definitely not normal. foreign key relationships are a best practice.

I've encountered this before, and it seems the companies think that their schema would reveal their business processes ro something, so...

here's a second table with an instead of trigger:

CREATE TABLE EXAMPLE2(KEYID BIGINT REFERENCES MASTERKEYS(KEYID),

...

i got curious on this and built a prototype.

man what a pain to have to code a zillion triggers, just to have a master id someplace.

i could

I ended...

to actually implement this, I think each trigger would have to insert into a mastertable with a bigint identity() column, and use an OUTPUT clause for the collection of values,...

just to clarify the whole Oracle thing, Oracle doesn't have a native identity() function like SQL does; instead, you use a SEQUENCE and a TRIGGER to duplicate the functionality, and...

yeah pseudo code is bad. your placeholder proc implied something completely different.

it is absolutely not possible to call a procedure with an update statement.

you can call a function, which returns...

ok, here's some working code, witht eh update you seem to need; but here's the rub:

this is what it does:

for each matching record, if today happens to be sunday,...

ok great that helps a lot...i can do the table and the joins.

just need one more piece of information.

what do you want to do to RESULTDATE? i thought you either...

the problem is the @in_name.

since you are executing a string you put together, it is possible that instead of the expected customer name being 'dba-vb'

i could pass ''';DELETE FROM CUSTOMERS;DELETE...

if you have two logical steps, you'll need to update statements, each witha WHERE statement to control which rows get updated.

what does the stored procedure do?

your pseudo code doesn't have...

with a CASE statement, you can't. case statements return data or help filter a WHERE statement.

a case helps goes something like this:

SELECT

CASE

WHEN status=1

...

tfader is there any common themes applied to the parent-child fields, even when no FK is created?

for example, if the following are all true, i have a script i call...

same issue that i saw in the other thread i think;

the mix of old vs new style joins is causing a cross join and increasing the rows processed:

select distinct a.empnumber,...

...from Vwrevenue as a inner join #revdates as b on a.empnumber = b.empnumber, #daterange

i think because you inner join one table, and then use a comma to join the old...

Thanks Wayne;

I think the minor tweak Lynn suggested is better; the change you made gives some inaccurate/unexpected results, even though the code runs without error:

i think it's because i was...