I\'ve used the Anki spaced repetition software (SRS) for almost 18 months and can now read and write over 2,500 characters, plus a larger amount of \"words\". So it\'s safe to say I\'m a fan of the software, and the principle behind it. But, from the start, I\'ve always drilled myself on the flashcards both ways: that\'s to say, if I write a flashcard for 拼搏, Anki will produce two flashcards for me to be tested on, one Chinese -> English, one English -> Chinese.

Aside from the practicalities of going English -> Chinese (ie often I need pinyin, or some other hint, to distinguish one word or character from another which has the same English translation) this process means that the time I spend using the software is doubled. In fact, it\'s more than doubled, because \"producing\" the characters and words is more difficult than \"recognising\" them.

I\'m not asking if in general it\'s useful to try to learn how to write -- as opposed to just recognise -- Chinese characters. My personal hunch is that it really helps me recognize them better.

But I\'d like to hear from others using SRS who do everything both ways, ie recognition and production, with characters and vocab. Do you ever think time could be best spent doing something else? Should I move to recognition-only for newer, slightly less-common vocab? Or, if you only go one-way, do you have any regrets?