Closed GerHobbelt closed 5 years ago
Done as per #33.
This really is a culmination of #17, #11, #13 (in particular!), #20 and other unlisted work.
It still MAY happen that Qiqqa + QiqqaOCR are lingering after close when you've just clicked/imported several PDFs via Qiqqa Sniffer as Qiqqa MUST wait for QiqqaOCR to finish the (small) OCR batch it was handed.
As James Jardine already noted (possibly quite long ago) in the sources, QiqqaOCR was a then-quick-hack to circumvent unsolvable issues with Tesseract and memory leaks, so this will only improve once we are able to embrace Hong Kong, pardon, Tesseract 🤡 into the Mother Mainland. That's what https://github.com/jimmejardine/qiqqa-open-source/issues/37 is meant for.
Closing and decluttering the issue list so it stays workable for me: fixed in https://github.com/GerHobbelt/qiqqa-open-source mainline=master branch, pending #15 / any maintainer rights/actions.
The only remedy is to reboot or
kill -9
Qiqqa et al, i.e. open the Windows Task Manager as Adminstrator and end all Qiqqa processes. ('end process tree')