Re: [Sks-devel] Launching a new keyserver on keys.openpgp.org!

[Andrew Gallagher]

On 19/06/2019 09:42, William Hay wrote:
> Thus spake Valentin Sundermann:
>> I think the best way forward would be to implement SKS Recon, this way
>> the SKS instances would not fall behind the hagrid ones (what's good for
>> the general network I guess).
>>
>> I'd suggest to provide an in/out sync interface where something like an
>> "sks recon adapter" could be plugged in. Such an adapter would strip
>> away all identity information in- and outwards.
> 
> That doesn't seem likely to work.  AIUI the sks recon protocol doesn't just 
> ensure that all members of the network have a copy of every key but 
> that they have the same version of each key.  If the recon adapter 
> only deals in stripped keys then the reconciliation could never finally 
> succede

This is correct. To support SKS recon between servers that support a
different subset of key material would require a full implementation of
fake-recon as discussed in the mega-thread here:

http://nongnu.13855.n7.nabble.com/SKS-apocalypse-mitigation-td228252.html

tl;dr: you probably have better things to do with your life. :-)

-- 
Andrew Gallagher

signature.asc
Description: OpenPGP digital signature