WKD lookup

Jürgen Schäpker Juergen.Schaepker at giepa.de
Sun Nov 6 14:06:32 CET 2016


Hi,


 >This doesn't seem to be a concern of this (draft) standard. HTTP Host: 
is well
 >standardized and defined. You should solve this there, in your 
webserver/proxy.
 >This is just sensible layering of protocols. You should strive to solve 
the
 >problems in the correct layer. Sometimes a little layering violation is 
still
 >the lesser of evils, but I really don't think that applies here.

You will not increase the adoption rate for OpenPGP if the standard just 
creates unnecessary requirements for no good reason.

I see no advantage why the hash should be calculated from local-part 
only and not from the full Email-address.

 >*Especially*
 >with IDN! That only makes it worse. How many independent IDN 
implementations
 >in how many layers do you want to add?

One might be enough, on MUA and WKD server. libidn ToASCII should be 
able do it, not?


best regards,
JS


More information about the Gnupg-devel mailing list