3 ms·
I think the idea is that it being case preserving is sufficient to allow decrypting the filenames. And I guess it relies on the 31 (32?) bit blocks, to make co
by ncruces 13d ago
I think the idea is that it being case preserving is sufficient to allow decrypting the filenames.
And I guess it relies on the 31 (32?) bit blocks, to make colisions unlikely for short filenames.
- jasomill 13d agoThis is true, but I'm not sure if I trust users to preserve case in case-insensitive filenames when copying files around, possibly at some point passing through something like a web server that may apply case normalization. Sort of like timestamps on files, they're handy for things like synchronization, but I hesitate before relying on them as a primary source of truth for anything mission critical.