[SQL] select within aggregate?

2005-05-06 Thread Vortex
Hi!! Of course the following is possible in various ways but i wonder if there is an elegant and simple solution: A part from apache log: remote_hostvarchar(50) request_urivarchar(50) ts timestamp with time zone Now i want to see which host has done how many requests. This is

Re: [SQL] select within aggregate?

2005-05-06 Thread Vortex
On Fri, 6 May 2005 13:34:26 +0530 "Ramakrishnan Muralidharan" <[EMAIL PROTECTED]> wrote: > SELECT abc.remote_host , c , abc.request_uri , a.t FROM abc , ( select > remote_host , count(*) as c , max( ts ) as t from abc group by remote_host ) > as a > where a.remote_host = abc.remote_host and ab

Re: [SQL] select within aggregate?

2005-05-07 Thread Vortex
On Fri, 6 May 2005 07:49:23 -0500 Bruno Wolff III <[EMAIL PROTECTED]> wrote: > You could use max(ts) as long as you can have two requests occur at the > same time. This might be a reasonable assumption, but it might have been > better to include a sequence so that each record would be guaranteed t