Current implementation does not always differentiate answers given by user from answers that were automatically created by the system. This has several side-effects that affect the flow of games:
User can see that all the answers have already been answered, if all of those questions have been opened and empty answers created. This might require refresh to take effect, at least in some cases.
Answering just one question will mark the whole game as complete, while the user has only given less answers than it is required. This happens due to creating of empty answers on initial question dialog open.
Different parts of code could show completion percentage as a lot higher value that it really should be. Empty answers are to blame for that.
Code has to be changed to always include the check for answers really being given by users and ignore the one that were automatically crated. The best approach seems to be to use the is_answered attribute as this seems to be the attribute to determine just that.
Current implementation does not always differentiate answers given by user from answers that were automatically created by the system. This has several side-effects that affect the flow of games:
Code has to be changed to always include the check for answers really being given by users and ignore the one that were automatically crated. The best approach seems to be to use the
is_answered
attribute as this seems to be the attribute to determine just that.