On Wed, 12 Jul 2006, Tom Lane wrote:
It's probably worth pointing out here that the MySQL behavior they seem to be expecting is considerably further from the spec than Postgres's behavior. If I'm reading between the lines correctly, they are expecting foo and Foo (both written without double-quotes) to be distinct identifiers. But these are the same identifier per spec, because the spec *requires* case-folding of unquoted identifiers.
Thanks very much, Tom. It turns out that the problem was index names. I tried to solve that, but being unfamiliar with the entire application structure, I ended up defining the same index name more than once. Sigh. I do appreciate the insight about case. I've always used only lower case for table and field names, even back in the DOS days when I was writing database applications in C. Rich -- Richard B. Shepard, Ph.D. | The Environmental Permitting Applied Ecosystem Services, Inc.(TM) | Accelerator <http://www.appl-ecosys.com> Voice: 503-667-4517 Fax: 503-667-8863
