Re: How to express dereferenceability? (ISSUE-91)

On Monday 19. January 2015 16.13.30 Ruben Verborgh wrote:
> Note BTW that I personally don't agree
> that a server should express dereferenceability:
> - the server doesn't know what the client wants to do
> - the client can deference easily anyway
> However, I launched this issue because others think differently.

Yeah, so, basically I think it all boils down to how strict you want to express things. To express that 
something is dereferencable is a very strong statement, and in 20% of the cases, it is probably wrong, 
because that's how much of the semweb is down at any given time :-/

So, a bit of history, around 2004-ish, when I wrote my RDF::Scutter, it would by default only follow 
rdfs:seeAlso links, because we thought at the time that with the 60s interval that well-behaved the bots 
would use, and the huge number of URIs a typical RDF document would have as compared to a HTML 
document, simply blindly spidering the entire semweb would be infeasible. Instead, people should make 
extensive use of rdfs:seeAlso links so that we'd get all the data without GETting all the URIs anyway.

I still think making a commitment on behalf of anybody that something is dereferencable at any given time 
is a too strong statement. Instead, I think saying "this object is a resource that you really should 
dereference if you are interested in the present subject", and then there is some historical support for 
rdfs:seeAlso actually meaning something like that. 

Ah, BTW, there's http://www.w3.org/wiki/UsingSeeAlso
which hasn't changed in a decade :-)

Kjetil