mirror of
https://github.com/Llewellynvdm/Tomb.git
synced 2024-12-24 11:05:30 +00:00
37 lines
1.2 KiB
Plaintext
37 lines
1.2 KiB
Plaintext
|
Overview
|
||
|
|
=========
|
||
|
|
|
||
|
|
|
||
|
|
What's a key?
|
||
|
|
It basicly is a gpg simmetrically encrypted, ascii-armored file.
|
||
|
|
It's encryption key is a function (see below, on KDF section) of your tomb
|
||
|
|
passphrase.
|
||
|
|
|
||
|
|
|
||
|
|
Layout
|
||
|
|
======
|
||
|
|
|
||
|
|
Before coming to the gpg part, there could be some "header" lines specifying
|
||
|
|
metatada. They're done like this:
|
||
|
|
_FIELD_params_params_and_more_params_
|
||
|
|
|
||
|
|
where FIELD should be the description for the header.
|
||
|
|
Pay much attention to the fact that there should ONLY be ASCII characters there,
|
||
|
|
to avoid encoding issues and whatever. Needs something more? Use base64encode.
|
||
|
|
(Of course, you're free to pack params into a single field, base64encoding
|
||
|
|
whatever you want).
|
||
|
|
And every header field should be in only one line.
|
||
|
|
|
||
|
|
KDF
|
||
|
|
===
|
||
|
|
|
||
|
|
Key Derivation Functions, are functions which will make your key stronger
|
||
|
|
spending some CPU time: the basic idea is that you have to compute that function
|
||
|
|
just once in a while, but an attacker that wants to bruteforce has to compute it
|
||
|
|
for every passphrase he's checking. This will make the bruteforce much more
|
||
|
|
expensive.
|
||
|
|
|
||
|
|
The header line format is _KDF_$method_$params_$params_... where $method is the
|
||
|
|
method we are using (ie: scrypt) and params is something that it needs (ie:
|
||
|
|
salt).
|