On Thu, 30 Jan 2020 23:34:59 +1100 Chris Angelico <ros...@gmail.com> wrote:
> ... I wasn't advocating for the use of a database; my first and > strongest recommendation was, and still is, a stateless system wherein > the files themselves are the entire indication of which documents have > been downloaded. Yes, I like stateless systems, too, but that system isn't stateless. As I understand the problem of a "crudely persistem download manager," there's a collection of to-be-downloaded URLs (which may be empty) and some data that's been downloaded (which may also be empty). You can certainly encode a lot of that directly in the file system, but it's still state. Using a database instead solves a lot of the tricky bits that the bare file system doesn't (which is what ChrisA said in what I snipped). It's just Greenspun's Tenth Rule with the words C, Fortran, and Common Lisp crossed out and Python and ACID Database written in crayon. Dan -- “Atoms are not things.” – Werner Heisenberg Dan Sommers, http://www.tombstonezero.net/dan -- https://mail.python.org/mailman/listinfo/python-list