I have spent a lot of time generating JTCS wargame map files by using QGIS. In terms of terrain, the files I generated are not very good; many rural areas, towns, and cities are not well-differentiated. The same applies to rivers. As for elevation, the elevation files also took a lot of effort to generate successfully, encountering various problems along the way. The road layer is particularly troublesome because you have to consider grid connectivity. Even knowing the 2^6 permutations and their corresponding JTCS map codes, it's still very difficult. Just the first step, exporting the data, is hard, and the second step, considering connectivity, is even harder; replacing codes is a secondary concern. I also haven't had time to deal with the grid edge contours, which is really a hassle. Unfortunately, I don't know how to use Wolfram Mathematica to recognize maps and then export data. If only someone could tell me how to generate these maps, that would be great. I've been using Gemini 2.5 Pro for a long time and have struggled to generate the code. Claude 3.7 Sonnet's generated code is somewhat better, but I've still tried for a long time with the road issue without good results, especially considering the hexagonal grid problem—the rows, columns, and adjacent cells of a honeycomb structure are very particular. It often happens that only a half-segment of a road is generated in one cell, while the adjacent cell, where it should have been generated, is left empty. Furthermore, the roads frequently appear as if they have been compressed.
Under the article,I send my QGIS with Wolfram attempt process.You can open "Pokrovsk.map" by cold war map edit.exe,if you want open it on other map edit.exe,you need to change the code characters,you can use wolfram to replace code characters.
I might send some pictures to illustrate the situation. For the pokrovsky.map
, I've been creating path information manually, which I found to be very frustrating and time-consuming. I'm looking for a more effective method to solve this. Other pictures describe my attempt to create the map using code, but it hasn't been very useful so far. If anyone could share their methods for creating path, terrain, or landform layers, I would be very thankful.
(Yesterday, 02:02 AM)Jason Petho Wrote: This would be a fascinating project if you can get it work and would make mapping much easier.
Right now, I do all the mapping inside the map editors, but if the base could be done in QGIS, that would be A HUGE time saver!
Thanks for your reply. I was thinking about those enormous maps – like Normandy, Herbstnebel, NL40Oos, Borisov, Odessa, Uranus, Die Brucken thebridges, Koltso, bayarea44, Overlord, and really any map over 400kb. If I had to structure them all by hand, I'd probably go crazy and get completely worn out.
(Yesterday, 02:02 AM)Jason Petho Wrote: This would be a fascinating project if you can get it work and would make mapping much easier.
Right now, I do all the mapping inside the map editors, but if the base could be done in QGIS, that would be A HUGE time saver!
QGIS is also quite effective for generating map terrain – my code demonstrates how the terrain layer is created.
When it comes to landforms, achieving realistic results requires balancing the overall structure and the proportions of different landform types. This often involves significant effort, potentially using AI, because otherwise, the output can be inaccurate. It's a lengthy process.
Getting the JTCS path layer right is genuinely challenging. Similarly difficult is accurately implementing details like various rock formations, waterways, bushes, and ferry crossings, particularly along grid edges. I suspect these elements present a similar level of complexity to the path layer, and perhaps even share similar code structures within the JTCS map path system.
(Yesterday, 02:02 AM)Jason Petho Wrote: This would be a fascinating project if you can get it work and would make mapping much easier.
Right now, I do all the mapping inside the map editors, but if the base could be done in QGIS, that would be A HUGE time saver!
Thanks for your reply. I was thinking about those enormous maps – like Normandy, Herbstnebel, NL40Oos, Borisov, Odessa, Uranus, Die Brucken thebridges, Koltso, bayarea44, Overlord, and really any map over 400kb. If I had to structure them all by hand, I'd probably go crazy and get completely worn out.
The monster maps I had created for Vietnam had taken months to build, hex by hex, each map. Being able to even build the skeleton of the map with GIS data would be a massive time saver, even if having to go in after and do the details.
(Yesterday, 02:02 AM)Jason Petho Wrote: This would be a fascinating project if you can get it work and would make mapping much easier.
Right now, I do all the mapping inside the map editors, but if the base could be done in QGIS, that would be A HUGE time saver!
QGIS is also quite effective for generating map terrain – my code demonstrates how the terrain layer is created.
When it comes to landforms, achieving realistic results requires balancing the overall structure and the proportions of different landform types. This often involves significant effort, potentially using AI, because otherwise, the output can be inaccurate. It's a lengthy process.
Getting the JTCS path layer right is genuinely challenging. Similarly difficult is accurately implementing details like various rock formations, waterways, bushes, and ferry crossings, particularly along grid edges. I suspect these elements present a similar level of complexity to the path layer, and perhaps even share similar code structures within the JTCS map path system.
The bright side with QGIS is one can create their own data, or modify existing data to make it historical. As noted above, even being able to add road, rail, hydrology, forest and town/village areas would be a huge time saver and then fill in the map later with the proper details.
That's a massive challenge and impressive indeed. I'd probably recommend looking at the Middle East or Vietnam map editors as the new WW2 games will be based on those - and they have a large number of new tile types.
(Yesterday, 02:02 AM)Jason Petho Wrote: This would be a fascinating project if you can get it work and would make mapping much easier.
Right now, I do all the mapping inside the map editors, but if the base could be done in QGIS, that would be A HUGE time saver!
QGIS is also quite effective for generating map terrain – my code demonstrates how the terrain layer is created.
When it comes to landforms, achieving realistic results requires balancing the overall structure and the proportions of different landform types. This often involves significant effort, potentially using AI, because otherwise, the output can be inaccurate. It's a lengthy process.
Getting the JTCS path layer right is genuinely challenging. Similarly difficult is accurately implementing details like various rock formations, waterways, bushes, and ferry crossings, particularly along grid edges. I suspect these elements present a similar level of complexity to the path layer, and perhaps even share similar code structures within the JTCS map path system.
The bright side with QGIS is one can create their own data, or modify existing data to make it historical. As noted above, even being able to add road, rail, hydrology, forest and town/village areas would be a huge time saver and then fill in the map later with the proper details.
That's a massive challenge and impressive indeed. I'd probably recommend looking at the Middle East or Vietnam map editors as the new WW2 games will be based on those - and they have a large number of new tile types.
I agree with you. The ability of QGIS to automate base layers like roads, rail, hydrology, forests, and towns would indeed be a significant time-saver for large maps. Manual creation is very time-consuming, as you've experienced with your Vietnam maps.
You're correct,the Middle East and Vietnam map editors could provide new insights with their expanded tile types. It may a good recommendation, and I'm interested in exploring them. However, acquiring those newer games has been challenging due to their cost and inconvenience of purchasing them in China through sites like Matrix Games or Wargame.com.
Even with current tools, addressing the JTCS path layer remains the primary challenge, particularly concerning the hexagonal grid's unique connectivity. Issues like partially generated roads or compressed sections often occur, despite considering the 2^6 permutations. This underscores the complexity of this automation project, even for generating the "skeleton" you mentioned. Success in this area would be a significant development.