Re: [keyassure] I-D Action:draft-ietf-dane-protocol-00.txt

[Ilari Liusvaara <[e-mail]>]

On Mon, Dec 13, 2010 at 02:00:01PM -0800, [e-mail] wrote:
> A New Internet-Draft is available from the on-line Internet-Drafts directories.
> This draft is a work item of the DNS-based Authentication of Named Entities Working Group of the IETF.
> 
> 
> 	Title           : Using Secure DNS to Associate Certificates with Domain Names For TLS
> 	Author(s)       : P. Hoffman, J. Schlyter
> 	Filename        : draft-ietf-dane-protocol-00.txt
> 	Pages           : 9
> 	Date            : 2010-12-12

Various unclear things:

- What protects against app managing to load junk keys out of wrong format
  certificate (those junk keys might matematically work an be weak)[1][2]?
- Since hash #0 is identity anyway, is there reason not to merge 1 and 2
  (and also 3 and 4)?
- Do type 3/4 entries match end entity certificate as well, or only other
  certificates?

And then what to do on encountering various kinds nonsense, what should
happen when encountering:

- Record with first two bytes of 01 00?
- Record with first two bytes of 02 02?
- SHA-1 record that is 21 bytes total (truncated record)?
- SHA-1 record that is 23 bytes total (extra stuff after the end)?


[1] draft-hoffman-keys-linkage-from-dns-03 had protection against this.

[2] Granted, it seems pretty unlikely that application can pull anything
out of wrong format certificate. But if there ever will be another 
certificate format using ASN.1, that might not hold anymore...

-Ilari