At 21:14 +0100 3/8/02, Roger Baklund wrote:
>* Paul DuBois
>> At 17:00 +0000 3/8/02, Corin Hartland-Swann wrote:
>> >This works for a varying number of rows (usually 5,000 to 10,000) before
>> >it silently exits the loop. I'm using MySQL 3.23.49a
>> That's also your problem. With mysql_use_result, it's *required*
>> that you completely finish the query before issuing another one.
>> I suppose you could write the IDs to a file, then read them back in
>> and use them to issue the UPDATE statements.
>Why not make two connections, SELECT from one and UPDATE through the other?
I thought of that, too, but I don't think it would work. The SELECT locks
the table until it's done. That will lock out the UPDATE, which needs a
WRITE lock no matter which connection it's done over.
Of course, if anyone tries it and find that it works, I'd be interested
to know that.