@@ -28,7 +28,10 @@ is that can be used to decrypt the footage (or at least claim that they do not
28
28
have posession of a device capable of decrypting / someone else has to decrypt
29
29
it)
30
30
31
- ## Potential User Flows:
31
+ ## Potential User Flows on First Launch:
32
+
33
+ Either of these flows could be implemented (or both). Perhaps one would work
34
+ better, this would probably require user testing to find out.
32
35
33
36
### 1. one device records, another device decrypts
34
37
@@ -63,9 +66,29 @@ unlock the footage.
63
66
64
67
### 2. single device, generated passphrase derives private key
65
68
66
- TODO
69
+ Upon launching the device for the first time, the device will randomly generate
70
+ a pass phrase that can be used to derive a private key. This phrase is
71
+ displayed to the user, and they are asked to write it down and store it in a
72
+ safe place (perhaps multiple places).
73
+
74
+ After the user has confirmed they have done this, the device forgets the phrase
75
+ and private key, and stores the public key only. The user is then able to
76
+ record footage.
77
+
78
+ #### Benefits
79
+
80
+ * This method only requires a single device.
67
81
82
+ #### Potential Problems
68
83
84
+ * It would potentially be more easy for a user to misplace the passphrase than
85
+ it would be to misplace an "unlocking device".
69
86
87
+ ## Potential Enhancements
70
88
89
+ ### Add additional "unlockers"
71
90
91
+ The user could potentially choose multiple unlocking keys (recipients) to add
92
+ redundancy. For example we could allow them to scan further QR codes, or to
93
+ enter in additional passphrases, or choose from a number of pre-defined
94
+ recipients (e.g. ACLU or some other org).
0 commit comments