Re: [HACKERS] please? - Mailing list pgsql-hackers

From Theo Kramer
Subject Re: [HACKERS] please?
Date
Msg-id 3752EA9A.C719EDFE@flame.co.za
Whole thread Raw
In response to Re: [HACKERS] please?  (Bruce Momjian <maillist@candle.pha.pa.us>)
Responses Re: [HACKERS] please?
List pgsql-hackers
Bruce Momjian wrote:
> 
> > > > > Oracle OCI has oopt() and Informix Online has dirty read that do the trick for
> > > > > me.
> > > > > --------
> > > >
> > > > Please give me more information.  How does dirty read fix the problem?
> > >
> > > It allows me to read uncommited records without blocking.
> >
> > I suppose it somehow lets you know whether the read was dirty or
> > clean...
> 
> Not that I am aware of.  Never heard that of Informix.

I also cheat. I use a 3 buffer approach, compare fields and see if a record
has
changed before I do an update.
--------
Regards
Theo


pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: s_lock_stuck (was Problems w/ LO)
Next
From: Tom Lane
Date:
Subject: Re: [HACKERS] Backend sent 0x45 type while idle