OK, is anyone else seeing this. I am getting the wrong record number reporting when records are omitted from the found set. Get ( RecordNumber ) reports the record number as if all records were found...but it is intermittant. Have reported the bug to FM.

The calculation must be unstored. You could also have some corruption but I don't recall vs. 8 having that 'losing its index mind' problem like 7 did; which would incorrectly display record numbers. What 8 version are you using? I haven't used 8.0v3 yet.

8.5

Get ( RecordNumber) just displays the current record number - it has nothing to do with a found set. But I haven't tried 8.5 yet. Are you saying that you can view the last record in a found set with 20 records and it still says 20 even after omitting some records out of the set?

You are certified so I assume you know to check that the calc is unstored. Sorry I can't help more ...

I now see 8.5 in the subject - I look at a User's profile or the bottom of the first post.

Actually it has everything to do with the found set since it is the record number IN the found set. On validating a field it returns the wrong record number. But yes, even after omitting it thinks it is the 20th record in your example. It doesn't seem to do this in other situations that I can find, but validation...interesting.

Anyways, I have uploaded a file to illustrate the point. Hope this helps. I know I haven't described this in the best way.

Get___RecordNumber__.fp7.zip

Edited July 31, 200619 yr by Guest

The problem I see is the name of the field, "Get ( RecordNumber )". This is not a good name for a field, and Filemaker does warn you that the "selected name cannot be easily used in a calculation formula". I believe your validation formula fails because it refers to the field's content, instead of to the current record number.

Not that I can detect. No matter what you call the field the result is the same. The wrong record number is returned.

When you change your calc name, it works for me. One should NEVER name a field a function. It DOES matter. If it isn't correcting itself, it may be because it's vs. 8.5 but if you create a new calc and test it, I'll bet it works just fine. Creating a new calc works just like changing your existing calc - both produce the correct record number when you eliminate using a function as a field name.

OK, I see the problem now (after I've made a simpler file). Strange bug, and it seems it's been there for a while: I get the same results in 7.0v3 and 8.0v3.

I cannot make it break. Can you explain what you tried? I'm using 8.0v2.

A single text field, validated by:

Get (RecordNumber) = 4

Create a few records, and fill the field with serial numbers. Now omit some records and/or sort them descending, so that the original record #4 is at a different position in the found set. The validation allows modification of the ORIGINAL #4 record, but not the current one.

Sheesh LaRetta, upgrade to 8.0v3 already!

It's just a simple patch and it's fixed a few nasty bugs.

I am not ready to upgrade. I still have some network testing to complete in 8.0v2. I want to then compare to 8.0v3. The only thing that currently bothers us is speed (I've designed around everything else) and that is small trade-off compared to getting answers to FM's internal workings ... and those answers only come from the slight variations between versions. I wish I would have been doing this from 7 forward.

Thanks, Michael, I see the Get(RecordNumber) issue as well now!

Beleive me, it has nothing to do with the field name...that is only for display purposes to illustrate the point. Sorry for the confusion. The problem isn't what is displayed in the field, it is what is calculated in the validation of the field. See the validation script for the field. It should validate OK when you edit the fourth record of the found set, but not anyothers. However, it doesn't work when records are omitted BECAUSE the wrong record number is being returned. Like I said, sorry for the confusion. I didn't really have a good way of getting the point across. But glad that I am not going crazy (completely) and that someone else can see the error even with my poor explanation! :P

Edited August 1, 200619 yr by Guest