Cleanup #27
Cleanup #27
Conversation
jnvsor
force-pushed
the
cleanup
branch
2 times, most recently
from
1b8921d to
be45fe5
Compare
|
Looks like this is not conflicting with changes introduced by #25. Could you resolve those conflicts? Also, I'd appreciate if you could split up the clean-ups into multiple commits. It would be much easier to review. |
|
Here's a branch that will merge - I'm going to start working on splitting it up now |
The salt struct exists mostly for e4crypt's input handling and to keep a single type across multiple subcommands. It's of no use to pam_e4crypt so we get rid of it. Since we're not parsing user input and not using a salt we'll also get rid of salt_parse and just use the raw contents of the salt file as a salt.
We're storing salts in files, and we don't have mtab keys any more, so we can ignore the filesystem encryption salt.
|
That should do it. Almost all of it was removing the user input parsing and refactoring related boilerplate |
|
LGTM |
|
You may want to rewrite the salt generation commands in README.md |
|
Honestly at this point I think we're better off using fscrypt since that seems to have a large userbase and more features. The only features it's missing that are blocking me from using it is the (planned) ability to backup keys and to store pam keys on a separate home partition. |
|
I do agree with @jnvsor: fscrypt is far better maintained and has more features. Still, I'll keep maintaining this module in the future. As for the pace... well by know you probably can tell that my maintainership of this module is rather lacking. I'll do some updates to the README during the next few days, including a reference to fscrypt. |
|
Yeah that's probably true. But pam_e4crypt works with a simple setup. And I'll need some investments to migrate to fscrypt (aka works for me now, I'll check that in the far future ^^) |
Breaking change: Now takes the file contents verbatim. You don't
need the 's:' prefix any more, and the file can contain binary
data such as \0
You can simply remove the s: prefix from your salt file to get the
same decryption key after this patch.
Depends on #25 (I "cleaned up" the code mtab depended on)