Stay tuned for a large new experimental dataset applying the open source Tesseract OCR package to a year and a half of CSPAN coverage next week!