OpenID Email Discovery

Artur Bergman sky at crucially.net
Fri Jan 4 09:59:12 UTC 2008


On Jan 4, 2008, at 10:16 AM, Trevor Johns wrote:

> On Jan 4, 2008, at 12:45 AM, Artur Bergman wrote:
>
>> On Jan 4, 2008, at 7:28 AM, Trevor Johns wrote:
>>
>>> 6. I can't see how this can be used securely. DNS is highly  
>>> vulnerable
>>> to attack.
>>
>> Which is why the internet isn't working at all. Ever, Never!
>
>
> Hey, that's not fair!
>

Fair or not, I am tired of hearing how un-secure DNS, when everything  
we do is based on it, and it being the worlds largest working  
distributed database.

> DNS is well designed to handle denial of service attacks, but  
> there's relatively little protection against false information  
> within a localized part of the system.
>
> For one, there's cache poisoning:
>
> 	http://en.wikipedia.org/wiki/DNS_cache_poisoning
>
> Aside from that, there's nothing stopping an upstream DNS provider  
> from injecting falsified records.
>
> This is why we have SSL CAs and discourage self-signed  
> certificates. Encrypting information is easy. Making sure it came  
> from who you think it came from is not, particularly when you're  
> dealing with previously unknown parties. Thankfully these attacks  
> are a little tricky to pull off, but they're certainly not outside  
> the realm of possibility.
>


There is SSL connecting to the provider that is being refereed from  
the srv/txt field. Which is no different than what you are referenced  
to from an A or CNAME or MX

Not that it matters anyway, since people just click OK.

Cheers
Artur



More information about the specs mailing list