[longiron]

DAve Babb, on 17 November 2011 - 12:46 PM, said:

But for the future, when OR moves away from pure MSTS compatability I'd like to suggest the following as features that would be nice to have...

Access to "Coupling state" of a bit of rolling stock, and the option to have animated coupling's to show for instance chains connected.

Interlocked door operation. The ability to prevent doors opening unless the consist is stopped brakes set and prevention of the brakes being released without doors shut.

Some form of 'suspension' effect, even if its just a representation of a disconnect between the bogie and the body to allow a bit of body shake / sway.

Controls in eng file to run animations such as cooling fans.

Dave,
Good ideas. Some of them like "interlocked doors" and "ENG file parameters to control cooling fan animation" are much easier to accomplish. Suspension effects and coupling state require significantly more development to do them right.

All will be added to the TO DO list

[Matej Pacha]

longiron, on 17 November 2011 - 02:35 PM, said:

Suspension effects and coupling state require significantly more development to do them right.

IMO, any suspension effects will consume a lot of CPU, several times higher than now. Consider that the suspension system is always a spring with a damper in parallel. I hope we can do some simplified models in future but if you mean some kind of real simulation, please be careful: You can picture the whole chassis of a locomotive as a system of 16 + 4 suspensions what is a pretty complicated system of 40 2nd order differential equations. This means that it is very hard to keep a simulation like this always stable. And what about wagons? If the train consists of 40 or more 4-axle cars, it's about 1600 or more equations and I'm not sure that it can be solved in real time within 0.1 second time step or less.
Anyway, to simulate it properly we have to have more information about routes (e.g. rail elevation in curves), to simulate all forces more precisely (tilting effect) and to have always-stable solvers for differential equations. What more - every model should have a set of suspension system parameters and I'm not sure that players or modellers can adjust this kind of stuff.
Finally, I'm quitting with negatives. We are working on some tilting simulation and I'm sure we can make a locomotive frame-bogie connection elastic a bit so you will see the frame lifting when applying a tractive force.

Matej

[Genma Saotome]

I agree w/ Dave -- if a form of rocking can be done with a very simple method then perhaps that's all that is needed. If it turns out that comes with a high price tag in performance, then maybe it is not such a good idea after all and should be withdrawn.

IIRC there is another thread somewhere in the OR forums that mentions two speeds at which rocking most often occurs, so if there is anyone who wants to try this, there is a bit more info on the matter.

[atsf37l]

Two things I'd like to comment on: Motion and interlocked doors.

For motion, you have an excellent resource on your team in the person of Barry (Captain Bazza) Munro. His 'Bazzamotion' techniques have his narrow gauge C-16's and K-27's and their tenders bouncing along over the rails in a very effective animation. Cabooses too. The bodies rock independent of their chassis/trucks. These animations don't seem to hinder the frame rates in any significant way.

As far as interlocked doors I vote to either eliminate the interlock or make it an on/off (0/1) option. Classic case is the RGS Galloping Goose from Tom Werb. If there wasn't any snow on the ground the geese ran with the doors open. Also the operator would frequently climb in the goose, release air, shift into first and take off and THEN close his door. Likewise, the doors would be popped open while the goose was still coming to a stop - remember, most of us work in a more nostalgic and not so regulated era! With those rigid OSHA complient interlocks this operation would not be possible. In MSTS it is bad enough that you have to be stopped with brakes set before the doors work. Fortunately Tom has built in a nice delay on closing the doors so you can Ctrl[9] Ctrl[7] and then start out and the doors will close about a goose length down the track. :oldstry:

[Genma Saotome]

DAve Babb, on 19 November 2011 - 11:07 AM, said:

As for coupling states, I'm thinking that we already know if the vehicle is coupled or not because the light settings are already aware of it.
I think it would be nice to have maybe just a 2 stage animation that shows the coupler in one of two positions. You could use it to couple air pipes, lift chains to hooks, even position tail lights( although it would be nice to be able to script this in the wag/eng file, so it would be easy to change the type of light based on the time period etc).
Not sure you'd need to actually fluidly animate between the two different positions, an "instant" change would be acceptable to me.

I've proposed the idea of "assembly" of multiple shape models to make a single car: car body, couplers, and trucks as separate models, which would allow easy reuse of the coupler and truck models. One could do doors too I suppose. Anyway, if the most popular coupler model has animation built in then the appeal of reuse of that model increases rather significantly. For myself, I'd be happy with animation being limited to moving the coupler in and out, representing slack and having that slack have an effect on the physical location of cars. Consider most North American couplers move a minimum of 3 inches; on a 100 car train that's 25 feet of slack the must be pulled out before the last car moves.

That said, being able to see the coupler open and close, see the airhose in a connected position and dropped... that'd be pretty cool too.

[CRQ5508]

The number one thing I'd like to see in rolling stock is harmonic rocking. Except on the best maintained track with deep, smooth ballast, good ties, and continuous welded rail, cars always rock from side to side. Running several routes in MSTS that I know have less-than-ideal track in the real world, the fact that the trains are running along very smoothly really kills the effect. If there was a way to control sway from route to route so you can have just a little sway on some of the better maintained lines, to rocking that threatens to derail the darn train on other routes, that would be even better (I've seen quite a few of the later in my time.) Now, I know it's very hard to actually model bad track in MSTS, but maybe the implementation of harmonic rocking could be put on the rolling stock (like pivoting on the trucks) but still somehow the overall amount of rocking can be affected from route-to-route. Maybe something along the lines of a slider switch (in lack of a better term) where you can go from an off position, completely removing the rocking, to a "high" position where the cars are swaying back and forth several inches in either direction. It will be really nice to see a train in MSTS where every other car is rocking in the opposite direction to the car in front of it, especially if the rocking matches up with train speed. :oldstry: