>>>I wonder if a DTOC() would help?
>>>>
>>>
>>>Wait, that wouldn't work...you lose the minutes and seconds that way, don't you?
>>
>>End of day mind lapse -- see my reply to myself where I replaced that with TTOC(). That's more out of curiosity. That more than doubles the field size, I believe, and is more expensive in indexing.
>>
>>Jay
>
>hehe...all of this to get a primary key. Pretty stinkin funny!! BTW, in case you didn't get cc'd, Jonathan came up with the way to do it w/o duplication.

Saw that!

It's easy to think that the values we deal with are "reality" in some fashion. I remember an article about 10 years ago in the late, lamented Programmer's Journal which dissected the IEEE floating point standard. Up 'till that point, I'd just thought of floating point values as integers with decimal fractions. But, the values are not any where continuous. At the max and min of the ranges, the increments are millions, even billions apart. That was an eye-opening experience. From what little AL programming I have, I try to think of what Fox may be doing "under the hood" when working with various features. And, it can help.

Yeah, PK's are often a challenge to come up with!

In any case, we've managed to localize where the problem probably occurs. That means it's probably not in other places (SCAN, etc.) and that's a good thing %).

Jay