Between looking for work and finishing articles, I have been working on a book on Achaemenid warfare which bears a certain similarity to a 2018 Innsbruck PhD dissertation and should be released this year. In Austria you make the mechanical fixes and the changes in response to the committee’s comments after the thesis is accepted, not before (in Canada, you are normally given a list of changes by the committee, make them, and pass the revised version back to the committee for them to approve before you are granted the title).
I never converted to citation-management software, preferring a simple word processor file with bibliographic information and notes on everything I had read, wanted to read, or thought I might one day want to read. When I was assembling the different files into a dissertation, I stripped out the metadata and dumped the individual entries into the bibliography then sorted it alphabetically with Tools -> Sort. So one problem I had is that some works in the footnotes were not in the bibliography, and some notes were in different formats than others. To sort this out I went through each chapter recording the works cited, then removed duplicates and standardized the format, then combined the eight separate lists into one and removed the duplicates again. I checked that list against the bibliography, making sure that everything in the footnotes was in the bibliography.
And that leads to the important question, out of the roughly 1,232 works in the final bibliography (77 pages x 16 citations per page), how many do I actually cite?