From: | Jonathan Bartlett <johnnyb(at)eskimo(dot)com> |
---|---|
To: | CN <cnliou9(at)fastmail(dot)fm> |
Cc: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: BLOB or BYTEA field |
Date: | 2003-02-19 22:27:04 |
Message-ID: | Pine.GSU.4.44.0302191423130.20576-100000@eskimo.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
BLOBs are actually created using lo_create - you are then returned an OID
that can essentially be used as a file descriptor. However, in my
programs, I've opted for bytea for the following reasons:
* You can't query for BLOBs and know if there is a dangling reference
* BLOBs use OID fields. Running out of OIDs is a problem in PostgreSQL
because they are only four bytes wide. This makes BLOBs useless for large
databases.
* bytea fields are easier to manipulate
I think I had some other reaons, but I can't think of them right now.
I also just liked storing the binary object itself as an attribute rather
than a somewhat dangling entity.
Jon
On 18 Feb 2003, CN wrote:
> Hi! Daniel,
>
> >
> > you should use oid Fields for BLOBS. That works well.
> >
>
> I have the same question. Could you make it more specific? oid is an
> unsigned 4 bytes integer. How do you save binary data from Delphi in
> bytea column "using oid fields"?
>
> Thanks you!
>
> CN
>
> ---------------------------(end of broadcast)---------------------------
> TIP 1: subscribe and unsubscribe commands go to majordomo(at)postgresql(dot)org
>
From | Date | Subject | |
---|---|---|---|
Next Message | Martijn van Oosterhout | 2003-02-19 22:41:16 | Re: 7.3.1 takes long time to vacuum table? |
Previous Message | Jean-Luc Lachance | 2003-02-19 22:14:25 | Re: Table Partitioning in Postgres: |