On Tue, 07 Nov 2017 13:13:40 -0500, John H
wrote:

I'm thinking a 'ballistics database' as used above might be a database of casings or bullets that
have been recovered from a crime scene.

.... and tested and cataloged and entered into the database in some
searchable form.
My bet, not that many.
I know on TV they are always saying "that bullet matches an open
murder from 10 years ago" but it sounds like TV bull**** to me because
that same show has some geek comparing the bullets under a microscope.
How cumbersome would it be to physically compare 11,000 bullets a
year, to maybe a few hundred thousand from past years, just from
murders? I understand computers could narrow the search but the minute
differences still require actually looking and using more than a bit
of opinion, the main flaw pointed out when they talk about the problem
with forensics.
The classic case is a guy in the US who was positively identified as a
murderer in Europe from fingerprints, confirmed by the FBI, Interpol
and local "experts" and it turned out he was never even in Europe.