>>>It's funny, though, that it occasionally gave "illegal operation", but then did read all the memos in the scan-endscan loop. Maybe the weak spot was on a deleted record's memo?
>>
>>That's a good point and I'd be interested to know. Perhaps George can tell us? (I'll copy this to him)
>
>The weak spot we believe now was created by a network error. The problem was indicated by Memo fields indictating they had data but when we looked inside the memo fields there was no data. We then noticed that the problems were only after a certain time and date. There were approximately 200 records out of say 270000 records. What I thought might have happened was that the memos we're pointing to addresses that were no longer valid in memory causing the corruption. If this were true we should be able to replace all these 200 records memo fields with "" and that would hopefully fix the problem. We did the replacement and that seemed to fix the problems.
>
>Dan

Memo fields will sometimes read a space or return in the memo as not empty. We got around it by changing the search to alltrim().
Brad