> > Sorry, I don't follow you here - what do you mean to do? Remove the > > event completely so we can't wait on it? > > > > I'd like to use the win32 provided recv(), send() functions instead of > redirect them to pgwin32_recv()/pgwin32_send(), just like libpq does. If > we do this, we will lose some functionalities, but I'd like to see the > performance difference first. -- do you think that will be any difference? I personally strongly doubt this will make a diffenrence. Anyways I think we might be looking at the wrong place. Here was my test: 1. drop/create table two fields (id int, f text) no keys 2. begin 3. insert 500k rows. every 50k get time get geometric growth in insert time 4. commit I am doing this via type dump.sql | psql -q mydb I rearrange: every 50k rows get time but also restart transaction. I would ex Guess what...no change. This was a shocker. So I wrap dump.sql with another file that is just \i dump.sql \i dump.sql and get time to insert 50k recs resets after first dump... Merlin
