On 18 Oct 2012 09:39, Dan Kimmel wrote:
I think Vern is trying to differentiate between a callback and an
exit. There's really not much point in differentiating, but I'll
offer this: Callbacks are usually done by passing a procedure
pointer. Exits are usually done by calling a program by name. In
the case of the callback the calling program has already loaded
the function into memory. An exit is loaded and executed by the
operating system at the behest of the calling program.


FWiW the term "exit", as I was familiar, was used generally [by OS and LPP development] to explain any scenario whereby some [typically non-system] code was enabled to be called by/rom within, some system-code; irrespective of the means to effect it. That is, the processing by the system made an "exit" from its normal processing, in order to call some other code. Whether or not the means to effect the 'exit' was implemented as an actual or effective 'callback', was immaterial. And FWiW, the common implementation of the registrations [WRKREGINF] of some typical types of exits was done both to better ensure system integrity and the consistency with those similar forms of system code invoking what are called 'exit programs' at what are called 'exit points'.





Return to Archive home page | Return to MIDRANGE.COM home page