that revolutionizes the gaming market and becomes a catalyst which games look up to being.

Personally I find the Unlimited Detail thing like cloth physics back in the day. It will just become mediocre and standard in gaming at some point sure. But will it be revolutionary and something great? Doubtful.

It's like Skyrim am I going to just stand their and be like "oo look at the butterflies, look at people walking around cutting wood!". Yeah I might do that for a little while but if the game doesn't draw me in and the gameplay sucks. The fluff and outer portions of the core game aren't going to make it a lasting impression for generations to come.

Why do people try to breath life into old games? Cause older games have a longer and more lasting impression even though these games have older graphics, sprites, and things that are considered "technologically inferior" to modern games sure. But does that mean these games are bad? Far from it.

I agree wholeheartedly. I only look for new graphics because I know people that I wish could enjoy the depth of the older games, but they can't get past the graphics.

Voxels are a slightly different issue. I get excited for the same reason I get excited about the holy grail of real-time ray-tracing (or photon mapping, or whatever). I'd like the game engines to eventually reach a point where the technology "just works" and doesn't require a Carmack-level genius to squeeze the performance out of it. Each step that takes us closer to the holodeck level of game creation makes me happy. Don't get me wrong -- the impressive part of the holodeck to me was NOT the holograms and forcefields and yada yada. The impressive thing to me was "give me a story in the style of Sherlock Holmes, which I will play as the character Sherlock Holmes" and it goes off and does that. That's just magic technology, I realize, but imagine if the technology could ever reach the point where you could feed a book and some maps into the computer and it could generate a simulacrum of that world, which could then be tweaked up instead of having to be manually created from scratch. That's not quite as totally crazy. The best we can hope for is somewhere in-between, but every tool that moves us closer to virtual worlds that just do what you expect makes me happy.

But yeah, the game's got to be there first. That's actually the problem I have with Tamriel Rebuilt for Morrowind. HUGE landmasses, cities, people -- and NOTHING TO DO. OK, that's a bit harsh, but there are no quests yet, and everybody responds with the same generic rumors. Of course, past a certain point, Daggerfall felt that way to me (GASP! Heretic! Sacrilege!) But I liked enough of what was there that I'm really excited to see what it will be like with the option of saner dungeons, and mods and more story and etc and etc.

(Wow, sorry to have gotten so off-topic. Blame it on heatstroke -- I've just been sawing and stacking wood from a recent windstorm...)

It's been a while since I've left the project; but there were quests at least in Map 1 even back then.

The thing is that you have to create things in order. First the landscape, then the buildings, then the people and basic dialogue, and finally the quests. You can't implement quests before any of the other components are there. You can certainly plan them; but not put them in the game as long as all their prerequisites aren't met.

Sorry Gez. I understand all that. And my frustration comes as someone outside the project. I know that there is a LOT of stuff in TR that the contributors see, but the "public consumption release" is pretty much quest-less.

Or at least it says it is, and I've not discovered otherwise. Though I haven't spent a lot of time since TR2 came out, so my data may be old.

We need a topic split here actually to discuss "next-gen renderers" Look at second tech demo of UD, where authors are interviewed by some magazine(don't remember exact name and don't remember where i saw it). It answers to maaaaaaany questions discussed.