Sergey - Thanks I was beginning to feel abandoned. I did some futher testing on their network myself. Sometimes all the records are appended and the next time I run the program only a portion of the records are append.

I have a control total for the file so I know if the all the records are download. I use a global error handler, but it never errors out, just the wrong number of records are appended. If the total don't agree I have them run the routine again until they get the control total.

At time the downloads is extremely fast and sometime very slow. Speed doesn't seem to be the problem as to getting the full download. It sometime fails to get all the record when it is fast and sometime when it is slow.


I am thinking I will just copy the file from the network to the workstation's harddrive and then append from the workstation and perhaps eliminate possiblities of a network problem in the append.

I used to have a program that would append several millions records for a sdf file off a network every few weeks and never had a problem. This is the first time I've used a tab delimited and wasn't sure if converting tab delimited files were problematic.

Thanks
Gaylen

>Hi Gaylen,
>
>A couple of questions. Is it possible that user picks the wrong file the first time around? By any chance do you have an error handler in place that could hide error in the APPEND command?
>As a workaround, you can create another cursor, do second append, compare record count and act accordingly.
>
>
>>I have an append program with the following code:
>>
>>

>>lcdirectory = Sys(5) + Sys(2003)
>>SET DEFAULT TO "H:\commission download"
>>lcFileName = Getfile("txt","Get USIS Download")
>>
>>SET DEFAULT TO  (lcdirectory)
>>Create Cursor usisxls (field1 N(9),field2 c(10))
>>Select usisxls
>>Append From (lcFileName)Delimited With Tab
>>
>>

>>
>>Ocassionally the number of records that are appended is only a portion of records that should have been appended. If I run it a second time the complete file is appended. The download usually consist of approx. 3,000 to 20,000 records, and size is not a factor as it fails on randonly for both file sizes.
>>
>>Any ideas as to what might be happening.
>>Gaylen