Yes, I agree TTS for xournal is probably not the best feature to implement. I was just answering question of the original poster's inquisition regarding TTS.


What I was referring to with the stroke capturing would use a block (read: word), the strokes, and probabilities to guess which word was written, rather than doing it on a character by character basis.

But this is just something that's possible if we start separating strokes. The more immediate benefits are word wrapping, auto-newline, auto scrolling, text styling, paragraph justification, text boxes, etc, which would IMO improve xournal greatly for note taking.

Ignoring hand writing recognition for a second:

Consider being able to define a box of text, and have the words that you've written re-organize themselves as you reshape/resize the box. Consider also, that the page automatically scrolls and newlines as you are writing (even if it doesn't understand *what* you are writing) without having to do this manually. With this in place, a user could easily write as fast on xournal as they would using paper even given the N900's relatively small screen. Remember, that as part of these scenario's xournal is only grouping strokes together as words and does not need to concern itself about what those strokes represent.


}:^)~