Not a grave problem, but it seems to me that some code could be optimized in this respect:
When I add words/characters via the queue and study all the items that are ready, often only one variant [either simplified _or_ (not _and_) tradititional] is added in the first session, even if I wait long enough to be theoretically able to work off both types.
The really conspicuous issue arises when I get back to Skritter the next time: Although I may have, for instance, 200 items ready for review, the new words/characters of the _other_ variant (e.g. simpl. if trad. was revise during the first session) wait for me to be answered before I have started to revise the items that are to be reviewed, i.e. "New word" prior to the ">100% items").
Short question: Bug or feature?