that's more than the guys at Enron and MCI Worldcom can say

They're preparing to say (on the stand): "Unfortunately I don't know any accounting, so this is very confusing...

CREATE TABLE #Temp(

Col1 int,

Col2 int)

INSERT #Temp

EXEC p_Test 'YourParam'

Put all the DDL at the beginning of your calling proc to prevent excessive recompiles.

Unfortunately I know some accounting, so this is very confusing to me. 

Please post the DDL and some sample data with an example of...

A way to do exactly what? 

There are lots of ways to do this. 

• You could have your weekend table populated with a row for each weekend day and then just subtract your start date from...

Yes.  That's typically how one deals with a set returned from a called proc.

UPDATE SCH_SET

    SET fac_send_date = DateAdd(n, fdate_cnt*24*60 + fhour*60 + fmin, f.book_date)

    FROM SCH_SET t, FAC_BOOK f

    WHERE upper(set_id) = 'A000001' and upper(pcode) = 'A'

If that's what you want...

SELECT ISNULL(i.FEmpID,o.FEmpID) FEmpID, i.FTime FTimeIn, o.FTime FTimeOut, CONVERT(char(5),o.FTime - i.FTime,14) Hours

FROM Punches i FULL JOIN Punches o ON i.FEmpID = o.FEmpID AND o.FType = 'OUT' AND...

It now sounds as though you should not be using a method incorporating the identity property.  Alter the stored procedure so it optionally takes a prefix and/or number.

CREATE...

If the first character is input but the numeric portion is automatically increasing, then you could use bbron's suggestion of a table holding the last number used.

CREATE TABLE SeqNos(

Type char(4)...

About the only reason to do this with dynamic SQL is if those responsible for the front-end code are not conversant with SQL.  As that is where your clauses are...

If you have any indexes on the ID columns, this will probably be fastest:

DELETE FROM StagingIn

WHERE NOT EXISTS

(SELECT *

 FROM Plans

 WHERE ID1 = StagingIn.ID1

  AND ID2 = StagingIn.ID2

  AND ID3 =...

No, as I wrote, you only need one local variable:

declare @DidItError int

insert statement ....

set @DidItError = @@error

update....

set @DidItError = @DidItError + @@error

if @DidItError > 0 then

rollback

else

commit

> Maybe I have to save the @@error to a var after every statement, then at the end evaluate that var to see if there was errors?

Yes, the @@error value...

CREATE TABLE test(

IntID int identity(0,1) CHECK (IntID < 260000),

StringID AS CHAR(IntID/10000 + 65) + RIGHT('0000' + CAST(IntID AS varchar(10)),4))

You could also do this with a view instead of a computed...