| From: | Bruce Momjian <maillist(at)candle(dot)pha(dot)pa(dot)us> |
|---|---|
| To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
| Cc: | Leon <leon(at)udmnet(dot)ru>, pgsql-hackers(at)postgreSQL(dot)org |
| Subject: | Re: [HACKERS] Fwd: Joins and links |
| Date: | 1999-07-05 19:08:17 |
| Message-ID: | 199907051908.PAA07909@candle.pha.pa.us |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
> Leon <leon(at)udmnet(dot)ru> writes:
> > We should make a real reference in one table to another! That
> > means there could be special data type called, say, "link",
> > which is a physical record number in the foreign table.
>
> There is no such thing as a physical record number for a tuple in
> Postgres. The closest you could come is an OID, which isn't really any
> faster than any other joinable field --- you still need an index to
> support fast lookup by OID.
Actually, there is:
select ctid from pg_class;
ctid
------
(0,1)
(0,2)
...
The number is the block number offset in the block. It doesn't help
because UPDATED rows would get a new tid. Tid's can be used for short
periods if you are sure the data in the table doesn't change, and there
is a TODO item to allow ctid reference in the WHERE clause.
--
Bruce Momjian | http://www.op.net/~candle
maillist(at)candle(dot)pha(dot)pa(dot)us | (610) 853-3000
+ If your life is a hard drive, | 830 Blythe Avenue
+ Christ can be your backup. | Drexel Hill, Pennsylvania 19026
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Bruce Momjian | 1999-07-05 19:09:19 | Re: Re[2]: [HACKERS] Fwd: Joins and links |
| Previous Message | Tom Lane | 1999-07-05 19:02:25 | Re: Re[2]: [HACKERS] Fwd: Joins and links |