I know all the solutions that you wise people have come up with, are not useful in my case.

Dan ,

It's all about re-using the same resource many and in different ways. If I can create a level with say 5 models and have say use it for 1000 different situations then the solution you are saying would need 5000 individual models. Think about the loading time and HD space needed for those models . And dont forget the time I would have to spend modelling it. And if situation arise for changes then all time that would be wasted on tweaking that model.

And when I say collision detection would be faster if the model involved has less polygons. That means only the models whose AABOX is in direct intersection, is checked with polygonal collisions. Try to create a big level with all geometry in model, and the same level with smaller models, then the level with smaller models or parts model should run faster .

If you still cant see what I am talking about then ask and I will post a picture for your refrence .

ulillillia,

It's a solution that wont let me see the level in real time .

ventilator,

You have got the idea what I am talking about . But you are wrong about the modeling program and as such. Because I see things differently and will like to do as I see fit .

Btw thanks all of you who are trying to solve a situation that needs only a minor feature change on the part of conitec. And that would be to let the user have the freedom of scaling his models as he likes, uniform or non-uniform.

I never said that the model would have only 8 or 80 polygons. And believe me when I say that collision detection would be faster for polygon-precise collisions.

In a game the collision detection is the most time consuming part that slows down the game. And where there are many polygons that are used, then the time to check for every polygon that might collide, is great.

I think that, instead of giving advantages, I should have only asked for this feature to be implemented. Because all the interpretations that have been done, are your own specific point of view and vague in nature and dont co-relate with what I am putting forward.

An example what this feature would help in, is people like small demo downloads and instant play startup.

I would like to voice my request once again for the feature. We should be able to scale in either direction - x,y,z. As 3dgs is going for the, all model level. we should have the feature to scale in even a single axis.

Well, as for most of the purposes such a feature could have, do as ulillillia suggested ... simply add that script, change the skills and voíla ... only disadvantage is that you can't see it in action straight away, I give you that. So yes, for aligning models by scaling them, this is not the right solution. I think a model scaling in WED is somewhat useful, especially for fences or model walls. Since we can tile skins too, this would be pretty neat indeed.

However, for random things like your example rocks, the script Uli provided is good enough, isn't it?

Cheers

It's for situations for creating a pile of rocks, or mountains. Roads of rock for mountain area or the very well known american rocks pile(I dont know the location name). It's for situations where precise scaling is needed to adjust more than one models. Like ledges in the mountains. I think these are enough of examples that would persuade conitec to implement the feature.