Atomic Systems

Many questions from a route developer about tools/features for development:

(1) Will terrain projection be using the Goode Inverse Homolosine projection as does MSTS (a skewed quadrilateral), the now commonly seen Universal Transverse Mercator used by USAPhotoMaps and GoogleMaps (a real square), either of those (by user choice), or something else?

(2) Will the distance of visibility be fixed as it was in MSTS (at 2000m) or will it vary depending on the size of the object -- for instance a large steel bridge is visible at a distance far greater than 2km while a small house vanishes at a lesser distance. Also a consideration for distant mountains.

(3) On seasons and precipitation: IMO one of the great flaws in MSTS was the failure to seperate terrain textures for seasons from terrain textures for precipitation; will the new sim have 4 seasons, each having 3 precipitation types (i.e., sun, rain, snow)? Essential for those locations where you get seasonal rains but never snow.

(4) On water: IMO another big failure in MSTS was making water textures global to the route. Ocean and pond water do not look the same. Water should be just another terrtex assignable to an area. Will the new Sim offer greter flexibility in this regard?

(5) Terrain Sculpting: Will there be a variety of shapes available for editing terrain (e.g., circles, spheres/cones, rods, etc.)?

(6) Object placement is Cad work by another name; will there be normal cad functions like snap-to, place-parallel, place-perpendicular, etc?

(7) MSTS grade specification is an ergonomic crime. Will the UI allow keystroke entry of digits for grade? OH, FWIW, the default grade should be the same as the section snapped-to, not level as grade changes are always relative to the preceeding section.

(8) Distant mountains in MSTS worked well except for one situation: where the foreground was both flat and extended further than 2km (e.g., looking across a large lake or bay). The DM display would then greatly shorten the forergound. Will the new sim accomodate this situation in some way to keep those hills as far away as they actually are?

(9) Rail barge/ferry is conceptually alike to a train moving under it's own power...just slower and tied to another object. Might the new sim allow for this feature?

(10) Lastly, IMO a train is a train is a train -- they're all operating under the same basic rules. Allowing the player to move from one to another (or to a static location like a tower) would greatly increase the diverse experience of playing the game. It also, architecturally speaking, lays out the logic for multiple players. Is anything like this under consideration?

Some good questions there.. you should get some interesting answers... eventually..

Isn't Genma Saotome a giant martial arts Panda ?!! Good points Genma. Put this big Panda on the payroll.

> Isn't Genma Saotome a giant martial arts Panda ?!!

Yup. And when I'm not Genma I'm an expert level enterprise data architect.

Two more questions came to mind...

(11) You may recall in MSTS there are file load points 1000m in front of the nearest tile approached by the camera, a distance which happens to be within the visibility range. Might the new sim be smart enough to do file loading well in advance of the visibility range? FWIW using something akin to a upfront compilation (rather than reading .w files as you go along) could have allowed MSTS to attach a relationship link between the track shapes of a path to one or more objects to display where the choice of objects was determine by their size, distance, and location ratio in the route rather than some arbitrary distance attached to a model found in many routes. That would also spread the workload of those file accesses.

(12) Can snow be managed as a vertical offset layer identical to terrain (offset specified by user and vertex editable)? Invisible of course in non-snow seasons.

Thanks in advance for any reply.

GenmaSaotome wrote: ... an expert level enterprise data architect.

I'm not sure what that is but sounds impressive

There are some excellent points here:

(6) Object placement is Cad work by another name; will there be normal cad functions like snap-to, place-parallel, place-perpendicular, etc?

Is very important. But needs two variations:

(1) place parallel or perpendicular to (x, y, z) for platforms, fences, etc. (Must be applicable to roads as well as railways - no fun fencing roads without being able to use gantries in MSTS... actually it would be good to have gantry placement on roads to place street lights telegraph poles and fences or hedges.

(2) place parallel or perpendicular (x and z only) for buildings. The y axis of buildings needs to remain plum even on gradients if we don't want lots of leaning towers!

