On Mon, Jan 23, 2012 at 3:06 PM, Robert Haas <robertmhaas@gmail.com> wrote:
> Not being responsible from the maintenance of any PostgreSQL drivers
> whatsoever, I don't have a strong feeling about which of these is the
> case, and I'd like us to hear from the people who do.
I'm just gonna come right out and say that GUC-based solutions are not
the way -- bytea encoding change is a perfect example of that.
IMSNHO, there's only two plausible paths ahead:
1) document the binary formats and continue with 'go at your own risk'
2) full auto-negotiation for all wire formats (text and binary).
merlin