Feb 7th 2021, 21:41:36
Hi All,
I'm very sorry for the upset & disappointment many of you will have had with the end of reset yesterday/today; I very much appreciate the amount of work that can go into countries over the reset.
Just an FYI as to what all happened, and where things are;
- I initiated a quick reboot to apply some changes that had built up, partly to do with trying to get a new dev environment set up
- We'd had a hard drive fail a few months back, but I'd forgotten to install grub onto it, and the server choked on boot
- Creation of USB recovery disks ensued, and the reset clocked by
- We agreed rolling back the end time would be best, so I did that
- I checked the code, and found that I'd written a function that should return everybody's goods in the case of a market table discrepancy years ago, and thought, perfect, that'll sort itself out
- After hearing that that was not the case, I examined the code more closely, and it appears that we erroneously added a block that set everything to zero right before re-calculating based on the market table, negating the bit that would return everybody's goods
- I have since released and pushed in a fix for that
Unfortunately, there's no way for us to track what countries had on market at the end of the reset - those tables get wiped at end reset; and the code that was *supposed* to handle discrepancies didn't work properly, and in fact erased the knowledge of what was there.
The only consolation I can really give is that we're hoping to actually get some development of some exciting new features done here in the next year; Slagpit has shown interest in development again, which has drawn pang & I into actually looking at things again, which is kinda exciting.
- qz
I'm very sorry for the upset & disappointment many of you will have had with the end of reset yesterday/today; I very much appreciate the amount of work that can go into countries over the reset.
Just an FYI as to what all happened, and where things are;
- I initiated a quick reboot to apply some changes that had built up, partly to do with trying to get a new dev environment set up
- We'd had a hard drive fail a few months back, but I'd forgotten to install grub onto it, and the server choked on boot
- Creation of USB recovery disks ensued, and the reset clocked by
- We agreed rolling back the end time would be best, so I did that
- I checked the code, and found that I'd written a function that should return everybody's goods in the case of a market table discrepancy years ago, and thought, perfect, that'll sort itself out
- After hearing that that was not the case, I examined the code more closely, and it appears that we erroneously added a block that set everything to zero right before re-calculating based on the market table, negating the bit that would return everybody's goods
- I have since released and pushed in a fix for that
Unfortunately, there's no way for us to track what countries had on market at the end of the reset - those tables get wiped at end reset; and the code that was *supposed* to handle discrepancies didn't work properly, and in fact erased the knowledge of what was there.
The only consolation I can really give is that we're hoping to actually get some development of some exciting new features done here in the next year; Slagpit has shown interest in development again, which has drawn pang & I into actually looking at things again, which is kinda exciting.
- qz
Finally did the signature thing.
EE Patron