[Genma Saotome]

Bill, I think it's fair to recommend qGIS over TransDEM on several accounts: It is free, it is still underdevelopment, and it can do more. Beyond that I don't know enough to really comment. What did impress, tho I really don't know / understand the details, is the qQIS projection to TM or UTM offers hundreds of target locations, which I'm guessing a pre-configured projections for specific locations. Seems to me such targeted outputs would lend themselves to minimizing the 15d constraint you mentioned. Is that guess correct?

FWIW I'm using qGIS Valmeria 2.2.

[Mike B]

Genma Saotome, on 11 January 2018 - 07:37 PM, said:

Bill, I think it's fair to recommend qGIS over TransDEM on several accounts: It is free, it is still underdevelopment, and it can do more. Beyond that I don't know enough to really comment. What did impress, tho I really don't know / understand the details, is the qQIS projection to TM or UTM offers hundreds of target locations, which I'm guessing a pre-configured projections for specific locations. Seems to me such targeted outputs would lend themselves to minimizing the 15d constraint you mentioned. Is that guess correct?

FWIW I'm using qGIS Valmeria 2.2.

When did that come out? 2.18 Las Palmas is the latest I've been able to get from their web site (https://www.qgis.org). They're warning that 3.0 is coming soon...

[Genma Saotome]

Mike B, on 11 January 2018 - 09:17 PM, said:

When did that come out? 2.18 Las Palmas is the latest I've been able to get from their web site (https://www.qgis.org). They're warning that 3.0 is coming soon...

Actually, I only now understand their numbering scheme. Valmeria is two point two whereas Las Palmas is two point eighteen. Dunno why Valmeris wasn't 2.02 (about really does say 2.2)....

I upgraded about 30 minutes ago.

[BillC]

Genma Saotome, on 11 January 2018 - 07:37 PM, said:

Bill, I think it's fair to recommend qGIS over TransDEM on several accounts: It is free, it is still underdevelopment, and it can do more. Beyond that I don't know enough to really comment. What did impress, tho I really don't know / understand the details, is the qQIS projection to TM or UTM offers hundreds of target locations, which I'm guessing a pre-configured projections for specific locations. Seems to me such targeted outputs would lend themselves to minimizing the 15d constraint you mentioned. Is that guess correct?

FWIW I'm using qGIS Valmeria 2.2.

Dave,

I wasn’t advocating TransDem over QGIS, just replying to Simon E post. TransDem is useful only with trainZ. However Ziegler’s method of converting different projections to UTM, projecting to the local tiles is useful.

I known about QGIS for about 6 years, when living in Oakhurst, stumbled upon a book in the Fresno library Desktop GIS by Gary E. Sherman. Sherman is the creator of QGIS at the time called Quantum GIS. QGIS is the open source clone of ESRI ArcGIS. On the web there is for ArcGIS resources ArcGIS 10.1 , which is actually a tutorial on GIS. Drilling down under Guide Books Map projections Projected under Limitations. For both TM and UTM is where the 10 – 12 deg constraint.

[Simon E]

BillC, on 12 January 2018 - 08:40 AM, said:

Dave,

I wasn’t advocating TransDem over QGIS, just replying to Simon E post. TransDem is useful only with trainZ. However Ziegler’s method of converting different projections to UTM, projecting to the local tiles is useful.

I known about QGIS for about 6 years, when living in Oakhurst, stumbled upon a book in the Fresno library Desktop GIS by Gary E. Sherman. Sherman is the creator of QGIS at the time called Quantum GIS. QGIS is the open source clone of ESRI ArcGIS. On the web there is for ArcGIS resources ArcGIS 10.1 , which is actually a tutorial on GIS. Drilling down under Guide Books Map projections Projected under Limitations. For both TM and UTM is where the 10 – 12 deg constraint.

