Field validation by calculation issue, possible bug or just an oversight?

[logicgator]

Greetings,

Curious if anyone has ever experienced a field validated by calculation that is validating correctly (when anything on the record is modified - it is meant to only require a value when another field has a specific value), but when certain repeating fields on the same record are modified, it acts as though the original field is suddenly invalid/empty?  Additionally does not seem to bring me to the field in question when I select "No" from the custom message (doesn't seem to bring me anywhere).

The validation calc is simple: If(IsEmpty(Self) and Field1>0;0;1)

I know the field is in a validated state because 1. it's not empty, and 2. other fields being modified on the record are committed without any error, not to mention the first repetition of said repeating fields as well.  It's only when the second repetition and onward are modified.  Not all the repeating fields in this file either, just a few.

(Edit: another reason I know it's being validated is that if I remove the value from Field1, I don't get any errors at all no matter what I modify.)

I have checked to be sure that these repeating fields are not involved in any calculations pertaining to the validation of the original field.

It seems like a bug based on the erratic conditions of the issue.  The file I'm working in is also very old and has been converted a few times over the years with FileMaker version upgrades... chance it could be corrupted in some way?  Any ideas will be greatly appreciated.

Edited June 28, 2023 by logicgator
additional info

[comment]

Can you reproduce the problem (and only the problem) in a brand new file? If yes, please post it here.

--
A side note: your formula could be simplified to:

not ( IsEmpty ( Self ) and Field1 > 0 )

[logicgator]

Thanks 😎 - simplified formulas are one of life's small joys. 

Works as it should in a brand new file (of course).

[comment]

17 minutes ago, logicgator said:

Works as it should in a brand new file

Ok, then make a copy of your original file and run a recovery on it. If the recovered file does not exhibit the problem, then your file is very likely corrupted. Otherwise it contains some "feature" that you're missing.

 

[logicgator]

21 hours ago, comment said:

Ok, then make a copy of your original file and run a recovery on it. If the recovered file does not exhibit the problem, then your file is very likely corrupted. Otherwise it contains some "feature" that you're missing.

 

So I did this today, and the recovered file also has the same issue.  What's weird is that any new repeating fields that I create also trigger this false invalidation if any repetition after the first one has been modified.  Even if I duplicate them from the other fields that work.  Can't think of any feature that could be contributing to this.  Guess I'm looking at a file rebuild, or changing over to separate tables/portals for my repeating fields.

Thanks for your input!

[comment]

There is always a possibility that some corruption survived the recovery, though I don't think it's very likely. In your place I would take the recovered copy and strip it to absolute minimum required to demonstrate the problem. If at that point you still cannot see what causes it, post it here.