The OCR feature in master is broken, there are several reasons for this: - https://github.com/naptha/tesseract.js/issues/369 - https://github.com/naptha/tesseract.js/issues/346 - and the fact that the worker JS file is served with an unversioned URL, and, in the nginx config we provide, served with "public, max-age=31536000, immutable", leading to version mismatches as the old version gets cached
5.9 KiB
5.9 KiB