[foaf-protocols] what if

[Nathan]

Henry Story wrote:
> On 18 Jul 2010, at 19:22, Nathan wrote:
> 
>> we simply stick the public key in an http header instead.
> 
> No need to change the protocol, please! We already have lots of implementations that show how foaf+ssl works very well.
> 
> We are really not all in such disagreement. We all agree that authentication agents should at a minimum understand RDF/XML and RDFa, and that producers should
> publish either one of those. 
> 
> The disagreement is on how to say that, where to say that, and how much flexibility to give for future syntaxes. 
> 
> Perhaps we should be arguing more about wording now. I will try to put something together. But we can also come back to it later.

I hate to say it, but if it was simply a header, there would be no 
disagreement, barely any discussion needed and we'd all be a lot further 
down the road now.

At the same time though, obviously i still see benefits in the way we 
currently do it, because it ensures we get some linked data back - 
popping it in a header would negate that.

Similarly though, if the protocol does need to change to handle all 
variations in the future (or even simply new recommendations) then maybe 
alternatives are worth considering.

And hopefully you realise that in no way am I attempting to undermine 
any of the hard work done so far, simply throwing in an option in to the 
mix which may make things easier for us all.

Best,

Nathan