I don't think I was being rude or impolite. It is impolite to suggest someone who is trying to help is being impolite. You should not assume from email what is impolite. I only advised you to spend time clarifying the problem and not guessing. I provided code that demonstrated that a replace can be effected within a UDF called from another replace contrary to your guess. Had you submitted a small piece of code like Sergey did which clearly demonstrated the initial problem, there would have been 2 or 3 entries in this thread. Further, you would have been helped faster and not had to put up with my being impolite.

http://fox.wikis.com/wc.dll?Wiki~SignalToNoiseRatio

>>OK. So its a real reproducible bug. Who's going to submit it to MS? I've never attempted doing a replace in a UDF that is called from a replace command. I'd redesign this code.
>
>I already changed the code, but I spent ~4h. on Friday debugging it and some time during this weekend not to mention not very polite replies from you and Jim (and partially it was my fault of not explaining the problem well, but you can understand my frustration at the time, the original message was written).
>
>I think, since Garrett already participated in the thread, I don't have to do anything else about it.
>
>I probably will create a topic in WiKi about this "discovery"
>
>Thanks again for help.
>
>>
>>>Thanks, Sergey. Notice, that gather memvar in UDF has the same effect, as replace in UDF, but insert from memvar works in UDF.
>>>
>>>So, what should we do with this info? :)
>>>
>>>>Hi Nadya,
>>>>
>>>>Hee's a code that reproduces your problem with min of code

CLOSE DATA
>>>>create cursor curT1 (RecNum I)
>>>>create table OutPut (Num1 N(2), Date1 D)
>>>>append blank in OutPut
>>>>replace Num1 with 100,  Date1 with Repl1()
>>>>RETURN
>>>>function Repl1
>>>>replace RecNum with 1 in curT1
>>>>RETURN Date()
>>>>

Comment out Replace command in the UDF and you'll get "Numeric Overflow" error.
>>>>
>>>>>>>Yes, and apparently a goodly number of the reported ones too! Since we can't
>>>>>>>KNOW, given the present circumstances, my recommendation is don't waste the >effort.
>>>>>>
>>>>>>Hey, cool, since I don't work for MS anymore, I don't have to let this post go without comment.
>>>>>>
>>>>>>But it's not worth it, so I won't bother.
>>>>>
>>>>>Garrett,
>>>>>
>>>>>Could you please comment about the nature of the problem? Can you replicate it, because it's weird. My tests didn't work and Mike's did...