SQL Power Software Forum

Please let me congratulate on a great product, once again.

I have one more suggestion. I have a demoralized table of donors and donor donations. I would like to create an entity for donors (master table for donors). The process I am looking will involve two separate processes. First, identify duplicate donors and second generate a complete record of donors from multiple donation records. This can be driven by rules. An example of one such rule is that the most recent record of donations is a starting point for the master and the remaining blank fields can be populated from the other donation records.

ismail

I like this idea!

This could be a refinement to the "auto-match" function which, frankly, I hope nobody uses at the moment... as is, it's too imprecise. If you could specify to the MatchMaker a set of rules for choosing the master record within a cluster, that would make auto-match a more predictable process. In your case, the rule is easy (the master is the record with the maximum date value), but I wonder how rich we can make the rule system.

I have a feeling we could take a lesson from email filtering criteria GUIs on this one. And if you have any ideas about how to design a good GUI for this, we'd be glad to hear them.

Of course, the second part of your request (filling in the blanks in the master record) is already implemented in the "merge rules" facility.

-Jonathan

Concerning the "auto-match" feature, is there a plan to make it more precise? Right now, it doesn't seem that there's a way to configure it other than through the merge rules. For example, it would be great (for me, at least), if we could be able to create some sort of "if, then" configuration.

Hi JWiley,

There isn't any way to tweak the automatch to your liking at the moment, but having the ability to do so does sound useful. We would still highly recommend that you go over automatch results to ensure that they really are what you want them to be.

Are there any particular conditions that you would like to be able to set in such a configuration?

In my table, I have records that have different dates and different values in duplicate fields. However, I would like to have the merge process have a rule that records should be merged on the later date (which I could input) and by a given value.

This message was edited 1 time. Last update was at 2008-10-17 16:09:07

So are you talking about:

1. a rule system in auto-match that lets you give hints on how to decide which record is elected to be the master during the auto-match process; or

2. a rule system that applies during the merge process by overriding the master/duplicate chain that was previously established during match validation?

If that's too abstract, it would probably help us both if you could clarify with a concrete example (make up 3 or 4 rows of fake data, then walk us through the match validation and merging process as you'd like it to work).

We really appreciate your input on this--I want to be sure we have a clear understanding of what you're suggesting.

-Jonathan

Jonathan,

You hit the nail on the head. Your description is perfect.

So is it option 1 or option 2 that perfectly describes what you're looking for?

-Jonathan

Sorry, it would be option #1.

Ok, great. That's the one we've been planning to implement!

As always, I can't promise any specific timelines, but I can assure you it's on the list. Also as always, we do give priority to enhancement requests from our paying support customers.

Either way, you can subscribe to bug 1686 to get automatic notifications when we work on this feature.

-Jonathan

Thanks a lot. I will subscribe to this bug.