- From: Henry Story <henry.story@bblfish.net>
- Date: Wed, 10 Oct 2012 15:34:56 +0200
- To: nathan <nathan@webr3.org>
- Cc: public-xg-webid@w3.org
- Message-Id: <6178E435-7F75-44EE-BE61-7D7967542E43@bblfish.net>
On 10 Oct 2012, at 15:19, Kingsley Idehen <kidehen@openlinksw.com> wrote: > On 10/10/12 9:06 AM, Henry Story wrote: >> On 10 Oct 2012, at 14:56, Kingsley Idehen <kidehen@openlinksw.com> wrote: >> >>> On 10/10/12 7:51 AM, Henry Story wrote: >>>> On 10 Oct 2012, at 13:34, Kingsley Idehen <kidehen@openlinksw.com> wrote: >>>> >>>>> On 10/10/12 5:12 AM, Henry Story wrote: >>>>>> But I think the question is here: is a 301 also a signal of a change of WebID, since >>>>>> the document has moved permanently, do #uris in the old document also need to be >>>>>> moved permanently? >>>>>> >>>>>> Henry >>>>> Should it matter how routing occurs if the end result is a WebID and public key association that mirrors what's in the X.509 certificate? >>>> That would be the simple answer - and that is I suppose the current position. >>>> >>>> But I am wondering about this, because when you have a 301 redirect the resulting document relative URLs get interpreted according to the redirected to URL. >>> Yes. >>> >>>> See Tim Berners Lee's point here: >>>> http://lists.w3.org/Archives/Public/public-webid/2012Apr/0006.html >>>> >>>> So what happens then to #uris. After all they are also relative, right? >>> Yes. >>> >>>> But now the question is >>>> the other way around if you do a #uri on a moved document do you have a #uri on the moved document? Do you query the one or the other? >>> Linked Data URIs simply denote entities. In the case of the WebID protocol, these entities are of type: foaf:Agent. All you ever do is de-reference the URI in the certificates SAN. Everything else is HTTP message exchange that's in effect routing. As you know, a hash URI simply delivers implicit rather than explicit indirection within the context of the Linked Data meme. >> But for example some of the things we could look at here is what the HTTP/HTML specs say >> about 301. Are browsers not meant to create an association directly to the permanently moved >> resource once the have received that message, and never come back to the intermediary resource? >> Are blogging engines not meant for example to take this as a signal, to change the URLs of the links in the blogs to the new links? I think "Moved Permanently" has some important semantics. > > Yes, to all of the above because they are all about realms where URIs denote Web documents. \ Good so perhaps we can have Nathan's view on this now. Nathan do you think that 301 redirects can be a signal for people who are linking to other people's WebID to change it? Should it be? Would this be one way to solve a problem of WebIDs changing. Clearly it is easy for the user to change his WebID certificate. The bigger problem is for the user to tell all the others to change their references. A 301 seems to be just such a signal - useful if you have some time in between a move. Of course if it you loose your server is hijacked then you will have more trouble - but you can have recourse to justice then still: that is where DNS is useful. Lost private keys cannot be dealt with in such a way. > With regards to the Linked Data based WebID protocol a URI denotes an entity of type: foaf:Agent. What's being denoted by a URI and the explicit semantics of the entity relationship graph to which it eventually resolves is what matters re. the WebID protocol. >> >> Perhaps something to bring up on the LinkedData and HTTP mailing lists. If someone would like to do that... > > That's fine, but we will arrive at the same destination, eventually. Linked Data is just about denotation applied to entities in general as opposed to Web documents. The evolution of denotation on the Web and Internet is as follows: > > 1. Internet -- de-referencable names denote machines via DNS protocol > 2. Web of Linked Documents (WWW) -- de-refrenacable names denote entities of type: Web Document, via HTTP or any other data access protocol > 3. Web of Linked Data (GGG) -- de-referencable names denote entities of type: owl:Thing, via HTTP or any other data access protocol . > >> Social Web Architect >> http://bblfish.net/ >> > > > -- > > Regards, > > Kingsley Idehen > Founder & CEO > OpenLink Software > Company Web: http://www.openlinksw.com > Personal Weblog: http://www.openlinksw.com/blog/~kidehen > Twitter/Identi.ca handle: @kidehen > Google+ Profile: https://plus.google.com/112399767740508618350/about > LinkedIn Profile: http://www.linkedin.com/in/kidehen > > > > > Social Web Architect http://bblfish.net/
Attachments
- application/pkcs7-signature attachment: smime.p7s
Received on Wednesday, 10 October 2012 13:35:35 UTC