against using user@host identifiers

[Brad Fitzpatrick]

On Thu, 2 Jun 2005, Xageroth Sekarius wrote:

[snip]
>  but then, what exactly does OpenID hope to achieve? The same
> thing LID does? or more? less?

I didn't understand the top of your message but I can reply to this.

OpenID's relation to LID:  I have a lot of respect for the LID project,
and I wish I would've found it earlier.  LID and OpenID are nearly
identical and have a lot of the same ideals, namely that URLs are
identities.

But I think LID tries to do too much and is too heavy to get adopted.
XPath, vCARD, FOAF.... c'mon.  Just because you /can/ bundle it all
together, that doesn't mean you should.  Every new dependency means
the bar is raised and you get less developers.

I think profile exchange (perhaps exactly as LID does it) can be built
/atop/ OpenID.  But I see no reason to bundle them together.  I think
they're separate layers, just as HTTP is a simple base layer for
transferring files... it's not tied to HTML or XHTML or MP3, etc.

Also, it seems LID requires the identity URL to be a dynamic page.  Is it
possible with LID to have an identity URL that's a static HTML page with a
link tag that punts the hard work to another URL/service?  I hope I
misunderstood, but it looks like LID is ruling out a lot of web publishing
tools that generate static files from being LID URLs, or forcing them to
have ugly /cgi-bin/ LID URLs.

- Brad