W3C home > Mailing lists > Public > public-webcrypto-comments@w3.org > November 2012

Re: security of a client-side JS API?

From: Arthur D. Edelstein <arthuredelstein@gmail.com>
Date: Thu, 1 Nov 2012 08:39:38 -0700
Message-ID: <CADHWJb5VDYRRNZyAkM+f-fmcMwG-PeVY0iKGzAxdCfd0uLnzEg@mail.gmail.com>
To: Eric Rescorla <ekr@rtfm.com>
Cc: Zooko Wilcox-OHearn <zooko@leastauthority.com>, Ryan Sleevi <sleevi@google.com>, public-webcrypto-comments@w3.org
Hi All,

On Thu, Nov 1, 2012 at 3:24 AM, Eric Rescorla <ekr@rtfm.com> wrote:
> As Zooko says, WebRTC provides a mechanism for establishing an
> end-to-end cryptographically protected data channel (for those who
> care, SCTP over DTLS. These channels can be created and accessed by
> JS.

Thanks for the suggestion about WebRTC. Please forgive my ignorance --
if the WebRTC data channels can be accessed by client-side JS, doesn't
that mean that messages can be read by the web app launching the
channel?

My concern is that browsers currently provide nothing out-of-the-box
for users who want encrypted communications that cannot be read by a
web app provider. This situation has resulted in a rampant online
privacy problem, as I'm sure everyone here is very aware.

Many users are going to perhaps have difficulty understanding the
nuance that the WebCrypto API doesn't give stronger privacy
protections than existed before. So I'm struggling to see the benefits
of a new crypto functionality for web apps that may require users to
make new security decisions (such as providing keys or signing
documents).

Best regards,
Arthur
Received on Thursday, 1 November 2012 15:40:05 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 1 November 2012 15:40:06 GMT