Hi, it seems like the dashboard may be failing to update after some point in your game

The problem is usually that I have no way to reliably test app with the end game, so something that I didn't expect happens and breaks the dashboard program.

Do you see any error messages in the dashboard application's terminal window?

And if you want, you can send me your "output/db/$gamename.sqlite" database to blubblubblobb[at]gmail[dot]com and one or two of the save files that were not picked up to .

I might have puzzled out why the budgets stopped getting updated in my singleplayer game: the save file somehow gained a second player who controlled the same country:


And based on the code in , skipping adding BudgetItems would make sense, since "this" country would have an ID of 0 and a different country with an ID of 0 would show up in .
I'm quite certain I was playing singleplayer this entire time, so, unfortunately, I have no clue how that second player showed up. In a new game, I've tried being in observer mode when an autosave is made and disabling my wifi connection before starting Stellaris, but neither caused a second player to show up.

In any case, removing that second player results in the empire-specific tabs showing the data as expected. It's a shame data from the intervening years is lost (or I guess it might be technically retrievable from the stored GameStates, but doing so might be a bit tricky for me), but it's not that much of a loss.

As a side note, I saw some warnings when parsing the save file:
- "WARNING - test 2357.07.01 Cannot establish ownership for system None and country 0"
- "WARNING - test 2357.07.01 Could not find war matching '\x11HFree Advancement Cooperative\x11! - Fjasivlin War'"

The modified save file with only one player can be found here: https://www.dropbox.com/sh/dayxq0x7wk0q7uw/AADDajmejR7uKswPWK06VVPTa?dl=0

As for the second case (the observer game), it might be a fluke? I accidentally deleted the database with a poorly thought-out script, and upon re-parsing the autosaves the empire-specific data showed up just fine. I'm not really sure what happened there, to be honest.

So in the end, the first case may or may not be a bug depending on whether two players with the same country is a reasonable thing to expect from a save file, and the second one appears to be an issue I can't reproduce.

The main problem is that the data model used in the dashboard application assumes that there is exactly one country from whose perspective events and such should be recorded. So if that country switches partway through a game, it is currently not obvious to handle that correctly...

Unfortunately, switching the player country and multiplayer are much harder to support at this point. Perhaps I can make a simplified version of the dashboard more geared towards observer mode that only records economic stats, then the perspective should be less of a concern.

Would that help?

An observer-only mode wouldn't help much in the first case, since that's supposed to be a singleplayer game and I'd prefer to have the empire-specific data if at all possible.

I think the issue less to do with perspective and more to do with the save file containing two players with the *same* country, which, on its face, is rather nonsensical. Handling that situation, whether it be by erroring out, not adding players with duplicate countries to other_players, or some other strategy, might be a viable alternative.