FMPRO-L Archives

January 2011, Week 5

FMPRO-L@LISTSERV.DARTMOUTH.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
"Richard S. Russell" <[log in to unmask]>
Reply To:
FileMaker Pro Discussions <[log in to unmask]>
Date:
Mon, 31 Jan 2011 16:06:26 -0600
Content-Type:
text/plain
Parts/Attachments:
text/plain (16 lines)
On 2011 Jan 30, at 22:39, Mike wrote:

> Hi Richard,
> 
> I can verify your result. I even remember, but can't find, a discussion of this from a number of years back.
> 
> I recall from those discussions that the right hand table of a relationship must be an indexed field, but perhaps this isn't as solid a rule as I remember. You've used a global, which can't be indexed, but seem to have achieved the same result as the 'x' comparative operator (I forget the proper name) in your relationship. An interesting side fact is that you can use 'Allow creation of new records in this (your 'B') table'  for your relationship and this can't be done with the 'x' join.
> 
> Still, it looks like you really want the 'x' join. This will eliminate the need for the global field 'B:ASeq Socket' and keep your tables cleaner.


As it happens, a cross-product ("X") relationship is exactly the behavior I'm getting now, but which I don't want.

Unmentioned in my original question (since it wasn't directly relevant to the issue at hand) was that there's stuff going on out past Table B where it's important to know what the active record is in Table A, and I'd lose that 2nd-hand connection unless Table B is aware of the code for the active record in Table A.

If, however, I can't figure out a way around my current problem, then I guess I get to explore alternatives where I store the active-record code as a global in Table C instead of Table B (except that I suspect it'll produce the same kind of oddball results).

ATOM RSS1 RSS2