(sorry for my bad English, I hope you will understand some
)
What I did (methodology):
- backup of all original autosaves for not overwriting original sequence when replaying
- load the game year by year, always using original saved year: after end of turn and after going to year+1 for testing results, I have reloaded this year+1 from the original autosave.
- I went through the years 1681 (oldest backup), 1690, 1691, 1692, 1693, 1694, 1695, 1696, then straight to 1715, 1716, 1718, 1719, 1720 and 1721 (year where I reported the problem of self-attack)
Example of sequence used when testing:
a) load year 1690AD
b) skip with space bar all unit turns (no exception) when cycling through the units, propose default choices for cities, peaceful proposals if diplomacy screen is coming
c) when all units are idle, ENTER is proposed for end of turn. BEFORE doing so, I quicksave (F5), THEN from Essen I move toward the forest, hence declaring war on Bismarck, and see the result.
d) whatever is the result in step c), I reload the quicksave, and simply end the turn WITHOUT declaring war. All diplomacy screens and city screen are answered with default choice, or most peaceful choice.
e) at the year +1 (in this case 1691AD), I try again the attack for checking if the end of turn changed something
f) I start from point a) again: I load from the original backup the original autosave for the year+1 (1691AD in this case), and start the sequence from b to f again.
Summary of what I found:
- the problem regarding this region of the map (Essen and Bismarcks territory) starts in year 1693, with game crashing (Windows error report) when trying to attack the squares contiguous to Essen.
- a possible new bug (or unexplained behavior) can be found in year 1694 to 1695 with Qin
- the game does not crash in 1691 and 1692 as I can attack Bismarcks artillery from Essen, and 1720/1721 when attacking from Essen, as declaring war will repatriate Thomas Roe to his beloved Essen forest, next to the city, and the attack on myself allows the game NOT to crash!!
- if war is declared on Bismarck, and I move from Essen: IF there are no unit to attack and it is a simple move into enemy territory, THEN the game crashes; ELSE the attack proceeds by attacking the artillery or myself and does not crash.
- whatever is your choice in a given turn regarding city production or moving units within your borders, it has no impact on the Essen crash / self attack. The only exception to this is the order is sending a unit to Kyoto, then, when the unit/units reached Kyoto, declaring war on Bismarck send the unit to the forest close from Essen.
- Self-attack takes place on a corrupted background (crash when attacking from Essen), hence having perhaps unpredictable but constant effect. It is very probable that solving the origin of the crash would solve the self-attack problem. It seems unlikely that such event would occur again with a healthy map.
Specificities of tested years (all AD), from original autosaves
; common to all: I skip the turn for all units and choose default science/eco choice for cities (c.f. sequences a to f, above):
1691:
Bismarcks artillery in Essen forest. I declare war (via diplomacy screen or by direct moving: same final effect) on Bismarck. I move a tank in forest and attack. Everything goes fine, no crash and the game proceed
1692:
I refuse to give my World Map, but accept Stalin capitulation. I skip the turn for all units and choose default science/eco choice for cities.
Bismarks artillery still in forest. Like in my test for 1691, I move my tank (hence declare war). Like in 1691, after destroying the artillery, no problem, the game proceed
At this stage, I have not given yet the order to Thomas Roe to go to Kyoto (japan is a vassal of Bismarck), and there are no Inca unit to repatriate to my cities
1693:
Open border with Mehmed II.
No more artillery in the forest! I attack toward the forest or the mine just south of Essen, and the game crashes: Windows error report
1694:
Same as 1693, the game crashes if I try to attack.
POSSIBLE BUG?? :
Qin is my vassal, by his own choice (no capitulation, he proposed to be my vassal long ago when I started the conquest of simili-north America). He is at 11% of land, 6% of population and 103% of original land. I am by very very far the most powerful of all.
After hitting the [end of turn], he says that I am too weak and breaks free. How come as he is so far from the limits!!
This is not the first time I had this in the game, and Stalin, Ramses, Tokugawa, Cyrus, did it before, with Cyrus (as explained in my other post) even capitulating to me at the end of a turn when the AI plays - AND breaking free 5 seconds later within the same AI turn , while I was easily 5 times bigger than Cyrus!! No save games on this.
1695:
Qin broke free (see 1694). Declaring war on Bismarck and attacking forest or mine south of Essen still crashes the game.
Years 1696 to 1720 skipped as it is the same as 1695.
1720:
My Great Merchant Thomas Roe is reaching Kyoto (on the East coast of the main continent). As Japan is vassal of Bismarck, when I declare war, Thomas will then be automatically transferred to my closest city which is Essen. Problem, he ends in the forest next to the city. The game thinks
- Essen is in the forest OR
- the forest and Essen share the same coordinates OR
- a virtual city, called Apache seems to have been created by the system (according to the logged events). Apache was a barbarian city I razed (1 population) long long ago (at least 100 turns) in the simili-north-american continent I conquer. If checking F1 before attacking: I have no city at this name, however, Thomas is repatriated to this virtual city and I destroy it when attacking myself.
Files uploaded:
- a single zipped file: original-BUGreport-I attack myself-Qin-autosaves years 1691 1692 1693 1694 1720.zip
This file contains:
{error report generated by Windows when crashing, years 1693 and above}:
12f7_appcompat.txt
{quiksave, just before declaring war on Bismarck}:
... replayed-BUGreport-I attack myself-BEFORE declaring war on Bismarck-AD1721-QuickSave.CivWarlordsSave
{original autosaves with relevant years}:
... original-BUGreport-I attack myself-AutoSave_AD-1691.CivWarlordsSave
{no crash}
... original-BUGreport-I attack myself-AutoSave_AD-1692.CivWarlordsSave
{no crash}
... original-BUGreport-I attack myself-AutoSave_AD-1693.CivWarlordsSave
{crash}
... original-BUGreport-I attack myself-AutoSave_AD-1694.CivWarlordsSave
{crash and Qin}
... original-BUGreport-I attack myself-AutoSave_AD-1720.CivWarlordsSave
{no crash self attack}
link: http://forums.civfanatics.com/uploa...-autosaves_years_1691_1692_1693_1694_1720.zip
My conclusions:
- The system behaves as if there are 2 different Incas: me, and a virtual one : which can explain the reaction of Qin, Stalin, Ramses breaking free while I am so so far superior to them in land population and other. My virtual ghost is probably weaker than me (perhaps with only this Apache city, hence I am very weak at their eyes!!)
Of course, if a Inca unit goes into my virtual Inca city, it is like a rebel unit (barbarian?), it is logic to have my faithful units attacking those simili rebels!
- Due to the large size of the map, the number of civilizations, the long loading time of the game, all possible background processes (NAV 2006, Symantec Internet Security 2006, Lavasoft Pro 1.06r, Windows Defender, swap file, changes in the processor speed as it is played on a laptop switching from battery to plug with variation of processor frequencies
) it is very possible that the game get corrupted, perhaps way before year 1690, and the gap became apparent only later in the game with unpredictable consequences.
- Do not waste time on this if:
I am the only one reporting the problem.
a Warlords saved game is built using data already existing from previous turns (for accelerating backups: incremental), and not by reconstituting all logical links when backing up (cleaning process rebuilding the unit/map databases by comparing logical data with expected data): if it is the case, a long and heavy game can collect too many external problems along the way!
it is too boring.
- If independent from the attack myself problem, any reason for Qin saying that I am too weak? (1694/1695)
