[SP 0-6-0]

Dave, So what is a good fix or should I say compromise for what we've got?

Robert

[eric from trainsim]

It's a game. Stop worrying about global accuracy... ;)

[Lindsayts]

eolesen, on 08 September 2017 - 05:40 AM, said:

It's a game. Stop worrying about global accuracy... ;)

A couple of comments...........

First, Its a simulation, a lot of people in this world like railways, to a good portion of them a GOOD simulation will be the closest they will get to be able to really "play trains", these people will be the ones most likely to contribute, it would be a major mistake to run them down.

Second point. OpenRails original goal was to produce an ACCURATE simulation, this appears to be still the case, with many improvements such as decent steam locomotive performance. I would think an accurate world model would be vital to accurate simulation AND ITS NOT AN UNOBTAINABLE TARGET. There's at least two open source projects that produce such good world models and one of these has really excellent documentation.

I am disappointed that any one would think that in an supposedly accurate simulation that a substandard world model would suffice.

LIndsay

[Metro4001]

And I'm disappointed that an important discussion was derailed because of a spat over a 2 kilometer difference. That is something for the RE thread. You all speak of an accurate simulation, well what is there to sim accurately if we are not able to create prototypical activities? Even though Goku's editor isn't "official," between that and the MSTS RE, there are enough resources for route builders. The MSTS AE is just not suitable at this point. ONE bad character in a wag or engine file and it won't even load. When you have thousands of pieces of rolling stock this is unacceptable. Especially after you've tested them with RR and Conbuilder.

The Trackviewer is the basis of a good AE. Integrate Goku's consist editor and we have 50% of what we need already done. Hopefully someone takes it upon themselves to really delve in like Roeter (correct me if I'm wrong) did with Timetable mode. An incredible undertaking and even though I don't use it (quite complicated) I'm thankful that there's a way for users to simulate actual OPERATIONS in the sim that aren't dependent upon activities made for MSTS. A route editor is important but it seems like some people are more interested in perfect world building than having a way to use the world they create.

[Lindsayts]

Metro4001, on 09 September 2017 - 01:10 PM, said:

And I'm disappointed that an important discussion was derailed because of a spat over a 2 kilometer difference. That is something for the RE thread. You all speak of an accurate simulation, well what is there to sim accurately if we are not able to create prototypical activities? Even though Goku's editor isn't "official," between that and the MSTS RE, there are enough resources for route builders. The MSTS AE is just not suitable at this point. ONE bad character in a wag or engine file and it won't even load. When you have thousands of pieces of rolling stock this is unacceptable. Especially after you've tested them with RR and Conbuilder.

The Trackviewer is the basis of a good AE. Integrate Goku's consist editor and we have 50% of what we need already done. Hopefully someone takes it upon themselves to really delve in like Roeter (correct me if I'm wrong) did with Timetable mode. An incredible undertaking and even though I don't use it (quite complicated) I'm thankful that there's a way for users to simulate actual OPERATIONS in the sim that aren't dependent upon activities made for MSTS. A route editor is important but it seems like some people are more interested in perfect world building than having a way to use the world they create.

You said "A route editor is important but it seems like some people are more interested in perfect world building than having a way to use the world they create."

Now I would be the main advocate for an accurate world model, this is because........

1: I know an accuarte world model can be done as I use one in my own sim.

2: The basis of simulating a real life activity is being able to simulate it correctly, the reason why I started my research in the "world model" problem was I was p----ed off with MSTS's short comings in this area, causing incorrect train times.


If distances are not correct in your simulated railway world how is it possible to generate an accurate activity, that is what I am after, the "world model" problem is just ONE of the steps required. OR's train physics has been good enough for ages to be able to pickup even minor train config errors, what is wrong with wanting the same for the world.

Keep smilng ;) ;)

Lindsay

[Lindsayts]

If you cannot simulate a real world rail "activity" accurately for some reason, what is the point of writing an "accurate" simulator.

Lindsay

[railguy]

Since I used to work extensively with Geographic Information Systems, as well as land survey, I'll make this comment: The level of accuracy required for a representation of geography can and does vary, depending on what the representation is going to be used for. For example, Global Positioning Systems can be as accurate as 1 cm for locating a spot on the earth, or can be up to several hundred feet off. Either level of accuracy can be acceptable, depending on the accuracy required for the representation. Of course, the time and effort required increases as the level of accuracy required increases. So, the debate about the accuracy of the OR geographic "model" should not be a foggy debate about opinions of accuracy, but rather about what level of accuracy is acceptable for purposes of the sim. And, if one really wants to get into a debate about "accuracy," in general, there is NO train simulator out there that is an absolute accurate representation of the route it portrays. The basic limitations of route building and computer requirements prevent it, and likely always will. I'm intimately familiar with several prototype routes recreated in MSTS/OR. The simulation routes go from somewhat accurate to laughably ridiculous attempts at representing the real thing, but none can be said to be strikingly like the prototype. Even great geographic accuracy can't fix that.

Now, as to how that relates to the development of an OR AE, to restate what others have said above, having very high geographic accuracy will matter little if the only activities that are available for a route are some basic "Explore" activities, or what can be cobbled together out of the MSTS AE--the latter a likely growing problem as some of the new OR routes may not even load in the MSTS AE.

[Lindsayts]

As one of my purposes was to be able to simulate real life country passenger services to within 1 minute in time I chose an distance accuracy of + or - 0.1 percent for my world model. I found this was readily acheiveable even with my own mathematics.

I ended up generating the terrain using existing open source geo information tools as it saved my having program the required tools myself. Although I did find existing documentation for most tools not that good for beginners. Checking the accuracy of the above generated terrain i found the accuracy exceeded any method a could find to test it.

I DO get the impression that most think this is some kind of impossible dream, something way to difficult to consider, I took me, a person who new NOTHING about the subject around 3 months of research to get a reasonable handle on the basics and a far bit of mucking around to be able to get the Data I required into the form I required it in (I use geotiffs 1 degree of lat and long square directly, my detail tiles being 1 sec of arc square) Note 1:

Note 1: Data in Australia is availible for the whole continent to a resolution of around 30 metres (!DEM), the web site interface though is a REAL puzzle for the person uninitiated in geo information systems and it took some time a quite a bit of expeimenting to work out how to get geo tiffs exactly 1 degree square out of the system. Even then I had to convert the data myself to geotiff format.

LIndsay

[SP 0-6-0]

Lindsay, I don't feel it's impossible to obtain this goal. It can be done if the powers to be decide it's important enough.

Robert

[Lindsayts]

SP 0-6-0, on 10 September 2017 - 08:03 PM, said:

Lindsay, I don't feel it's impossible to obtain this goal. It can be done if the powers to be decide it's important enough.

Robert

I apologise if I sound a bit to direct, this is a subject I am quite passionate about, OpenRails is SUPPOSED to be an accurate simulation, it surprises me that apparently so many think one does not need an accuarte world. What is the point of having accurate "physics" if the distance to the next station is 2 percent to short. To some people OR is clearly a game so accuracy is of little consequence, as far as I am concerned its a simulation and always will be and accuracy needs to be present in all parts not just the physics and what rolling stock and the scenery looks like.

I will say this though I because of problems with the world model in MSTS did a lot of work on solving this issue, its DISTINCTLY possible that people simply do NOT know enough about the issue to understand its importance.

Lindsay