Re: interesting "group by" usage

From: fumi <fumi_at_tpts5.seed.net.tw>
Date: Wed, 7 Mar 2001 23:18:51 +0800
Message-ID: <985oc8$96m$2@news.seed.net.tw>

"Zhiliang Hu" <zhu_at_genomicfx.com> ????? news:3AA4276F.C9BB9ED9_at_genomicfx.com...
>
> I have a table that contains jobs done by different people
> at different times and I have a need to see a list of names
> with unique name entries. So I tried a
>
> > select fname, lname from job_done group by fname, lname;
>
> that works! ("not a group by expression" isn't it? ;-)
> -- indeed, when I tried the same thing through a web interface
> (via CGI <-> SQLPLUS) it complains "not a group by expression"
> Any thought why?
>
> Or is there an option that allows return of unique entries?
> (I didn't seem to find one).
>
> Zhiliang

Hi,
I have encountered this problem.
The reason is that the "interface" may re-write your SQL statement, ex., it puts "ROWID" before the FROM keyword.

Consult your manuals carefully,
there must be an option of connections or statements (such like "no cache" or "forward only") to overcome this problem. Received on Wed Mar 07 2001 - 09:18:51 CST