Denis,

Glad you could understand it okay.

The problem lies within the Intersect Buckets table. You'll need a DELETE statement to remove bucketID 2 (as has been SENT, similar DELETE method...

Just thinking on, if the CHANGED result (difference of CHANGES) is to be the difference and not the total of the weights after the changes have been made:

Remove the Changed...

I've got a script which I think answers the question. I haven't fully tested it; but it's all set-based so the performance on a large table shouldn't be too bad.

The...

There is a bug in the above script... It does not return correct results when a bucket has been reused..... Change the dates...

I couldn't quite figure out if you wanted to include buckets that have been brought IN or CHANGEd before the start date but not sent OUT before the report (therefore, like...

Are you using the Copy SQL Server Objects task?

Hmm, not seen that one before...

Drop me an email to {email now removed} and I can send you an example package that I've created.

Just out of...

Steve,

You'll need the Execute SQL task as the first step in your package, this will build the filename and pass it to a Global Variable.

You then need to follow...

Steve,

Thinking back I remember tearing my hair out over a similar problem and decided that I could actually hate an inanimate object....

Anyway, back to the problem. If the path is...

Apologies, my misunderstanding...

You need an Execute SQL task with the statement:

SELECT CONVERT(CHAR(8), DATEADD(dd, -1, GETDATE()), 112) + '.mdb' AS filename

in it to build...

Steve,

You can use DTS to build a package that will transfer data from a MDB (through the Jet engine) to a SQL Server table. Have a look at http://www.sqldts.com...

Stuart,

The following script opens a GV rowset and builds a comma-separated string for each row. You could incorporate the FileSystemObject in to the record loop and write each line to a CSV...

Can you post the statement that you are attempting to use too? This may help us in helping you...

Check out the performance difference of creating the temporary table first and the INSERT INTO it...

If the error statement is exactly as above it looks like the extract is trying to put a VARCHAR value ('XXXXXXXX') to a NUMERIC column (column 8).