Re: [TLS] Update spec to match current practices for certificate chain order

[[e-mail] (Martin Rex)]

Peter Gutmann wrote:
> Dave Garrett <[e-mail]> writes:
> 
>>I'd like to propose simply changing that second "MUST" to a "SHOULD" or
>>possibly even a "RECOMMENDED", thus allowing for clients to accept cert
>>chains in different orderings as they already seem to do. (no change proposed
>>for the first "MUST")
> 
> I suspect the first MUST can go as well, particularly if you're using code
> that handles cert chains in other formats like CMS/PKCS #7, where the "chain"
> can contain any old rubbish and the chain-assembly code has to build the path.

That would be a terrible idea.  I believe that there is a non-marginal
number of TLS implementations that requires the end-entity certificate
to come first in the list, not just ours. 


>
> For example my code looks for a cert containing the site name 
> ("www.whatever.com") and then builds a chain up via the parent links until it
> can't find any more useful certs.  That just works no matter what the other
> side sends.

Adding workarounds for clearly bogus behaviour is a very bad strategy,
because it makes the situation worse for everyone in the long run.


-Martin