There seems to be a memory leak in the OCR feature. If you try to OCR a large (say 500+ page) scanned text my computer alerts that I've run out of application memory (see attached photo). This is bonkers and needs to be fixed! You've got to memcheck this or something...