Re: Re: 7.2 items - Mailing list pgsql-hackers

From Peter Eisentraut
Subject Re: Re: 7.2 items
Date
Msg-id Pine.LNX.4.30.0106272051470.729-100000@peter.localdomain
Whole thread Raw
In response to Re: Re: 7.2 items  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
Tom Lane writes:

> Peter Eisentraut <peter_e@gmx.net> writes:
> > Tom Lane writes:
> >> What for/why bother?  A toastable bytea column would do just as well.
>
> > There's still a 1 or 2 GB limit for data stored in that.
>
> 1 Gb, I believe ... but LOs are not a lot better; they'd max out at 2 or
> at most 4 Gb, depending on whether the code always treats offsets as
> unsigned.

That can be fixed by adding a 64-bit aware equivalent of the existing lo_*
functions.  I suppose it'd be a lot harder to make regular data types
handle long values.

-- 
Peter Eisentraut   peter_e@gmx.net   http://funkturm.homeip.net/~peter



pgsql-hackers by date:

Previous
From: Jan Wieck
Date:
Subject: Re: functions returning records
Next
From: Hannu Krosing
Date:
Subject: Re: Re: 7.2 items