Hello Tim,

> << Both Table and Query. >>
>
> Actually it's quicker just to simply count up to the current position.  The
> maintenance alone on such an index is very time-consuming since it has to be
> constantly synched to the current state of the table by rebuilding portions
> of it.

Can you make it transparently ? (without needing to do
anything from us)

--
Charalabos Michael - [Creation Power] - http://www.creationpower.com -
http://www.creationpower.gr

Steve,

<< Well, to me that sounds like a great reason to dump the stats. >>

Yes, it is a compelling reason.  The stats don't prevent fail-safe writes -
it just makes them take longer.

--
Tim Young
Elevate Software
www.elevatesoft.com

Roy,

<< If the fail-safe applies to fileserver mode as well then no stats becomes
a no brainer (DBISAM is good but I do have to do repairs occasionally) >>

Yes, it does apply to fileserver mode also.

--
Tim Young
Elevate Software
www.elevatesoft.com

Michael,

<< Can you make it transparently ? (without needing to do anything from us)
>>

Do you mean adding into EDB the "counting up to the current position" part ?
If so, then my answer would be no.  It would be dog slow and not even worth
the effort.

--
Tim Young
Elevate Software
www.elevatesoft.com

Rita,

<< Well you asked for feedback and you got it, most peeps say dump the
stats. >>

Yes, I did.  I'm kind of surprised at the feedback so far, although
pleasantly so.

--
Tim Young
Elevate Software
www.elevatesoft.com

Tim Young [Elevate Software] wrote:

> Would you consider the removal of the index statistics, and the
> subsequent removal of the ability for ElevateDB to provide option 1)
> above significantly negative ?
>

I say remove them.

--

Tim and others

Since DBGrid scrollbars works only in some DBISAM conditions and not in ALL
situations. Then my vote will go to REMOVE STATS.

PageUp and PageDown keys do the task of page records.

I agree with others that users like precision DBGRID but...

Eduardo

I'm Joining this feedback very late, but - if the difference between having
and having not is a few lines of code, is it possible to allow the end user
to do a Conditional Compile using the DEFINE option in the compiler option?
(Or has this been mentioned). I scanned through the huge thread, but didn't
see this.

Cheers

Adam.

Adam

This has been asked a few times and Tim has each time said
quite clearly that it is 'one or the other'.  There are
fundamental structural differences.  

No sitting on the fence!

Malcolm

Hi Malcolm

> This has been asked a few times and Tim has each time said
> quite clearly that it is 'one or the other'.  There are
> fundamental structural differences.  

I was going through the threads, but didn't catch that reply - thanks for
clearing it up.

I guess it's too late to put my 2c worth in anway now - Beta release
happening any day now!  

Cheers

Adam.