Chuck,

On Sunday 12 December 2004 04:01, Chuck Williams wrote:
> I'd be surprised if the function call overhead was significant, but
> nonetheless I can't argue with optimizing the sum case.  However, it

In case you have performance data, I'm all ears.

...

> <soapbox>
> I maintain the belief that max is *required* to implement reasonable
> multi-field searching (1).  I can't imagine a case where the current

I agree that a maximum score over fields is useful.

...
 
>   > -----Original Message-----
>   > From: Paul Elschot [mailto:[EMAIL PROTECTED]
>   > Sent: Saturday, December 11, 2004 2:05 PM
>   > To: [EMAIL PROTECTED]
>   > Subject: Re: Boolean Scorer
>   > 
>   > Chuck,
>   > 
>   > On Friday 10 December 2004 23:12, Chuck Williams wrote:
>   > > Paul,
>   > >
>   > > Would there be a way to get the best of both worlds?  E.g., could
> you
>   > > factor the specializable score combination differently, so that
> one
>   > > method was called with each new score to generate a state entity,
>   > while
>   > > a final method computed the score from the state.  For both sum
> and
>   > max,
>   > > the state entity could just be a float, not requiring an array.
> The
>   > > final operation for the sum with coord case would do the coord.  I
>   > > haven't looked at the code carefully enough to see if this
> actually
>   > > works, but it seemed worth mentioning.
>   > 
>   > It's simple enough to do some abstract method call instead of
>   > initializing
>   > a sum or adding to it. The problem is that as long as such a call is
> not
>   > effectively inlined by the JVM, it will cause a performance hit for
> the
>   > sum
>   > case.
>   > 
>   > The latest version of the advanceAfterCurrent method that computes
> the
>   > score is java protected. It can be overridden to make the best
>   > of it in another world.

I'm afraid I should have left the last three words out. I'm sorry if they
can be interpreted in a negative way, that was not my intention.

Regards,
Paul Elschot


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to