"Bruce" <bruce@aristotle.net> wrote...
> I'm getting rather frequent crashes from the Access 2003 replication
> conflict viewer. I recently synchronized two Access 2000 format
[quoted text clipped - 4 lines]
> Jet 4.0). There was also data synchronized in addition to the design
> change. The data synchronization incurred some data conflicts.
How many conflicts?
> I
> opened the Access 2003 replication conflict viewer on one of the
[quoted text clipped - 4 lines]
> element '56' does not exist", followed by a "run time error 440,
> automation error" dialog.
How many fields in the table?
> Does this indicate database corruption?
Not that I know of -- but then I have never heard report of this, even back
when I was the dev owner of the viewer.
> Also, if I make a design change in the design master, do I need to
> then synchronize each replica directly with the design master to
> propagate the change, or can I synchronize the design master with the
> hub and then synchronize the hub with the other replicas in the set?
You can synchronize any replica. I doubt the design change has anything to
do with the error, fwiw.

Signature
MichKa [MS]
NLS Collation/Locale/Keyboard Development
Globalization Infrastructure and Font Technologies
This posting is provided "AS IS" with
no warranties, and confers no rights.
Bruce - 20 May 2004 16:06 GMT
> "Bruce" <bruce@aristotle.net> wrote...
>
> > I'm getting rather frequent crashes from the Access 2003 replication
> > conflict viewer.
> How many conflicts?
The times I have seen this happen the number of conflicts has been
small...always less than 10.
> > I
> > opened the Access 2003 replication conflict viewer on one of the
[quoted text clipped - 6 lines]
>
> How many fields in the table?
There are 172 fields in the table. The option to show only the
conflicting fields is selected.
Bruce
Michael \(michka\) Kaplan [MS] - 20 May 2004 17:27 GMT
"Bruce" <bruce@aristotle.net> wrote...
> > How many fields in the table?
>
> There are 172 fields in the table.
This would be the main source of there being a problem. Which is not to say
it wouldn't still be a bug. But the bulk of the testing was done on much
more reasonably sized (and often properly normalized) schemas.
> The option to show only the
> conflicting fields is selected.
Note that this means the viewer has to load every item and compare them to
see if they are the same or not. Depending on the complexity of the fields
and the amount of change, the time it takes to "compare every item and
display the changed ones" vs. "display every item" can vary in terms of
which is more intensive.
My advice would be to do some schema design, specifically to properly
normalize the data, as a design that is so far from 3NF as to have tables
with 172 columns in them is fraught with either unnecessary duplication,
wasted space, and/or data inconsistencies which replication will only be
hurt by.

Signature
MichKa [MS]
NLS Collation/Locale/Keyboard Development
Globalization Infrastructure and Font Technologies
This posting is provided "AS IS" with
no warranties, and confers no rights.