Re: [TLS] Next steps for draft-agl-tls-padding

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

On 04/07/2014 01:40 PM, Watson Ladd wrote:
> I read the F5 explaination. It's worth keeping it around so in the future,
> when tempted to upgrade a binary protocol we ask the guy proposing the
> upgrade how to distinguish old and new.
> 
> In particular,  tossing it down the memory hole out of misguided
> professional courtesy would probably mean making the same mistake in the
> future.
> 
> In particular treating this extension's proper use as something to be
> passed on by word of mouth hurts new implementors who don't know the magic
> number. Embarrassment for F5 is nowhere near as serious.

I agree with Watson here.  Including some variant of Xiaoyong Wu's
explanation [0] in the padding draft would be useful for future
implementers.  The global network *does* still have SSLv2 endpoints on
it, and they connect to and listen on ports that SSLv3/TLSv1.{0,1,2}
peers also use.

I think Xiaoyong and F5 have done us all a favor by making this
situation clear.  This shouldn't be considered an embarrassment for them.

	--dkg

[0] https://www.ietf.org/mail-archive/web/tls/current/msg10423.html