Subject IBO4 - OrderingLinks and POS
Author guardian@pobox.com
Hello Jason,

My email has been down but I looked into the ReleaseNotes
and it mentions using the POS attribute. I put it in the
OrderingLinks and the program seems to need more to do
something. I looked at the SearchingLinks Demo, and I
think I picked up enough to get it to work for now.
But the fact that I have to do:

SQLWhereItems.Add( 'OM_Company STARTING WITH :Company' );

in the PrepareSQL makes me wonder if I am doing this
right. Could you give some more explantion of how
it should all work together. Maybe it needs a artical
in the GSG or something, but I don't understand how
the v4 is better than the v3. In v3 I could

seSearchingLinks.ParamRefinePos := 2;

Which seems to do the same as the POS attribute.
Please clarify?

Thanks
Michael L. Horne



--- In IBObjects@y..., "Jason Wharton" <jwharton@i...> wrote:
> There should be some text in the ReleaseNotes.rtf about how to do
this.
> In short, you add the POS attribute to your OrderingLinks property.
>
> MYCOL=ITEM=1;POS=0
>
> HTH,
> Jason Wharton
> CPS - Mesa AZ
> http://www.ibobjects.com
>
>
> ----- Original Message -----
> From: "Michael L. Horne" <guardian@p...>
> To: <IBObjects@y...>
> Sent: Tuesday, September 18, 2001 7:07 AM
> Subject: [IBO] Questions about new capabilities in IBO4
>
>
> > Hello
> >
> > Jason, I notice on the web site the list of improvements
> > from IBO3 to 4 the first one is:
> >
> > It is now possible to specify a base query which could
> > output a dataset of more than 100K rows and, as a user
> > navigates this large dataset, IBO automatically does
> > the refinement necessary to pull in only those rows that
> > are of interest.
> >
> > That sounds great, but how do we implement it. If I do
> > a locate it seems to start retrieving all records down
> > to where the new record is. IBO3 did that.
> >
> >
> > Also, same question about this, how does it work:
> >
> > Client-side filtering has been added, including the
> > capability to change the filter logic and refresh the
> > buffers without having to refresh the dataset from the
> > server.
> >
> > Thanks
> > Michael L. Horne