Re: Where can I get a complete list of all SQL Hints?

"Noons" <wizofoz2k_at_yahoo.com.au.nospam> wrote in message news:<3ef39be1$0$16256$afc38c87_at_news.optusnet.com.au>...
> "TM" <tonym101_at_hotmail.com> wrote in message
> news:7b718442.0306201427.1b798ddb_at_posting.google.com...
>
> > FOR READ is not a hint, but the ANSI SQL standard for defining a
> > cursor to be read-only. FOR FETCH is just a synonym that does the
> > same thing.
>
> Mind explaining why it is included in the section about tuning
> in the DB2 doco? Shouldn't it be in the standard SQL section?

It *is* in the standard SQL section. And it is standard ANSI SQL, not non-existent DB2 UDB hint syntax.
>
> > FOR UPDATE is not a hint, but the ANSI standard for defining a cursor
> > to be updateable.
> >
>
> Same.

Ditto.

> > > FETCH FIRST n ROWS ONLY
> >
> > OPTIMIZE FOR n ROWS is a soft limit that tells DB2 that you are likely
> > to only want the first n rows of the result set, but does not prevent
> > you fetching further rows if you change your mind. FETCH FIRST n ROWS
> > ONLY is a hard limit that is DB2 direct syntactical equivalent of the
> > ANSI SQL LIMIT TO n ROWS clause. It is not a hint and indeed alters
> > the result set.
>
> So why is it included in the doco in the section about
> tuning, with an EXPLICIT recommendation stating
> it changes the execution plan?

DB2 documentation is pretty extensive and I can find no such reference. I suspect you are misquoting as usual, or maybe quoting for a different product rather than DB2 UDB For Multiplatforms, but who knows... maybe someone did write that. If you expect me to take this claim seriously, post a link to the relevant HTML page of the DB2 UDB For Multiplatforms "doco".

> > Jeez... well quite. You pretend to know something about databases, so
>
> Quite wrong. Unlike you, I NEVER pretended to "know something"
> about databases. That is your stupid claim, not mine.

So why do you waste others' time posting ill-informed opinions here as though they were fact, stated without qualification or indication of possible uncertainty?

> > it's amazing that you seem to be unaware of the existence and
>
> Oh, I am the one quoting them and I am unaware of what they do?

Apparently. Or pretending to be, so you can misrepresent them as "DB2 hints".

> Hellooooooooo? Anybody home?

Stop trying to change the argument. Anyone who can be bothered to trace this thread back will see that you claimed FOR READ ONLY etc. were DB2 hints, whereas in fact they are DB2 support for standard ANSI SQL.
> Make sure you pass that on to your stupid organization so they stop the
> regular interventions here.

I don't work for IBM, but do Oracle pay you to come up with the smears and lies you post here?

> Oh yes: have you ever seen me interject in comp.databases.db2
> or other crap product ngs? Never. Yet I keep seeing you dickheads
> regularly trolling around here. Enough said about your true motivation.

Perhaps there's a conspiracy, or perhaps the DB2 forums are rather saner places where you get more balanced and professional opinions, not smears and lies about competitor products.

> > usefulness of LIMIT clauses, and don't recognize standard ANSI clauses
> > FOR READ ONLY and FOR FETCH ONLY but instead somehow imagine these are
> > DB2 hints! If you really work in the industry then I suggest you try
> > to keep your ignorance a little less high profile.
>
> The "ANSI clauses" are in the section of UDB dealing with
> tuning, TOGETHER with a clause that is UNIQUE to DB2 and
> locks people in to their code and way of doing things.

Smears and lies. FOR READ ONLY and FOR FETCH ONLY are in the ANSI specification.

> The Oracle hints are standard ANSI comments and inefective
> in ANY SQL processor that follows the most basic ANSI compatibility.
> Therefore, they are COMPLETELY portable and do NOT lock ANYONE
> into their syntax.

Whatever. Oracle hints are obviously the best thing since sliced bread.

> Can you spot the difference or does it need to hit you
> square on the head?

I can spot that there's little point wasting further time debating with a total zealot.

TM Received on Sat Jun 21 2003 - 07:35:29 CDT