Mike,
Even for one variable you could get great differences based on version used.
I checked viki reference and differences are almost ignorable. Was it VFP6 used in tests ? In VFP7 there is a much wider gap.
Check these threads when you have time (if you want of course:):
Thread #646212 , Thread #647967 and one more would be on those days.
Cetin

>For one variable that's true, but consider how many variables are in your system(s) and how many fields are in your tables and how many loops ...
>
>>Hi Mike,
>>
>>I guess you're right. I never really stopped to consider that, but it makes sense (although I wouldn't want to live on the difference). ;)
>>
>>Alan
>>
>>>Hey Alan!
>>>
>>>Your code will also be faster! The more fields in the current alias, the longer it takes VFP to figure out "ok" is not a field, before returning the memvar content.
>>>
>>>Check out http://fox.wikis.com/wc.dll?Wiki~EssentialMDot~VFP
>>>
>>>>I used to think that too, but as Jim pointed out, that's what the 'm.' is for. Once you get into the habit of explicitly pointing to what you want - ie - always using alias.field and m.variable, not only will your code become more readable, but Fox will no longer have to guess what you mean, and it will begin to feel more natural.
>>>>
>>>>Like I said, I had the same argument you do once upon a time, but I've since changed my mind.
>>>>
>>>>Alan
>>>>
>>>>>>Hi Srdjan,
>>>>>>
>>>>>> Because VFP read as first field OK from cursor test before variable.
>>>>>>
>>>>>
>>>>>Hi Martin :)
>>>>>
>>>>>I understood what is happening - What I am saying is that this situation shld be handled somewhat differently by VFP.
>>>>>
>>>>>That is all.