White Papers
Mapping Shopping Process To Microsoft Commerce Server Framework
Introduction
There is one world-wide lookup service that is used by almost every computer user today, and it goes unnoticed, because it works so well and does so in a fraction of a second. This is the Domain Name Service (or DNS) that works in the background whenever you browse the web using a URL, transfer files or in any way try to connect to some remote computer system using a name like tech.Calsoft.co.in The DNS service does a world-wide lookup and returns with the Internet address of the system that you are trying to talk to, and this information comes from a DNS server located at the remote site. Almost every organization maintains one or more DNS servers containing the Internet addresses of the machines on its network so that users world-wide can discover them and communicate.
Such a huge and powerful infrastructure in place, and all it stores are Internet addresses (and then some)!. The search percolates down a world-wide hierarchy and ultimately executes on a server at the remote site. What if this percolation and search can return other information? Yes, the DNS model can be extended so that an organization can store any public information about its resources that may dynamically change. Think of one or more servers on your network which handles this information base. It may contain port addresses for protocols or services on your network, the workload on a machine on your network, public devices on your network like printers and fax machines, whether a particular user (say the MIS Manager) is on his PC , even where he is logged in now, where to reach secretarial services to leave a message for a person… all kinds of dynamic information that can be useful to programmers as well as users of your organization. Also think of the possibility that you can maintain separate servers for public as well as private information. Now think of a world-wide setup like this where each corporation manages its information that is amenable to programmatic querying. It would be possible, for example, to locate a JetSend enabled printer that is currently up and running at Calsoft.co.in in California and send a document from a scanner operating in India. Programmatically or by user intervention.
Such a model can also extend the power of the web in searching for information. Instead of searching at well-known search engines which hold huge databases that are reaching critical dimensions, searches can spread across the world to reach individual organizations - a massively parallel search. For this the percolation mechanism must support querying that spreads out on various limbs of the Domain hierarchy in some ways similar to a multicast. Think of trying to locate your friend who is visiting many IITs in India. It would be easy to launch a query targeted at all IIT domains in India to look for your friend. The searches will be carried out in parallel by the servers there (i.e. if they allow it).
You are not just federating devices here. This is collective and massive collaborative computing - a world-wide intelligent network. The Web today is mostly a static information base that changes when system administrators change some information at their web sites. We are talking about making the Web really dynamic! Such a powerful distributed co-ordination model and a search mechanism has been within our reach for so long. Who will take the initiative to make it happen?