Flow services implementation

I’m a newbi in Integration Server and flow services, at first glance I wonder: If the flow services have input and output parameters, why don’t they run in a context (or variable scope) of their own, instead of getting the hole pipeline as an input?

When I m doing a debug on a flow service, it gets dirty with all the variables from the other services invoked, and worst than that is when the variables have the same name…

Why not implement a simple variable scope like the programming languages???

Am I missing something here?

Indeed webMethods by default uses the complete pipeline. However, if you use the ‘scope’ variable in the properties tab for the invoke, you can enter the name of a record which will be passed to the invoked service as its pipeline.