Right, the whole time all I was trying to agree on is that the user should know that the
EncryptedSQLStore can’t decrypt the data inside it with the passphrase (or any other
authentication method) when trying to open the store and not on ‘read’ like it’s now. All
my suggestions were how that could be done and it seems that you have just another
solution for that. As I see it, we’ve just misunderstood each other (or it might be just
on my side).
To wrap it up, if there are plans that EncryptedSQLStore#open would be the method to throw
InvalidKeyException and not other methods (well, of course other methods might be throwing
this too, but that’d be just in some corner cases in which the database was tampered after
it was already opened), I’m all good with that.
—
Tadeas Kriz
tkriz(a)redhat.com
On 21 Jan 2014, at 15:39, Bruno Oliveira <bruno(a)abstractj.org> wrote:
Yes, we are under the data encryption domain right?
Either way you are free to suggest whatever you think is ideal for you, if the team agree
on that, I would respect their decision. Even if it doesn’t sound right for me.
--
abstractj
On January 21, 2014 at 12:36:24 PM, Tadeas Kriz (tkriz(a)redhat.com) wrote:
>> So, does that mean that when user calls ‘open’ on EncryptedSQLStore,
> he’ll get an exception if he used wrong passphrase (or for example
> keystore) ?