This is a maintenance evolution. The only addition is a new command-line parameter -pdfnometa, which is false by default.
If set to true, PDF metadata (title, author, subject) is not extracted from a PDF document, except for the number of pages of the publication, which is needed for proper navigation in a reading system.
This is useful if the PDF metadata is considered too bad.