[imp] oracle_create script

[Hanser, Kevin]

Ooops... yea, trailing ; would be good.

As far as it working w/oracle versions other than 8i? 
No idea... I am by no means an oracle expert.  In fact, quite the opposite.
It just happens that we have an oracle database running at our company, and
so that's what I used :)  I'm learning as I go...

As far as the "NOT NULL" goes... it's probably a good idea too, as others
have also pointed out.

Now the question I have is this: I created my database with this script,
should I add those constraints on now?  Or is it something I really don't
have to worry about since IMP/Horde is working fine anyway...?

thanx

Kevin

-----Original Message-----
From: Brent J. Nordquist [mailto:bjn@horde.org]
Sent: Thursday, October 12, 2000 3:51 PM
To: 'imp@lists.horde.org'
Subject: RE: [imp] oracle_create script


On Thu, 12 Oct 2000, Hanser, Kevin <kevin@ShopsForMe.com> wrote:

> Ah, then it's probably a good idea to have them :)
> 
> here's the updated script:

OK, some questions for the Oracle users before I commit... recognizing
that it's been a loo-o-ong time since I used Oracle :-)

(1)  If I commit this as oracle_create.sql, are users of Oracle before 8i
going to have problems?  Should I commit this as oracle8i_create.sql, or
will this work fine for both?

(2)  Compared to the old oracle_create.sql, this script removes PROMPT,
SPOOL, and "/" as terminators -- I'll defer to your judgment on those.  
However, this one looks suspicious because there's no ";" terminator
either:

> create index CHANGED on ACTIVE_SESSIONS (CHANGED asc)

(3)  I think the following three fields need "NOT NULL" attributes, based
on the old oracle_create.sql script, as well as the others (like MySQL):

> CREATE TABLE imp_addr (
>     username varchar2(255),
>     address varchar2(255),
> [...]
> CREATE TABLE imp_pref (
>     username varchar2(255),

Let me know your thoughts; I'm ready to commit as soon as we have
consensus.