I’m not sure it had millisecond on the old mono. Interpreting trailing zeros as seen on Lastmodified image:
7 is 100 ns
6 is 1 us
5 is 10 us
4 is 100 us
3 is 1 ms
2 is 10 ms
1 is 100 ms
0 is 1 s
The first timestamp of the image evaluated at https://www.venea.net/web/net_ticks_datetime_converter:
Portability meaning what? There’s both mono version issue and filesystem issue (FAT32 is 2 seconds…). There’s also a question of restore accuracy if (for example) times are rounded to some approximate time.
The idea was given in the target framework upgrade that an in-spec system would follow the manual, and install latest from mono-project (unless they can’t, e.g. Synology). SynoCommmunity has v5.18.0.240-12.
https://repology.org/project/mono/versions shows that (ignoring mono-project for now) distros that got v5 generally got v5 beyond v5.12.0, so possibly the majority of users are past danger point (and not noticing).
Survey for Linux users, which version of Mono do you have installed? isn’t a random sample, but got data. The impact of the problem “should” be rescans and a relatively small upload for the new metadata blocks. Personally, I favor Duplicati doing what it can within its limits and without attempting to dumb things down.
This is getting far from the original topic, so maybe discussion can go elsewhere now that it’s understood.