RFC on the GEM map?

[Mouthwash]

Has this ever been tried? I might give one a shot if it hasn't (preferably using the StrategicRivers mod).

 

[merijn_v1]

It has been tried several times. But all were projects abandonned. It's just too much work. It will take a lot of time to create all settler-, war- and other maps. And even more time to balance everything.

Don't be afraid you won't get any support though. There have been several requests for something like this. So enough people are interrested.

 

[binTravkin]

Let's continue this discussion.

First off - I don't like GEM.
Three particular reasons I noted by playing just a couple of games with GEM:
- ability to found a city that is BOTH on Nile and on Red Sea
- tundra in all kinds of places, where the climate is very far from tundra
- blocking mountains in places where there should be none (Aghanistan was passed through by large armies multiple times in history, same for many other places, Mexico, etc)

However, the author claims the resources are very accurate and the detail is also good.


But, I think, there are other ways to expand the playable area of RFC.

I will not elaborate right now, because it would take too long to explain properly, but what I would like to know is - can you list all the map-related work that needs to be done if migrating to a different map (assuming the map itself is ready to use)?

What I know ATM:
- starting locations including scripted spawns for barbarian and independent cities
- settler maps
- stability maps
- war maps
- balancing (some of the mechanics need different modifiers due to larger map)

Additionally, I read that it is possible to change the DLL in such a way as to allow extra civs (up to 50?).
Could this be done for RFC, too?
It would allow independents be more granular and possibly also some extra historical civs.

 

[merijn_v1]

Additionally, I read that it is possible to change the DLL in such a way as to allow extra civs (up to 50?).
Could this be done for RFC, too?
It would allow independents be more granular and possibly also some extra historical civs.

That's already done. RFC has many more than 18 civs. (IIRC 27 playable, 2 independents, celts, byzantines, natives, native america, barbs)
RFC Classical world has even more than 40 civs!

 

[binTravkin]

Ok, why not add more?
RFC already seems to need more spots for independents as they often seem to make nonsense cultural connections and shift troops between cities quite a distance away.

 

[srpt]

moving RFC Classical World to a larger map took 1 long weekend of steady work to get it basically running and about another week of tweaking to get it enjoyably playable, but that included drawing the new map from scratch. the only task I can think of that you missed on your list is the city name map.

 

[Crossphazer]

Although, if it isn't province-based, like traditional RFC, it will take you somewhat longer to manually draw the stability maps for each civ. But getting those maps done shouldn't represent a monumentally bigger task.

 

[srpt]

yeah thats a really good point. all I had to do was make 1 new province map and everything else just follows that. also RFCCW, like SoI, only has 1 city name map. so doing RFC as it is would be quite a bit more work, especially of the tedious, data-entry type.

 

[binTravkin]

Can somebody remind me where the stability and settler maps are stored?
I have forgotten where to look for them.

 

[merijn_v1]

They are in the DLL. (CvRhyes.cpp)

 

[srpt]

if I was going to do this (if I didn't have a job I would do it) I wouldn't use GEM. I'd draw a new map. when I drew the large RFCCW map I started with the locations of the critical city sites, (taken by multiplying their coordinates from the old map by 1.5) and drew the map around them, constantly thinking of city sites and civ cores etc, having already had the experience of playing the same civs in the same places and sequence on a smaller map. obviously the same circumstances would apply with a larger world map for RFC.

 

[binTravkin]

This is exactly what I'm doing.

I have identified these steps:
1. upscale x4 (x2 on each dimension)
2. straighten rivers
3. average out terrain height
4. average out humidity
5. average out vegetation / ice
6. automatically determine location of bonuses
7. add/change bonuses by hand
8. sort out any bugs and inaccuracies by hand, possibly add some features from other maps, like smaller rivers
9. create proxy functions for the various maps (based on upscaling/downscaling coordinates)
10. test & edit the places in code, which need to have more precise coordinates
11. change some basic values and modifiers to account for the larger map (stability calculation, unit movement speed, some others)
12. update the resources spread functions (I'd like sugar to spread, also I'd like the spread to be subject to conditions, not date-based)
13. optimize to create less load - the map size 248x135 or possibly slightly less height (cut a few lines of ocean off).

Step 1 is done.
Steps 1-6 are to be done by a script.
I have imported the map into a MySQL database and each automatic step creates new table.
I am now writing an export script to be able to output the map for testing.

 

[binTravkin]

Ok, now I need help.
I've exported the map.
It's a basic upscale - each plot in RFC original map is made into 4 plots in this map.
If there is a river running "around" the plot, it runs "around" all 4 as if it was one.

I used all of the part that contained "BeginPlot/Endplot" chunks together with a custom header, to create a save file.

Now, it properly recognizes the header and allows me to start loading it, but once loading starts, it will crash.

Does somebody know - is there any limitation as to the number of plots?
Maybe my map header is missing something?

The map is attached to this post.

 

[binTravkin]

Ok, I figured one problem out - the rivers were not quite running properly as I had imagined that Y is counted from top to bottom, not vice versa.
Now that is fixed, but the map still crashes even in RFC (it contains swamps and cotton - both from RFC).
Are there any good ways to debug it?

Attached screenshot how it looks in MapView 2.0.9.

 

[srpt]

I would copy the top part of the map file, everything down to "BeginMap" from the RFC mod you're using and also change the catapult, unit flipping and mercenary plots in the python and the DLL (I think only the catapult plot is in the dll). that might get it running.

 

[binTravkin]

Right, I thought of that too.
I don't know what the plots you named are though.
I figured I need to at least correct the starting locations for it to be loadable.
I only want to load the game without advancing a turn.
It should be enough to show that it can use the map.

 

[Mouthwash]

Lol. I had no idea people were continuing this conversation. Um, why are we using this weird, scaled-up map instead of the best Earth map ever made?

 

[Askiya]

Lol. I had no idea people were continuing this conversation. Um, why are we using this weird, scaled-up map instead of the best Earth map ever made?

Do you mean the Peters projection map? While it's more to scale, it also doesn't allow for historical advancement the way the current Mercator one does.

 

[binTravkin]

The scaled-up map as you see it is not ready.
I use it because I think it will be easier to make due to being partially made by script and the in-game coords passed through a proxy function.
Also, I outlined some flaws of "best Earth map", which I would have to fix for it to work properly with RFC, thus creating significant extra work even past point-by-point adaptation of all the settler and other maps.
I will move this to a new thread once map reaches advanced enough stage to accept feedback.

 

[Crossphazer]

He's referring to, well, the GEM map. It looks like a custom projection of the Earth, with the pacific shortened and an enlarged Europe. Europe is still bigger in binTravkin's projection, though.