Well, my message was an answer and not a question... :) But thanks for your answer anyway.

When I said that these conventions are not explained in the doc I was reffering to the Win32API help file that was shipped with the VFP 3.0 (since it seems this is the main Win32 API info resource for many VFP programmers).

Vlad

>Vlad,
>
>LP = long pointer which is moot under Win32 because all pointers are 32bit long pointers. In Win16 days you could have pointers which were just 16 bit pointers into into the code's single DS (data segment). The Infoviewer in VC++ can point you at these "base" kind of declarations.
>
>>Ok, I got it now! :) The "*" indicates a pointer, but if the type is LP..., then it is probably a macro for a pointer type (ex: LPWORD is in fact a replacement for WORD*). So, all LP... types are pointer types, in fact. Also, all P... are pointer types.
>>So, the convention was followed all the time, only that LP... types hide the "*". (But the "*" is always there, because it's part of the C language).
>>
>>Special care must be given to types like LP...*. These are pointer to pointer types. Ex: LPWORD* translates to WORD**, pointer to pointer to WORD.
>>
>>I wonder why these conventions are not explained clearly in the doc... :(
>>
>>Vlad
>>
>>>Sorry if I was a little vague. I mention the convention of using the pointer symbol ("*") to indicate that a pointer (reference) was used. They just used "lp". Then again, I really shouldn't have called it a convention, not being too familar with C++ .