Re: How to get around LIKE inefficiencies?

On Mon, 6 Nov 2000, Philip Warner wrote:

> At 23:12 5/11/00 -0400, The Hermit Hacker wrote:
> >
> >Except, if we are telling it to get rid of using the index, may as well
> >get rid of it altogether, as updates/inserts would be slowed down by
> >having to update that too ...
> >
> 
> So long as you don't ever need the index for anything else, then getting
> rid of it is by far the best solution. But, eg, if you want to check if a
> page is already indexed you will probably end up with a sequential search.

except, it appears that they both store the text of the URL *and* the
CRC32 value of it, and do other queries based on the CRC32 value of it ...