Re: [TLS] Encryption of TLS 1.3 content type

[Daniel Kahn Gillmor <[e-mail]>]

On 07/28/2014 10:39 AM, Martin Rex wrote:
> I also object to the removal of the ContentInfo from the outer
> TLS record protocol.  I'm not aware of the slightest rationale for
> this severe backwards incompatibility, that will not just break
> middle-boxes, but also applications that parse TLS records for the
> purpose of non-blocking operation.

Parsing TLS records is still possible, since the length field will be
present.

I tried to highlight your concern about non-blocking applications to the
TLS WG during the interim meeting last Sunday, but clearly failed to do
it justice because none of those attending the interim (including
myself) appeared to understand what specific toolchain you were trying
to preserve.

Can you provide more details?  Applications that just look at the raw
TLS records will still be able to tell when the communication shifts
from cleartext to non-cleartext, but they should not be able to see more
than that about the ciphertext.

In what case is this a problem?

	--dkg