On Thu, Jun 19, 2008 at 12:26 PM, Peter Saint-Andre <[EMAIL PROTECTED]>
wrote:
> How is contacting the page owner / creator less "direct"
> than finding an alternative representation of the content?
> If the web is to be truly social then ISTM that it's
> completely appropriate to provide links to content
> creators, discussion forums, microblogging feeds,
> and anything else of interest.

Clearly, the owner or a creator of a page is not an alternative
representation of the page... In any case, the long term issue here will be:
"How many and which <link rel>'s are appropriate content for a web page and
if there are any links which are not appropriate, where should they be
placed?" I hope that people can look ahead to the future and realize that
there are potentially many links that we would want to make from some
"identity root" for an individual to services and resources that are
associated with the identified individual. Today, when there are few such
links, it "works" to have them all embedded in a "home HTML" page. But,
we'll eventually get to the point when overloading <link rel=> that starts
to feel very cumbersome. At that point, we're likely to look to some
mechanism like XRDS files and we'll undoubtedly regret the number of
historical "standards" that establish the expectation that some links are in
HTML pages and other are not.

I argue that we as we move forward, we have an obligation to try to
anticipate where we're going. Let's avoid building up practices today that
are fairly certain to be considered unfortunate in the future. This is why I
argue that we should limit embedded <link rel> use to links that are
directly related to the content in which they are found. In my reading of
the situation, I would consider a link to an XRDS file which describes that
page owner/creator as something which is "related to the page." That linked
XRDS file is, however, where you should look to find attributes of the
page's owner (such as JID, email address, address, social network
membership, current location, etc.).

bob wyman

Reply via email to