We've started working on the next version of the example sentence system. (I've just set up the text segmentation backends we're going to need). Here's what we're thinking:
1. Importing sentences from Tatoeba, hopefully with a way to contribute new ones back there. (Does anyone know any other good sources of sentences with translations that are free to use?)
2. Sentences for both Chinese and Japanese.
3. User-created sentences, with editing and remixing sort of like we have with the mnemonics.
4. Definitions of sentences in all the different source languages, like we have with normal vocab.
5. Better, editable sentence segmentation into real component words.
6. Better sentence cloze deletion, where just the characters/words of interest can be blanked out instead of the entire word segment containing the character of interest.
7. Perhaps: a way to indicate which words the sentence is a good example for.
8. A way to study the sentences directly, as if they were normal Skritter words, but perhaps not doing the writing prompts (or not by default, anyway).
9. iOS implementation, after we've got it running well here.
We're looking for feedback on 8 in particular. Does anyone want to be able to study entire sentence writings, or should we leave that out? What about sentence tones?
If y'all have other suggestions that we could work into this, we'd like to hear them--keeping in mind that sometimes when building features, easy-sounding things can be unfeasibly hard, and vice-versa. Now that we have ways to do automatic (although not perfect) text segmentation, there are some new possibilities in what we can do, like turning big chunks of text into component words.