Vime, I think what you are saying is that you need both a one time update - and a trigger for future updates...(I may be misundertanding).

See Martin's examples for the...

Hello Phil, I'm curious about your double from.  You mentioned that it is not ANSI (89/92 - or not at all?)...  Is there an ANSI statement that deletes based on...

No, not necessarily.  NVarchar's and Varchar, as just one example, can frequently vary in size within a column.  Therefore, updates to a column can cause page splits, etc. that will...

I took out your variable for my testing so you will need to add it back in...  Anyway, I think this does what you need - eventhough it is extremely...

Bruce, eventhough you have 4GB memory, standard is going to use no more than 2GB.  It probably just better to let sql dynamically determine memory utilization.

Regarding virtual memory - as...

JN, this brings us back to the distinct idea.  You just stated that "all other fields have same values repeated including key fields".  This means that you do not have...

As far as I know, there is no need to wait.  You can still PIVOT in SQL 2000 - just takes more typing.

See Pivot Table in BOL. RH

Check out PW's thread.  His derived table join still applies - only I guess it won't matter what field you aggregate. RH

It might be easier to leave col2 as NULL until the APP updates the column.  In fact, that would tell you even more about col2 since the app may set...

JN, what logic are you using to determine what the first occurrence is?  Is it arbitrary?

Also, are you saying that your select statement that feeds the INSERT statement has identical RECORDS,...

I would consider:

Class Table, Student Table, Attendance table.

The attendance table would house Class_FK, Student_FK, Datetime...or something like that.  You can always change the display of data though a report.

Your attendance...

8 bits = 1 byte (i.e. 10101010 = 8 bits or 1 byte)

16 bits = 2 bytes, 24 bits = 3 bytes, etc.

Disregard - I got it worked out.  The following routine helped illustrated it for me.

CREATE Table tblEXISTS (col1 int, col2 int)

INSERT tblEXISTS VALUES (50,10)

INSERT tblEXISTS VALUES (50,11)

INSERT tblEXISTS VALUES (45,12)

INSERT...

OK - another clarification.  I hate to ask open ended questions of people so I have rewritten the query using "IN".  This query is an example of how I am...

I'm not sure my post above properly explained my issue.  I rewrote the EXISTS as I understand it:

SELECT t3.* FROM

test t3, test t2, test t1

WHERE t3.location != 50...