Re: [mmox] Creating walled gardens considered harmful

[Jon Watte <[e-mail]>]

Morgaine wrote:
> On Tue, Mar 31, 2009 at 12:35 AM, Jon Watte <[e-mail] 
> <mailto:[e-mail]>> wrote:
> teleport grant or denial might validly be considered part of the 
> "mechanism" of teleport.
>
>

Fair enough.

>
>     Do we really want to use the term "landmark" in interop speak?
>     That sounds a little too SL specific? Else, would you define what
>     a "landmark" really means?
>      
>
>
> The term "landmark" is in extremely wide use in many applications that 
> involve maps or addresses, as well as in real life

I know of no serious GIS application that uses the term.

But the use in the real world is very different. In the real world, the 
Coit Tower cannot move. In the virtual world, that's just a few mouse 
clicks. If I had a delicious hot dog right next to the Coit Tower, and 
the Coit Tower then moved, in real-world speak, the Coit Tower landmark 
is no longer near anything it was near when I set that as a Landmark.

In general, I like the term "address" more, because it's more akin to, 
say, GPS coordinates. Also, when using Google Maps (arguably the most 
used GIS system ever), they are called Favorites.

But this brings up another use case point: Is the "landmark" referencing 
an object (that can move), or an address (that can't)? And does it even 
matter? Maybe these location favorites can simply be URLs with a known 
type, and the actual words of the URL are specific to the system that 
minted them. Then the only information that is important to the 
destination system is:
1) Knowing that this is a location URL, and not some other kind of URL
2) Some user-visible description (icon, title, movie, sound, whatever) 
for presentation
3) The process by which one of these URLs get turned into an established 
session

Sincerely,

jw