I’m still waiting for added rails support – I’d love to build some ridiculous railways through the Himalayas. That said, I dislike their control scheme for creating roads – it is IMO pretty clunky and not very well thought out.
I’ve bought the access though and have had a lot of fun flying around a bit. Be sure to check the forums for additional models, such as the A380 or the Basler Turbo.
Oh no, I agree that the path based modeling is the best way to do it – I just think the interface needs a little work. That said, I have not tried all too hard yet, maybe it just requires a bit more of a learning curve than I’ve given it yet.
Just two small annoyances to illustrate what I mean:
The free camera comes in the way more often than not. An option for clamping the camera to the ground and allowing it to rotate around a fixed point on the ground would make working on the terrain a whole lot easier, I think. So a mode like in any strategy game. This would of course only be for constructing things, not meant to replace the current model for moving around.
I have trouble with extending existing roads. It seems that you can not freely select from which end of the road you you want to continue – in my case it always continued adding new nodes from the far end at the current end, which resulted in a road drawn straight across itself again. This might be down to the system not being finalized yet and may very well get better once junctions etc have been cleanly added, as you’ll need to be able to continue from arbitrary nodes then anyway. I did try selecting the old node from which I wanted to continue, but it didn’t work for some reason.
Of course this might have already changed for the better, it has been a few months since I’ve tried it. Another small thing that has bothered me repeatedly is that it seems to want a joystick or steering wheel for controlling vehicles. While this is surely a boon for any simulation diehards out there, I would prefer a sensible and well adjusted key mapping for toying around quickly.
This is using a global cloud density map with resolution around 5km, which is driving the fractal cloud generator. Both the speed of clouds and dynamics of shape change are adjustable, although the wind speed is global and blowing always to the west.
To add to that, would the clouds be synchronized in a multiplayer game? I guess if they are procedural they should look the same for all at any point and time, unless someone hacks it to remove them in order to gain better visibility.
Run the calculations server-side as well. Cull any packets (related to objects) that are not in view/sensor range. Removing clouds would be removing a visible clue that something invisible might be about to hit you (and it’d save a tiny bit of bandwidth which might make financial sense in the long run).
Why would someone do that when they have a HUD with sensor overlay? It’s not like in a FPS game where you can only target visually – just the fact that you’ll be shooting over large distances necessitates some kind of target marker on the screens. Or did I forget something?