You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
All industries at the start of a game should have the same risk of closing (from inactivity/low production) whether they were built automatically by the map generator, or manually in the scenario editor.
Actual result
If the scenario starting date is changed after building an industry, the industry might have a risk of closing very soon after scenario start, instead of the usual grace period of 5 years from its construction.
Steps to reproduce
Start scenario editor
Place some towns and some industries
Save the scenario as name A
Change the starting year of the scenario to 10 years later (e.g. from 1950 to 1960)
Save the scenario again as name B
Start a new game from scenario A, let it run for a long time, and take note of how many of the placed industries close
Start a new game from scenario B, and similarly let it run and take note of industry closures
The scenario with changed starting year has more industries closing, because their build year (and thus recorded "last production date") is 1950, but the scenario starts in 1960 so they all appear to have been idle for 10 years.
Attached sample files made according to these instructions
In scenario A (start 1950), the first industry closure is recorded in February 1956, and 10 years after game start 3 industries have closed.
In scenario B (start 1960), the first industry closure is recorded in February 1960, and years after game start 7 industries have closed.
This isn't a great demonstration, simply because it's too small sample size, but it's at least convinced myself! Industry closure test A and B.zip
Thoughts
I see three possible places to fix this:
During starting a new game from scenario.
During saving a scenario in the editor
Each time the starting year is changed in the editor
All three would need a thing to reset the "last production year" for all industries. All three have the issue that there isn't really a precedent for "fixing up" things like this. In fact, for the first option, the game can't tell the difference between starting a new game from a scenario and loading a saved game, so that would require introducing a new flag somewhere.
The text was updated successfully, but these errors were encountered:
Version of OpenTTD
12.2
Expected result
All industries at the start of a game should have the same risk of closing (from inactivity/low production) whether they were built automatically by the map generator, or manually in the scenario editor.
Actual result
If the scenario starting date is changed after building an industry, the industry might have a risk of closing very soon after scenario start, instead of the usual grace period of 5 years from its construction.
Steps to reproduce
The scenario with changed starting year has more industries closing, because their build year (and thus recorded "last production date") is 1950, but the scenario starts in 1960 so they all appear to have been idle for 10 years.
Attached sample files made according to these instructions
In scenario A (start 1950), the first industry closure is recorded in February 1956, and 10 years after game start 3 industries have closed.
In scenario B (start 1960), the first industry closure is recorded in February 1960, and years after game start 7 industries have closed.
This isn't a great demonstration, simply because it's too small sample size, but it's at least convinced myself!
Industry closure test A and B.zip
Thoughts
I see three possible places to fix this:
All three would need a thing to reset the "last production year" for all industries. All three have the issue that there isn't really a precedent for "fixing up" things like this. In fact, for the first option, the game can't tell the difference between starting a new game from a scenario and loading a saved game, so that would require introducing a new flag somewhere.
The text was updated successfully, but these errors were encountered: