Hi Ed. This is a great explanation. You should consider uploading it as a faq.

>>When is my vfp-application a client/server-application?
>>
>
>There are many different client-server realtionships; in an environment where some of the services such as file access, object manipulation, or database manipulation is performed on some system other than the local PC, you are operating to some extent in a client-server environment.
>
>The simplest and most common client-server relationship, and the one that is least typical of what is meant by client-servr, si where some or all of the data files reside on another system, and VFP accesses those files across the network. The server is typically a file server; it may be a dedicated machine running under a network operating system like NT or NetWare, or might simply be another workstation whose files are made available across the network. The local workstation does all the work of data manipulation; it requests file services across the network interface. The local station sends low-level file requests and updated data across the network, and the file server retrieves or stores the data. The file server doesn't understand what the files it manipulates represent, it just provides low-level file functionality.
>
>Most people, when they refer to client-server application environments, mean that more of the workload than raw file or printer access is performed across the network. Typically, a database server handles the underlying file manipulation instead; the local station prepares SQL commands and pases the command across the network to the database server; the database server does the actual work of performing the query, processing the files, and passes back result sets and trransaction acknowledgements to the workstation. In this type of environment, the local station doesn't do the direct data file manipulation; the SQL statement passes over the wire, and the results come back. In most cases, the net result is that there's less information passing over the wire, at the cost of having a central server responsible for all file manipulations and processing the requested service for all stations.
>
>The advantage to this scenario is obvious if you have large files that would otherwise have to cross the wire in order for the local workstation to process the data. In many cases, the database server can provide enhanced capabilities not found in VFP, for example, enforcement of record and field-level security, automatic tranaction journalling (keeping a record of changes tot he database so that a set of changes can be backed out as a whole, and the sequence of processing is clearly documented), scalability and additional data integrity checking and fault tolerance are some of the compelling reasons to use database servers. Database servers are more in line with data-oriented client-server discussions, where the primary issues are network overhead or file system capabilities.
>
>With the adoption of COM, the Component Object Model, client-server has expanded to include n-tier applications. In an n-tier systtem, some or all of the business logic is moved into a COM server or servers, which provide a standardized interface for performing object-modeled business operations on the database. The mid-tier business objects may run either locally or remotely, and may in and of themselves talk to other business objects, or a database server. The n-tier model allows the encapsulation of business logic in addition to data services, in a way that permits many different application environments to reference those services using the common interface. The n-tier client-server relationship allows you to move not only data but logical services to another local process, or a remote system, and allows for a wide range of scaling options, and further insulates the front-end from changes in business rules or physical and logical data designs.
>
>The COM model is probably the most flexible and versatile of the client-server models in wide use today, and is typical of what many end-users imagine when talking about client-server applications.
>
>
>>VFP-help says when a local (client) user interface accesses data on a remote server. What do they mean with 'remote server'? Is it the physical machine or the software (e.g. SQL-server) which is running on that machine?
>
>The server idea is a software concept - the server provides common mechanisms and services; whether it runs on the same physical machine or a different machine is not critical to the concept. When speaking of a remote server, it's specifically referencing a process running on a physically distinct and separate system, and usually is referencing the COM model of processing, specifcally the use of DCOM, the distributed implementation of COM.