Subj is an OCR system which needs to much love to be alive that I
haven't. It's not required from any other package, it never had upstream
from he very beginning and it is a `tesseract` OCR to switch to. So it's
time to say good bye for me.
It would be nice if anybody could adopt it (but I do not believe this
could ever happen).
See also:
https://bugzilla.redhat.com/show_bug.cgi?id=1799265 and I
believe there are no opened bugs.
Package page:
https://src.fedoraproject.org/rpms/cuneiform
Regards,
Dmitrij