SQL commands execute on the SQL Server, so batch Inserts should not have any network traffic issues.

OK, a couple of questions:

1) Are you using an exact solution? (sounds like you are)

2) Do you just need a yes/no answer (possible, not possible) or do you need a...

Shawn Therrien (5/8/2009)

---

Subset Sum is a pattern we're using on a few procedures and we're doing it with cursors. None of us can think of how you would solve this...

It's not doubling the "ADD"'s anymore is it?

Technically, all permissions are set ON objects, and TO principals (users, roles, etc.).

The SQL for this would be something like this:

DENY INSERT ON OBJECT::[schema].[YourTable] TO Public

Assuming I have that correct,...

Why do you say that? I've done reports...

There are two basic ways to do this, both through a trigger:

1) Use the Update() function to determine if any of the column(s) was SET by the update (and optionally...

Yes, but normally you would do this through Permissions setings on the table: I.E:, just Deny INSERT access to the table.

"Sargable" effectively means "can be used to search via an index". The general rule being that a column wrapped in a function is not sargable (there are some exceptions...

Hmmm, actually. now I am not sure that my compile-time/run-time binding explanation was correct. However, I still do think that dynamic SQL would work.

Thanks for the correction, Gail. Hmm, does make me think that I might have gotten a couple of other things wrong today too... 🙁

The problem is that for the most part, objects get bound at compile-time, not run-time. True #temp tables get some kind of workaround for that, but a table like...

Frequently an APPLY can be as fast as a JOIN, however, there are many times when it will not be. It is very rare that an APPLY is ever...

GilaMonster (5/8/2009)

---

RBarryYoung (5/7/2009)

---

As for how he keeps his job, there are two obvious possibilities: 1) Civil Service, and 2) Nepotism. These two work pretty much the same, to one degree...

Actually, I'm more in the ambivalent camp myself. I doubt that I use either one enough in stored procedures for the performance difference to ever really matter, and if...