Re: Hash Discussion Overview II - the return

The WebID is an http URI is a marketing decision to help us make it
simpler to specifying things. Other specs make these types of decisions.
There is nothing stopping future specs being more relaxed, just as the
meaning of HTML is constantly evolving.

We are not saying that *only* http URIs can refer to agents. We are just
restricting ourselves here to HTTP URIs for works in different specs, such
as LDP etc, WebID Auth, etc... 

The Identity Interoperability work
 http://www.w3.org/2005/Incubator/webid/wiki/Identity_Interoperability
shows that you can work with a lot of other identifiers, but it is a lot more work
then to specify things, and it means that people looking to implement a WebID
have a much steeper curve to get in.

Apart from that is there a short argument you would like me to add to this,
for your position, in section 3, that does not cover it? 

I added the following for you:

"It is better not to have restriction here since there are technical solutions to get to the profile document for both hash URIs and non hash URIs."

Henry

On 8 Dec 2012, at 20:03, Toby Inkster <tai@g5n.co.uk> wrote:

> On Sat, 8 Dec 2012 17:35:35 +0100
> Henry Story <henry.story@bblfish.net> wrote:
> 
>> http://www.w3.org/2005/Incubator/webid/wiki/WebID_Definition/hash2
> 
> FWIW, I don't think WebID should place any restrictions on users'
> choice of URI; just as it shouldn't place any restrictions on what
> ciphers are used for the TLS sessions established.
> 
> I'm not saying that restrictions should not exist. People shouldn't be
> using, say, a Caesar cipher (look it up if you don't know) for TLS; but
> restrictions on TLS ciphers should happen in the TLS specs, not in
> WebID.
> 
> The WebID spec is the wrong *layer* to address this sort of issue. It's
> an issue that needs resolving (even if that resolution might be that
> the status quo is OK) at the linked open data level; or maybe even at
> URI.
> 
> So WebID shouldn't place restrictions on what URIs people choose to
> identify themselves with. I don't even think we should require
> HTTP/HTTPS; if people choose to use an FTP URI, chances are that most
> existing implementations of WebID would cope. If they choose to use
> an NNTP URI... well, I tend to be in favour of giving people enough rope
> to fashion themselves the very best noose possible.
> 
> Be liberal in what you accept; be conservative in what you omit. The
> spec should accept whatever URIs people want to use; how-to guides
> should steer people towards sane options.
> 
> -- 
> Toby A Inkster
> <mailto:mail@tobyinkster.co.uk>
> <http://tobyinkster.co.uk>

Social Web Architect
http://bblfish.net/