Re: [SQL] Serial field starting at 100 ?

Thanks for your reply,
however the syntax

>create sequence fish_seq start 100;
>
>create table fish(
>     fishID int default nextval('fDish_seq') not null,
>     fishName varchar(32) not null
>);

is exactly the same as this code that I'm currently using
in Postgres:

> > CREATE TABLE fish (
> >     fishID INT4 SERIAL
> >     fishName VARCHAR(32) NOT NULL
> > );
> >
> > SELECT setval('fish_fishID_seq',99);

except that the two sections are switched.
I would have thought that it would have been possible to set the
start value within the "CREATE TABLE" so I don't have to stuff
around in some separate clause using a sequence name that will
never be explicitly used anywhere else in the code.

Perhaps this is something that is intended for an upcoming version
of Postgres.

>On Fri, 7 Jan 2000, Richard Walker wrote:
> >
> > In SQLServer, you can do this:
> >
> > CREATE TABLE fish (
> >     fishID INTEGER IDENTITY (100, 1) NOT NULL,
> >     fishName VARCHAR(32) NOT NULL
> > );
> >
> > When inserting into that table, you don't mention the first field
> > and it will automatically use add 1 to the last value in that field,
> > starting with 100 for the first record.
> > If you wanted it to start at 25 and increment by 5, you'd put
> > fishID INTEGER IDENTITY (25, 5) NOT NULL.
> >
> > I'd like to do something similar to this in PostgreSQL, but the
> > closest I can get requires that I mention the implicit sequence which
> > seems unneccessary hassle:
> >
> > CREATE TABLE fish (
> >     fishID INT4 SERIAL
> >     fishName VARCHAR(32) NOT NULL
> > );
> >
> > SELECT setval('fish_fishID_seq',99);
> >
> > (Incrementing by 1 is fine by me).
> > Is there any way I can do something like:
> > fishID INT4 SERIAL (100, 1)
> > or
> > fishID INT4 SERIAL (START 100)   ?
> >
> > - Rick

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com