| From: | Rod Taylor <rbt(at)zort(dot)ca> |
|---|---|
| To: | Michael Meskes <meskes(at)postgresql(dot)org> |
| Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, PostgreSQL Hacker <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Strange bahaviour |
| Date: | 2002-08-12 14:24:24 |
| Message-ID: | 1029162265.42569.36.camel@jester |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Just go with tradition and label all of your variables with a
v_<varname>. Never use columns or tablenames prefixed with a v_.
It's a quick way for determining what is what. Forcing use of a prefix
in some places and not others would not be a nice thing -- especially as
the core takes on more and more abilities of plpgsql.
On Mon, 2002-08-12 at 10:00, Michael Meskes wrote:
> On Mon, Aug 12, 2002 at 09:40:12AM -0400, Tom Lane wrote:
> > The NULL is perfectly correct: that's the initial value of the plpgsql
> > variable. The above is essentially the same as saying
> > login := login;
> > It is not "incorrect".
>
> That's exactly what I thought is the reason. I just wonder if there's a
> way to make this kind of stuff more obvious for instance by using :login
> for the variable as with embedded SQL. IMO the actual behaviour, while
> of course correct, is a little bit confusing.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Don Baccus | 2002-08-12 14:26:44 | Re: [SECURITY] DoS attack on backend possible (was: Re: |
| Previous Message | Dave Page | 2002-08-12 14:23:53 | Re: [SECURITY] DoS attack on backend possible (was: Re: |