Re: DBD::PostgreSQL

On Monday, November 18, 2002, at 09:33 AM, Tim Bunce wrote:

>> Okay, then that's what I'll do. Do I check it like this?
>>
>> if (DBIc_has(imp_dbh, DBIcf_BegunWork)) {...}
>
> Yeap.

Great, thanks. I think I'll set the minimum DBI requirement to 1.20 in
order to properly support this feature. Any objections, DBIers?

>> would count as an error, I think. DBD::Pg doesn't report it as an
>> error, though -- it just returns false. Should I add an appropriate
>> call to do_error() in such a case?
>
> Probably. It's fairly important that a method doesn't return an error
> status without having recorded the error by at least doing
> sv_setiv(DBIc_ERR(imp_xxh), ...)

I'll add the call, then.

Regards,

David

--
David Wheeler AIM: dwTheory
david(at)wheeler(dot)net ICQ: 15726394
http://david.wheeler.net/ Yahoo!: dew7e
Jabber: Theory(at)jabber(dot)org