On 10/07/2014 09:08 AM, David Hedberg wrote:
> (Forgot to reply to the list)
>
> On Tue, Oct 7, 2014 at 2:36 AM, Michael Paquier
> <michael.paquier@gmail.com> wrote:
>>
>>
>> On Mon, Oct 6, 2014 at 10:19 PM, David Hedberg <david.hedberg@gmail.com>
>> wrote:
>>>
>>> Resending this patch rebased on latest master. Please apply or tell me
>>> if it looks wrong.
>>
>> At quick glance this patch is correct, and that's actually the same as the
>> one you sent before.
>
> Thanks for looking. Since I've never sent a patch to this project
> before I'm relying on my experience with development in another
> project, which is basically "if it's not in git yet and no one has
> said anything definite, resend or it will get lost". I don't know how
> to proceed here.
Sorry for the slow response. I've had your email lying in my inbox, but
haven't gotten around to it.
In theory, it would be nicer to do changes like this in a major release,
i.e. when we bump from 9.3.XXXX to 9.4.XXXX. But then again, the
pgsql-odbc project has never made much distinction between minor and
major releases.
I'm not 100% sure the check for EN_is_odbc3 is correct. If an
application wants ODBC2 behavior, but the driver is an ODBC3 driver, the
driver manager is supposed to bridge the differences. But I'm not
convinced it's wrong either, so I left it as you wrote it. (Of course,
no-one uses ODBC2 anymore, so it doesn't matter in practice). I also
fixed the expected output for the catalogfunctions regression test - it
contained the old column names.
I've committed this now. The change will appear in the next release. Thanks!
- Heikki