Subject | RE: [IBO] TDataSet vs. Native |
---|---|
Author | Ondrej Kelle |
Post date | 2001-01-07T08:31:02Z |
Hello again,
thanks for all responses. I didn't sleep very well last night because I
later realized that perhaps in my previous posting I let my frustrations out
too much and was afraid they would evoke some hard feelings. I'm very glad
this has not happened. I can see that this community consists of some
brilliant people and I sure wish you all the best of luck.
TDataSet itself is an abstract ancestor class, there are Prepare methods in
descendants like TQuery or TIBDataSet. I tend to use parameterized queries a
lot and have noticed the performance improvement with Prepare. From your
postings I have the impression that I've been missing something essential
all the time.
- searching:
I presume you mean incremental searching within a result set of a query.
I've always done this using parameterized queries, and I thought I was doing
it the right way.
simple fact that I need to do some studying before I can discuss these
points.
Thanks everybody for your comments, and your generous treatment of a
(somewhat confused) newcomer.
TOndrej
thanks for all responses. I didn't sleep very well last night because I
later realized that perhaps in my previous posting I let my frustrations out
too much and was afraid they would evoke some hard feelings. I'm very glad
this has not happened. I can see that this community consists of some
brilliant people and I sure wish you all the best of luck.
> -----Original Message------ preparing:
> TDataset/TDatasource has no modality to support
> two C/S basics - preparing and searching.
TDataSet itself is an abstract ancestor class, there are Prepare methods in
descendants like TQuery or TIBDataSet. I tend to use parameterized queries a
lot and have noticed the performance improvement with Prepare. From your
postings I have the impression that I've been missing something essential
all the time.
- searching:
I presume you mean incremental searching within a result set of a query.
I've always done this using parameterized queries, and I thought I was doing
it the right way.
> IBX has even less capability than the BDE to moderate dataHere I'm sure I must be missing something. Let me conclude here with a
> requests. It doesn't position or reposition correctly in the
> buffers and
> you have considerable work to do *in each control* to get datasets to
> display and refresh as the user wants them. You need to add
> TClientDatasets (= lots of fat) to emulate the key-sensitive
> behaviours; or just not support them. Native IBO does all of this
> stuff; and it doesn't require you to fetch all records in
> all cases. With
> IBX, all commits and rollbacks close all datasets in a transaction.
simple fact that I need to do some studying before I can discuss these
points.
Thanks everybody for your comments, and your generous treatment of a
(somewhat confused) newcomer.
TOndrej