A true zero-knowledge web application knows nothing about its users and their data. We have been fascinated by this simple idea since 2005 when we started this blog. Since then it became our obsession.
We focused exclusively on exploiting Ajax and browser-based cryptography to build applications that users can wholeheartedly adopt to manage their private data. We started with a password manager, but we have more ambitious plans.
The whole point of writing a zero-knowledge application is to avoid the need for a trust relationship between the web application provider and the users.
It requires a strict discipline. It’s easy to fall for a new fancy feature that leaks some kind of information to the server destroying the whole zero-knowledge architecture. Clipperz won’t do compromises, won’t take shortcuts.
What Clipperz does know about its users
Every time a user loads the Clipperz login page, the following information are logged by Clipperz web server:
- IP address (and therefore the geographic area)
- request date and time
- browser type and operating system
All of the information above can be linked to a specific account, once the user successfully logs in. Furthermore, for each account, Clipperz web application could save the following data:
- date and time of account creation;
- dates and times of every single access to Clipperz;
- the number of cards;
- for each card: dates and times of creation, modifications and access;
- an estimate of the amount of information stored in each card, inferred from the length of the encrypted text (the estimate doesn’t include any details about the number of fields in that card or the presence of a direct login configuration);
- date and times of every downloads of the offline copy.
Actually we are not storing all the data listed above, but this doesn’t make any difference since we could!
Delicious
Digg
Reddit
Facebook
Yahoo
Netscape
StumbleUpon
But how much is 'zero', in a zero-knowledge application?
What we do know
No matter what we do, as we are running a web-based application, so there are some data that will be sent to our servers anyhow; we are receiving the same kind of information you would provide any web server when accessing its static content with your cookie enabled:
Other than this standard data, we also have access to other data for each user account:
So we are not talking about “zero kelvin”; it is more like “zero fahrenheit”.
But this is all we got. Nothing else is stored intelligibly on our servers.
What we do NOT know
Let me please list some of the notable voices missing in the lists above:
Direct logins
Direct login is a very nifty feature of our service, and we are quite proud of it, because we have being able to implement it without leaking any further information other than the one listed above:
Other services are trying to imitate our direct login feature, but no one has being able to achieve our level of privacy and convenience.
Post new comment