-----Ursprüngliche Mitteilung----- Von: Robert Raszuk <rob...@raszuk.net> An: rrg@irtf.org Verschickt: So., 25. Apr. 2010, 23:52 Thema: Re: [rrg] Next pass Hi Heiner, > Maybe however you can admit that in case I start a trip here from > Munich, Karlsplatz Germany to San Francisco Golden Gate Bridge I don't > need the city map of San Francisco to make the right next hop here in > Munich towards the Munich airport. This means: I don't need approx. 99,9 > % of the world's road maps just to make the next hop at that place where > I am at that moment. Clearly you don't need world map to get to the MUC. To translate this to RRG language you can follow AIS default route. But once you get to MUC you better know which plane to take. And here comes a challenge. ... which I am well aware of: My composed TARA-map must contain this single-hop link between the airports of Munich and S.Francisco, although that bridges a forth of the globe. Your destination (content you are looking for) is more then likely to be very distributed - even globally. The times where user A just wanted to talk to server B has gone long time back. Distributed file systems are now reality, distributed video files are norm and there is no way to map chunks of data to the geographical coordinates. And not even due to size of few orders more then 6 million .. mostly due to the dynamics of changes which would be required. So coming back to your example in order to get to your distributed destination you better have a well defined timetable on the wall of MUC airport to direct you to the right destination. It could be mapping plane like we see in LISP or enhanced DNS like we see in ILNP except in latter (ILNP) you don't even need to get to any major hub to know which way to go. What makes you believe that no other concept than what I'd call mainstream-IETF is able to cope with new challenges?. I give you an example: A year ago Google map was not able to display a route from NY to San Francisco as perfectly as it is able today: At that time, when you zoomed out for seeing the entire blue route from start to end, and when you thereafter zoomed closer e.g. to see the details in the surrounding of Chicago, it failed. Meanwhile you can do that, too. Will say: If you have a particular problem and are supposed to solve it based on a particular architecture then you will be able to solve it, after the time it takes to solve it of course, and well, restricted by the given architecture. So this is no other sort of argument as was the partition argument I have been confronted several times. Heiner Cheers, R. _______________________________________________ rrg mailing list rrg@irtf.org http://www.irtf.org/mailman/listinfo/rrg
_______________________________________________ rrg mailing list rrg@irtf.org http://www.irtf.org/mailman/listinfo/rrg
