Re: ADO Data Control Concurrency Problem
From: Ralph (msnews.20.nt_consulting32_at_spamgourmet.com)
Date: 11/27/04
- Next message: Galen Somerville: "Re: Talking to USB"
- Previous message: lofepramine: "Re: scheduler"
- In reply to: Andrew D. Newbould: "Re: ADO Data Control Concurrency Problem"
- Messages sorted by: [ date ] [ thread ]
Date: Sat, 27 Nov 2004 09:02:46 -0600
"Andrew D. Newbould" <newsgroups@NOzadSPANsoft.com> wrote in message
news:6UkuDyrGtRpBFwB3@zadsoft.gotadsl.co.uk...
> <snip>
> >> Two users open the same row of a table in Oracle for editing ; one of
them
> >> commits the changes to the database. After this event, the 2nd user has
an
> >> outdated version of the row on his/her screen. ADO doesn't notify the
> >> application that a row has changed.
> >> Editing of row in table is done with the help of ADO Data Control. We
have
> >> used Cursor Location "adClient Side" for the connection.
> >>
> >
> >A 10 cent answer in four parts.
> >
> >1) ADO has nothing to do with it.
> >2) By requesting adClient you have basically asked for a bag of stuff as
it
> >exists at the time of the request that you can search thru and generally
> >play with - all logical connection with Oracle has been lost. ie, the
server
> >has no remaining interest in what it delivered.
>
> Not exactly true. Cursor location does not bare on whether the
> connection to Oracle has been lost just that the data resides in the
> Client or Server memory. The "ActiveConnection" needs setting to
> "Nothing" in order to break the connection (ie: Disconnected Recordsets
> of Firehose cursors).
>
You are correct. My use of the term 'lost' was a bit of hyperbola in
expressing the fact as you noted - the database engine no longer has any
interest in what was provided.
I would also like add to your clearer explanation a bit of unrequested
advice. <g>
Be very careful when reading about "ADO cursor types" to appreciate that
what the manual might say about a particular cursor and what a particular
provider and DB engine may actually do, below the surface, can be very
different - though the 'logical' result may appear to be the same.
-ralph
- Next message: Galen Somerville: "Re: Talking to USB"
- Previous message: lofepramine: "Re: scheduler"
- In reply to: Andrew D. Newbould: "Re: ADO Data Control Concurrency Problem"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|