On Mon, Oct 19, 2009 at 12:49 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> "David E. Wheeler" <david@kineticode.com> writes:
>> I'd sure love $, as it's like shell, Perl, and other stuff.
>
> This discussion has gotten utterly off track. The problem I am trying
> to solve is a non-Oracle-compatible behavior in plpgsql. I have got
> substantially less than zero interest in proposals that "solve" the
> problem by introducing notations that don't even pretend to be
> compatible.
OK. In that case, it seems like we should offer options #2 and #3
with a GUC or #option to switch between them. Nobody has made an
argument in favor of keeping #1 around. I'm still strongly of the
opinion that #3 (error) should be the default behavior to avoid silent
failures.
...Robert