Quote: "if you're offering that our work should be automatically saved before attempt to export (which, obviously, might fail), i'd definitely agree."
no, I was not suggesting that at all. That can have other nasty side effects....
The only reason I mention loss of work is because the potential loss of data should (rightly) elevate the severity of the issue.
The reason automatic saves can be problematic is because in any system where the user is not expecting changes to be saved, doing so presents unexpected behavior. The user may have just deleted some big chunk of code, not intending to save and test it by exporting! For functionality where a catastrophic crash is likely and can't be fixed for some reason, then a better approach is to check for dirty data at the
start of the export and ask the user then to save if they care about the changes before getting too far.
In this case even though there is a system configuration issue, I regard it as a bug. I don't see it documented on GitHub (search for jarsigner produces no results) I would be happy to write it up, I grabbed a screen shot of the error message before fixing the path. (It seems so 1980s to be mucking with the path environment variable!
)
