Hello:

In the distant past I worked on a Prime Information system (pick based) were we could group multivalue columns within the dictionary. In the Arev 3.12 system I'm currently working in, ocasionally I'll see numbers after the "M" in the S/M column of a dictionary. The system that I see these in originally came from an AREV 1.16 system.

Is there any way to group associated multivalues within a dictionary definition? If so, where do I look to read up on that? I know that I can associate multivalues within a window but can I do it any where else? Are the remains of what looks like assocated multivalues in our system something that is no longer supported?

Any information you can send my way would be appreciated.

Thank You,

Michael Slack

Arev does not support this information within the dictionaries themselves. You either have to create a separate file or create group items in the dictionary as Prime Information would create PHrases for the AMV groups.

For example in the dictionary of Accounts_Receivable create a Group item "Pay_Hist" which would contain the name of dictionary items "Pay_Date","Amount_Received","Pay_Method","Pay_Id".

I've waved my arms about the lack of this information since version 1.0 to no avail. (Revelation stored this information in the dictionaries as does every other Pick based system). If Arev had supported this feature, the same powerful AMV reporting features that are found in Prime Information/Universe could have been incorporated into RList. However this was not the case.

And having this information in the dictionaries would certainly make it easier to write a utility to normalize the data for export.

If you're seeing numbers in the dictionaries it means the application or parts of it were converted from Revelation, which supported this information.

While you can enter and maintain this information with the editor, none of the dictionary tools (DICT & MAKETABLE) support this.

ARev has never supported this 'feature' in the dictionaries.

Thanks for the information. I thought that something like that may be the case but I thought I'd give it a try.

Michael Slack

Thinking about this further, there may be a way to

express associated MVs in a dictionary. You

could use group (type G) records called AMV1, AMV2…etc,

adopting the convention that the first field is

the controlling data item. You could then

LIST FILENAME AMV1

and see all the related data together.

Processes could also read this data from the

dictionary record for export purposes.

You may even be able to interrogate the groups

information in windows/template records and

construct your AMV1…AMVn records automatically.

FWIW,

Steve

View this thread on the forum...