Unfortunately, I did the flush after I reviewed the extra terms. There was no effect. The the extra words were not counted in "studied today". Which is correct as I already reviewed them.
I will try to do flush again, when it happens again. The expected sequence will be:
Do the reviews and wait for extra review term(s) to appear in the schedule.
Do the flush.
It is expected that the number terms for review might decrease.
Is this correct?
Another question: What does the flush do exactly? Does it force some database updates to achieve data consistency?
The flush basically resets the summary data for the schedule. So if, for example, you have 10 terms to review, but the summary data mistakenly is saying 12 terms, then it should fix it back to 10.
I managed to reproduce the extra term in schedule.
In my second quiz of "Words for Japanese Basics" there were 92 terms to review, after the 40 word quiz, there are 53 terms to review (instead of 52).
I did the missed terms at the end of the quiz.
"Studied today" was correctly increased by 40.
The flush did not help. There are still 53 terms to review.
This is first time I seen it in past couple of days. Yesterday, I did not do any quizzes, so it might be somehow related to that.
EDIT: After the next quiz, the terms for review were decreased by 40, but "Studied today" were increased only by 39. (Possibly got same word twice today.) Flush had no effect.
Thanks - I am really quite confused as to what could be happening.
Try this: first, so you don't go crazy, I'd recommend taking your quiz size down to 10 or so. Then do it, one quiz at a time. Once you notice the discrepancy, tap the green number under the schedule "xx to review" <-- that number will link you to the terms that still need to be studied for the day. I'd like to see if you can find any of the 10 terms that were in the last quiz. Perhaps seeing that term will help clue in on something that was different about that term when you quizzed it.
Was this post really 8 days ago? I'm afraid I did not see it until now - not sure if I slipped into a time bubble or something.
I doubt that you recall, but when you got it right in the "post" quiz review, did it show that the mastery level went up? I wonder if there is a hint there, that even though you get it right, it doesn't show the mastery going up...
it really was 8 days ago. No problem thought, I am sure you are getting tons of messages so one might easily slip away. And this bug is really not going anywhere. :-D
To your question: I did not check if the mastery level increased or not during the post quiz review, but I can look for it in the future. :)
Is it sufficient to look for "green text" in mastery levels during the quiz? Does this guaranties that mastery was indeed increased? Or do I need to somehow dump mastery before the quiz and then compare it after the quiz?
I have also generic questions about the mastery:
Mastery is tracked for each vector of word separately and it shared across word schedules. Is this correct?
The overall mastery is arithmetic mean of enabled vectors in schedule. The overall mastery level of word can therefore be different for each schedule depending on enabled vectors. Is this correct?
How is mastery mapped to the Levels? Is it just linear function which takes the overall mastery as parameter?
How does the mastery for "Listening" vector work? It seams there is only one vector tracked, but there are 3 "sub-vectors" (definition, Kana, kanji compound).
If the word is answered incorrectly in the quiz the mastery for the corresponding vector decreases, if answered correctly in post quiz review it increases again But probably not so much to the original value, correct?
BTW: I have not encounter the bug in past days. (But I was not active during the weekend at all.)
There was no effect.