Hard code stat results so that the tests pass on a different checkout
w/ different mtimes..

Looks like this is the only way:
https://github.com/sqlalchemy/alembic/discussions/1086

add tests for migration, good example for future needs..

We were only using the ed448 code, and cryptography requires rust
to compile, and contains a LOT of other code, while edgold is much
smaller, and doesn't require a rust compiler, and will work on more
systems..  Keep around an old identity generated from cryptography
to make sure the two libraries are compatible.

one minor issue is that the identity object gets dump'd twice..

make sure tags/properties are unique...
support tuples in asn1coder, for the tag cache
make sure that only one of each uuid is stored in hashes, and that
it is the last modified version.
for containers, since we're uniquified them, just replace them..

this fleshes out hostuuid better..
add a more useful debug print..

If the identity changed, generate a new uuid, and set created_by and
modified appropriately...

the modified is the correct, not mtime, as someone can put an
"older" modified file, but at a later time, so it's assumed
that the most recently modified fileobject has the most up to
date data..

The later happens after importing a container.

git-subtree-dir: ui/medashare/btv
git-subtree-mainline: 4a28327b7d
git-subtree-split: 85f5c9a03e

https://www.bittorrent.org/beps/bep_0003.html

This uses an index to quickly look up what pieces are part of a
file, and then checks that they are all valid, this should be faster
as it is likely that the torrent has more pieces than files (few
large files, vs many, many small files)..