Sebastian Messmer

It's going to be part of CryFS and released together with it. On November 25, 2020 3:44:32 AM ignacio-reggiani wrote: > > Might as well leave the encrypted files where...

Potentially yes. But development of the recovery tool has somewhat stalled at the moment, not sure when it will be finished.

CryFS blocks are encrypted individually. Even if some blocks are corrupted, other blocks should still be readable. - If a block that is part of a file is corrupted, the...

This looks weird because it tries to build a dynamic library `libcryfs-cli.so`. CryFS should build everything statically by default.

CryFS in the `develop` branch now uses Conan for getting the boost dependency. Can you retry? I think with Conan, it should work now.

We just added support for the noatime option in the develop branch, this will be part of the 0.11.x series once that is released. This should hugely improve your performance....

Thanks for reporting. I'll try to reproduce this.

I did some investigating. With DokanY 1.2.0, CryFS gets a read request from DokanY but the path is garbage, just some random data and not even a valid string. This...

This is a verbose CryFS log from right before the crash: ``` [2019-05-27 19:05:25.377] [cryfs] [debug] read(/CrystalDiskMark000C1597/CrystalDiskMark000C15DA.tmp, _, 65536, 67239936, _) [2019-05-27 19:05:25.378] [cryfs] [debug] read(/CrystalDiskMark000C1597/CrystalDiskMark000C15DA.tmp, _, 65536, 67043328, _)...

I fixed the issue in https://github.com/cryfs/cryfs/commit/251272b517adfaee4b1499849d567e689ca85503 . However, I then ran into another issue, namely the benchmark would complete but the file system would crash after the benchmark completes. This...