Civ1 Scenario breaks on Turn 63/2740 BC

[Autumn Leaf]

Hi guys, I'm working on a scenario and I have encountered a very odd bug.

Everything works fine up until Turn 63 (2740 BC) but when I try to go to Turn 64 (2720 BC), the inter-turn period never ends. The mouse still moves, but otherwise the game is totally unresponsive.

It's happened twice now. The first save file was made as a freshly-generated Custom world. The second file was based on a freshly-generated Earth world. The only thing the save files have in common is that to save having the redraw the whole map, I exported the map from the first file as MAP.PIC and used it when generating the second map. I then altered the map to give it a sea border (rather than mountains) at columns 79/0, in case the game was having trouble with an all-land map. Nope, that's not the issue.

I got stubborn and played with the turn counter and year display. If I manually change both the Turn counter from 63 to 64 and the year from 2740 BC to 2720 BC, all is well and I can play on indefinitely. If I change them to (say) 60/2800 BC, all is well until it tries to go from 63/2740 to 64/2720, when again it freezes. If I change the turn but not the year, it freezes. If I change the year but not the turn, it freezes.

So, obviously there is something funky about Turn 63/2740 BC. These two files are my only games where this is happening. I used JCivEd 0.18a to draw the map, FWIW, but I've created other scenarios using JCivEd and this is the first time I've seen this glitch so I'm not inclined to blame the tool!

Still, I cannot for the life of me see what is special about Turn 63/2740 BC. Anyone have a clue to offer?

I want to use this world for a GOTM scenario (which is why no sample upload at this point) but if I have to include advice to "skip Turn 63/2740 BC and change the clock to 64/2720 BC" then I can't GOTM it.

 

[Mize]

This does ocasionally happen and I think it's been reported here. I also seem to remember it could be fixed by some slight hacking. It appears to be some sort of infinite loop in the AI's turn, or its inability to decide what exactly to do at a given point. Sometimes when I encounter this bug I am able to avert it or delay it by going back a few turns with a save and doing things differently, but it's not always up to your actions. Sometimes the AI just gets stuck.

 

[Autumn Leaf]

Thanks, Mize. Given that simply adjusting the clock averts it, I wonder what sort of indecision that would be. Restoring a save, no matter how many turns back, does nothing - it still freezes between 63/2740 BC and 64/2720 BC.

Maybe it's trying to spawn sea barbarians? But game 2, with an ocean at 79/0, should have averted that. I made sure to recalculate land values, so the game definitely knows where the ocean is.

Also, 64 is 8 x 8. I played on and now know it freezes again 8 turns later between 71/72. It seems to run in binary.

 

[Autumn Leaf]

SOLVED - more or less. It is indeed the lack of ocean. The game is doing something every 8 turns from 63/64 (generating Sea Barbarians is my working hypothesis) that fails if it can't find water.

I created a bunch of Earths today and played each through to 2700 BC (OCC, of course, just hitting enter most turns). The all-land maps failed every time, but maps that respected the poles, with oceans at rows 2 and 47 (counting from 0 at the top) and columns 78, 79, 0, 1, 2, worked. From this it looks like Civ does need ocean around all edges of the map.

I know from experience (my Riverworld scenario) that I can get away with less ocean than I used today. Now I just need to experiment a little more to find the minimum quantity and location of ocean that the game needs.