Regards

Darwin

I thought it had been made clear by Kuju that this development was a "fresh build" in terms of how things work and are designed...in fact everything!

Why, oh, why is everyone asking things like "Will it have the same xxxxxx as MSTS?" or "Will the xxxxx limitation be fixed?"

There will be things that will look and feel the same as MSTS.....or Trainz....or BVE. But it will be a similarity..not a development of.

Im pretty sure that the 2km viewing distance was put in MSTS so that more computers could run the game. We have to remember that MSTS was made when computers where still slow and not extremely powerful compared to today. The limit of 2km would have aided graphics cards as they would not have had to have loaded so much at one time.

But, as today there are much more powerful cards available and faster PC's, im sure there will be bigger viewing distances.

Already from what we have seen, the sim certainly seems to be extremely detailed. The only thing we can do is wait and see what the final product will be!

> ...2km viewing distance ...

Yes, it was probably done out of concern for hardware performance, probably to avoid the display of so many terrain polys.

That said, it was still done incorrectly. Take this example: two of the world's tallest buildings are in the city of Chicago, on the shores of Lake Michigan. From the tops of these buildings, on a reasonably clear day, you canlook eastward all the way across the Lake to the state of Michigan, a distance of over 100 miles. Howver, if you are in Michigan looking west towards those buildings, you'll not see them. A simple reason: the state of Michigan is huge and therefore can be seen at a distance of over 100 miles, whereas even remarkably tall buildings cannot. This means the distance of visibility is a function of the object size, not it's distance from an observer.

To apply this logic into a train sim means having the software determine how far it can be seen rather than asking the model builder to guess (or accept some default value).

At any rate, I think it is quite fair to hope that with a new sim some thought is being given to this issue because just "doing what we did before" isn't very good.

I'm not sure what that is but sounds impressive

Well I didn't want you all to think I was a Panda full time. FWIW, what I do is help companies sort out the mess they've built into their business data. You'd be surprised how utterly convoluted the data can be can be in a big corporation for concepts like customer and product.

...this development was a "fresh build"...

Yes, of course.

I could have written "I want the new sim to do XYZ" but instead chose to put in some context the reader would recognize and ask "You are not going to repeat this boneheaded decision from MSTS are you?" as I thought that way might stand a better chance of earning a reply.

Sorry if it bothered you.

it'd be interesting in america as the naked eye can see for 50km on accurate flat lands like dessert's on dessert in particular between mountain ranges is Nevada which i think, correct me otherwise is 30km of flat land standing at the one end the Sierra Nevada mountain range is visable on a clear day so in reality i hope that the new sim can do alot further than 2 or 3 km.

If you think about it a bit... a tall building... 200m high should be visible at least 2km away. A ratio of 1:10 height to distance. In truth it might be visible 20km away, a 1:100 ratio. Whatever the actual ratio turns out to be it's easy to "see" that the 2km limit that was used in MSTS imposed great limitations on the viewing realism in the game. Think gas storage tanks, blast furnances, large bridges, tall smokestacks, tall buildings. There are lots of structures in the modern world that are visible for far more than 2km.

Let's hope the new SIM can provide much more.


p.s. you mentioned the Sierra Nevada. A bit further west, not far from San Francisco, is a solo peak called Mt Diablo, about 1000m above sea level. But as there is nothing else tall nearby it provides an amazing vista: on a really clear winter day you can see 250km -- from Mt Lassen to well south of Mt Whitney. On the rarest of clear days, 325km. That's about 3/4 of the state of California. It's amazing.

The downside is building more than 2km away from the line.
Distant mountains (Lo Tiles) was intended to be an easy way to do this I suspect.
Terraforming from DEM data will hopefully take care of the landscape in the new sims.
But how much detail will we realistically need to put in to places to the left and right of our railway line??

steampsi wrote:

GenmaSaotome wrote: ... an expert level enterprise data architect.

I'm not sure what that is but sounds impressive

Is it legal????