If I wear my 'super retentive' had, I would argue that the SEEK() function as opposed to the SEEK command (action), [also] requires an extra cycle to package the return (T of F).I have got the constraint game down so well that my apps "know" the pointer is there (eg,they don't need to ask if it is there!) . At the top of my apps, of course, the constraints have to be enforced before the data is committed - so - in that case - The seek function is very helpful.

Did you know that SEEK command had the table and tag arguments? Even ZAP (now) has an 'IN' clause. Of course - it may have been that way for a while!:)


>Oops...hit send without typing reply :)
>
>There is no difference between the code I posted and using the SEEK command. SET EXACT is relavent when SEEK() is called, not when FOUND() is called.
>
>I agree that an EXACT flag for SEEK() would be nice. I don't like having to mess with SET's myself.
>
>>>>I agree - but with the function, I would have to "SET EXACT ON" twice!
>>>
>>>Why twice?
>>>
>>>

>>>=seek(cvar, "mydbf", "mytag")
>>>set exact on
>>>if found("mydbf")
>>>   ....
>>>endif
>>>

>>
>>Here's one reason why:
>>

set exact off
>>if found()
>>   set exact on
>>   .... do code that requires exact on
>>else
>>   set exact on && return EXAXT to 'system' state
>>endif

>>I have oftened hoped that MS would give us an 'EXACT' switch in the SEEK() function that would revert to the SET EXACT state after the function completed. They havent - but - and this is a surprise - is that MS has added a bunch of features to core xBase! My 'set-in-my-ways' apprach had condemned me to re-using old libraries - I bnever even thought to check VFP xBase for enhancements - until last night. I was warm and fuzzy all over!