[Libpqxx-general] Unnamed prepared statements

[Trigve Siver]

> From: Jeroen T. Vermeulen <jtv at xs4all.nl>
> To: Trigve Siver <trigves at yahoo.com>
> Sent: Sunday, August 23, 2009 9:45:55 AM
> Subject: Re: [Libpqxx-general] Unnamed prepared statements
> 
> On Thu, August 20, 2009 14:44, Trigve Siver wrote:
> > Hey Jeroen,
> > any news on this?
> 
> I'm playing with it now...  It's weird.  It looks like the backend will
> forget the unnamed prepared statement at a transaction boundary.  Any
> transaction boundary at all: commit, abort--even begin!
> 
> (The reason it doesn't fail the first time is that libpqxx doesn't really
> prepare the statement until it's first executed, so that happens inside
> the first transaction).
> 
> The only thing I can think to do about this is make sure we've got a clear
> picture, and document it.

Hi,
I'm going to vacation tomorrow so I'll be of no help for two weeks. One workaround could be that if the prepared stmt is unnamed we will always prepared it regardless what last unnamed prepared stmt definition was.

> 
> Jeroen

Trigve