let's consider a few things with this: 1) what is tesseract providing to ghostscript? 2) what happens without that functionality? 3) how is ghostscript utilized in other applications that may be relying on that functionality?
1) tesseract provides OCR. OCR is important not just for things like LLM ingestion of images, but also for accessibility (it's what makes it so a screenreader can understand a PDF). 2) remove tessarct, remove OCR from ghostscript 3) is the thing I really don't know right now. Ghostscript is in _tons_ of places. I know it's a part of poppler and cups. poppler is a PDF renderer and cups is our printing stack. I _don't_ know how ghostscript is used in those stacks, and if removing tesseract causes issues with functionality. I don't know if ghostscript is used by something that _strictly requires_ the OCR capabilities. I see things like diffoscope which does in-depth comparison of files...which I don't _know_ if it needs OCR for in-depth comparison of PDFs (or if it works nicely without it). this gets me to: a) Vendoring is not a great thing. b) we do have tesseract in the distro. c) debian is keeping tesseract up to date d) I don't _know_ what version is being vendored in ghostscript e) I don't know if tesseract would pass an MIR, but vendoring it to avoid an MIR is not really ok in my book Where this lands things for me is still ??? -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/2174616 Title: Remove repacked tesseract. To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/ghostscript/+bug/2174616/+subscriptions -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
