Yeah, there's an option to decompress the BIQ in the editor, but no way to do it in game.
As for getting around this, if you download the latest version of my editor, in the /bin folder, there's a Java .jar file that the editor uses to decompress save files. SAV files are compressed with the same algorithm, so that program will also work on SAV files. Currently, it has to be run from the command-line. Once you navigate to its folder, run:
java -jar BIQDecompressor.jar [input.SAV] [output.SAV]
where input.SAV is the file you want to decompress, and output.SAV is the name you want it to have when decompressed (I'm not sure off the top of my head if it will work if the names are the same, but I'd recommend having them different just in case something goes wrong. I can't vouch for the accuracy of this program as I didn't write it - I just put a command-line front-end on it - but I haven't seen it fail yet).
Steph, I think WildWeazel meant setting the retirement to be the next turn only if the era was changing (or something else allowed by scripting). Although unless that was changed in memory, the player wouldn't know about it until they loaded the autosave, by which point they wouldn't need to load it any more. So that may be a dead end.
You're right that there's certainly risks in modifying things in memory. I've accidentally crashed my game a couple times while playing around with it. Allowing scripting with reloading autosaves probably could progress quite a bit more quickly due to the lower risks and the familiarity with the format. I guess I'm favoring the memory for now in part because ideally (if it isn't buggy), it's preferable in a lot of cases, and if we're going to make progress with auto-save updates, it might as well be with your editor as a base since it's the best program at it currently.
I haven't spent a whole lot of time on this yet, but I've got the basic structure of the memory-updating program in place. There's definitely some polish to add to it, though, and there's no scripting yet. So what I've figured out is that this approach probably is feasible, it probably won't be very quick, and if there are bugs, they can be rather unpleasant.
