Re: [TLS] rfc7366: is encrypt-then-mac implemented?

["Christian Kahlo" <[e-mail]>]

> >But as stated before I'll have a look in the final RFC and compare it
> to the
> >currently deployed implementation.
> 
> See the more recent messages about this though, there are two possible
> interpretations, both equally valid, so don't go changing things just
> yet :-).

Done so. After reading some dozen eMails and bug reports at different
sites (Google, Mozilla) I think I'm now up to date what the whole
discussion is about.

I understood "TLSCipherText.length" in the MAC-function as the length
of the encrypted and padded data because the MAC is calculated exactly
over this - not anything else. This seemed quite logical to me. But I
now realize that fragment.length != length of the final data block on
the wire. It's more a question of wording - and how it is interpreted.

BTW: @all, after seeing all this: it is far, far easier to talk
to Peter or me directly in case of doubts about EtM than to open
tickets on systems all over the world we aren't aware of. ;-)

So, eid.vx4.net will keep its implementation for all those who
want to check and discuss RFC7366.

Best regards,
Christian