I know TransDem is only for Trainz.(Which itself uses some strange Projection of its own with a 720 x 720 m tile grid... Weird) I was just getting at the fact it seems to be exactly like what we want. Its qGIS that I use a lot and it does the same thing. (And as we have already established, we don't need all of its functions) We just need the Projections, Shapefiles, DEM And map tiles stuff., Im sure we would be able to put something together. Wouldn't it be very useful to use some GIS program, Pass the whole lot to Goku's Editor, And have him layout all your tracks (In the right gauge) , and roads, Automatically? (I think Goku, is heading there). We just need a SANE map projection in Open Rails (which is suitable for its purpose, and Known to us) , and a way to show It correctly, because one reason is the MSTS mess, is unusable in lots of places. (Think of Japan..)

[leir]

Hello,

just some tips for the original thema:
- There is mathematically no way how to make seamless planar projection, which will not distort majority of the projected globe on distances and angles greatly. There is just two possible solutions:
1. Convert all terrain plane to (at least) sphere. This is totally uncompatible with current MSTS/ORTS tile database concept and means a brand new simulator.
2. Use local projections. This is a standard approach for more than 2000 years. Every route should have defined own projection, which fits the globe on given route localization and orientation. This means that we should use standard projections defined for maps of given states or countries, which are optimized for minimal distortion. There exist very good maintained GPL/LGPL library for managing this task: Proj4. This library is a standard part of a tool/library called GDAL/OGR, which also have C# port, see eg. https://trac.osgeo.o...GdalOgrInCsharp . QGIS is using the same library, but in C++. The definition of route just needs to hold given projection definition, e.g. as a proj4 definition string. In the body of the ORTS should we then just create the Projection object from the string when initializing/loading the route and then call the transform() function when necessary. The data and whole object model then still should stay in planar coordinates.

[darwins]

I am coming to this late, but this new projection will be so much better for routes in UK.
Unlike orthographic projection centred on a particular lat/lon in Demex, it means that all routes using this should match up with one another.
Still I need to review how to get started. Using TSRE5 I set up a new route - just one tile to begin with - in Lancashire - 53.417142 N, -2.766766 W
Then having read the wonderful news here, I added TsreGeoProjection ( 48.8485 2.29143 -12340224 30290944 ) to the trk file, set the terrain with dem data, and applied a map...
It looks a lot more rural than my chosen location is today - hold on a minute - I find myself in Nottinghamshire rather than Lancashire - 53.1019 N, -0.934637 W ... hmm - I think I am going to need a new start tile - but how do I choose the right one?

[darwins]

Not too difficult to solve the above problem after all.
First "Jump" to the appropriate latitude and longitude.
Second create a new tile there.
Then set that as the start tile!
(Having deleted the original world tile and its associations manually, TSRE seems quite happy that I have a route ready to go.)
Being able to use Open Street Map to position things is great - but the trick now will be to see if I can overlay the saved OSM map with a historical map and use that.

[Kapitaen13]

The problem is, that we do not know the two values for the center point to the Start Tile.

Quote

TSRE will use new TSRE projection.
This line describes center point in the world. From this LatLon average distance in meters per degree is calculated, and TSRE use it for the whole route.
Last two numbers is center point in the internal coordinates, so it is possible to define center point at custom world tile location.
Any ideas of better definition in .trk file are welcome.

I have tried for several days but did not find the correct solution, although I had approached the Google maps Map except for one Tile.
But another approximation was not possible, because the first value changed again... ?

[darwins]

Now finding the choosing the Lat and Long from the centre of the route can be quite important. Even in a relatively small country like UK choosing values from the other end of the country can make quite a difference to the East West scale of the tiles.
Not feeling too sure now if we should use a standard centre for each country or state (which would allow routes to join up) - or use an individual centre for each route - giving greater accuracy to individual routes.
Also thinking now that if my route is centred on 57.4812 N and 4.2156 W then I can enter
TsreGeoProjection ( 57.4812 -4.2156 x y )
but should that be

TsreGeoProjection ( 57.4812 -4.2156 -12340224 30290944 )
Or are those other values going to change? - what Kapitaen is trying to work out? - and if so how much difference will they make?