After Updating Survey123 app, required fields are not enforced at record level.

2090
9
02-13-2019 05:42 AM
AdamDick2
New Contributor II

Since the latest update, a user can skip over required fields from record to record.  For example, when measuring trees, the user can skip entering tree height and move to the next tree.  When the plot is completed, only then will it warn you that required fields have not been filled.  In previous versions, the user could not proceed to the next tree until all required tree attributes for the current were filled out.

0 Kudos
9 Replies
by Anonymous User
Not applicable

Hi Adam,

Are the fields you are referring to that are not validating the required on focus change located inside a repeat?

Can you provide a copy of your survey xlsx form and details of which questions this occurs on, so we can take a closer look?

Phil.

0 Kudos
AdamDick2
New Contributor II

Hi Philip,

Can I provide my xlsx to you directly?  Otherwise, I will pare down to essentials.  It appears it is only integer fields that are not being warned if blank and required.  If I test the application on the desktop Connect app, the validation works as expected.

0 Kudos
AdamDick2
New Contributor II

Further, I just got an iPad back in from the field that still had the older version of the Survey123 app still installed.  I can confirm that with the same data collection form, the behaviour is different in the old app version than the new.  In the old version, if tree level measurements that are required are not filled in, you cannot proceed to a new tree.  In the new app version, you can omit required information and still proceed to a new tree.  To your original question, they are within a repeat.

0 Kudos
by Anonymous User
Not applicable

Hi Adam,

Yes, there is a change in the required fields validation between 3.1 and 3.2 inside repeats. We are currently working on improving how required fields are validated in repeats, mainly so that in repeat records that are not visible, ie when you use repeat count, can not be submitted if fields are required but empty. Previously the app would allow you to submit empty required fields in repeats not visible. This has been fixed in 3.2 and was a bug previously reported and requested by many users, however the downside now is that the behaviour of how the repeats get validated for required is now different. It will occur on submit versus as you change focus from each field.

We will continue to improve the repeat validation so keep an eye out for further improvements which each release.

Phil.

0 Kudos
AdamDick2
New Contributor II

Thanks Phil.  Two steps forward, one step back.  Per record validation is pretty important, I look forward to this functionality reappearing in future versions.

Adam

ANRGIS
by
Occasional Contributor

Since the addition of validation through equations for required fields, it seems like the functionality should be reverted so that required fields within repeats cannot be skipped.

For instance, we can have a validation question... To keep this in the realm of trees, say a question whether a measurement is required. 

This question is default to 'NO' so the repeat is hidden. I can set the required measurement fields to NOT be required since the repeat is hidden using an equation based off the validation question.

If 'YES', then the measurement repeat is visible and the required measurement fields are now required because the validation question is now true.

What it fails to do now, if there are multiple measurements, I can submit repeats without entering values within the measurement repeat section... If I'm not submitting the survey until later (say, I'm surveying many trees), I don't realize that I've left those required measurements blank. 

0 Kudos
DougBrowning2
Occasional Contributor III

See these recent posts.  It is well know that requireds are not working right in 3.3.

 

Bug when a field with a required is inside a group that has a relevant 

 

Survey123 required fields not working as expected 

?

Esri 123 mod have you considered a Sticky post or 2 for this kind of stuff?

   Like Current Open Bugs sticky (with workarounds)

   Common issues and how to fix them sticky (like commas in lists)

Even when I know about a post finding the right post with the fix can be tough.  There is a pretty long list of known issues now so having them all in one place would make it easier and cut down on all the repeat posts.

thanks

0 Kudos
ANRGIS
by
Occasional Contributor

Hey Doug,

I was merely giving expanding on my experience with it. Typically when these cases arise, Esri-staff will take note and add stakeholders to a particular issue. 
I agree however, that having a place of known issues would be nice to go to with perhaps a link of a workaround to get you by until the issue is resolved. It's also nice to validate that an issue you are experiencing is known and not something I'm missing.

Take care.

0 Kudos
by Anonymous User
Not applicable

Hi Adam/Doug/ANR GIS,

We have just updated new 3.4 beta builds to EAC which include fixes for required questions when using relevant, constraints, and validation in repeats. Please test them out and let us know if you still have any remaining issues.

 

You can find the latest 3.4 beta builds here: Survey123 Early Adopter Community

Phil.