Feedback: Basic Column Field-Value Validations - When?

Topic Labels: Base design
872 1
Showing results for 
Search instead for 
Did you mean: 
4 - Data Explorer
4 - Data Explorer

This is mostly feedback for Airtable on why, having been very excited, I’m now disappointed by it after only fifteen or so minutes… tl;dr at the bottom.

I was just introduced to Airtable via the Wintergarten / Marble Machine X Youtube channel (and I commend you for giving them a licence!!). I thought someone had finally built the hybrid database/spreadsheet I’ve been craving for probably 20 years.

But, from what I can tell, there is NO WAY to constrain fields with custom validations, such as e.g. min-max length, permitted characters, etc. This means that using Airtable for structured data capture, with a view to future data import, is STILL going to require data cleanup that otherwise could be enforced at the time the data is entered.

I can see this issue is bought up repeatedly in the forum here, but generally, the answer seems to be ‘no you can’t do that, try this clunky formula-driven-status-twin-field solution’ on an auto-closed thread, with no mention of whether this is in development, in the feature queue, or ‘wontfix’.

It doesn’t have to be complex or even retrospective. i.e. enforce the current validations when data is entered, and if the validations are later changed in a way that invalidates existing data, those fields just need to be highlighted/flagged (unlike e.g. typical SQL table constraints). (actual value-input enforcement could be optional I guess).

For starters, it could even be JUST user-provided-regex matching and then adding a better ‘simple’ form customised for each field type later. It would be fairly simple to implement this, and it must have been discussed internally?

Also, a unique-in-column validation (which I guess would have to flag both of the conflicting fields).

I would like to know if any user-specified field validations are planned, before deciding whether using Airtable is actually going to be worthwhile. These sorts of validations are trivial in Excel / GSheets but of course, the entity relationships are missing.

The inability to define validations on ‘columns’ is why I gave up on SmartSheets ages ago.

/sadface I’ve just checked, and there doesn’t even seem to be a regex matching function, so a simple character and length restriction on a text field is going to be an unwieldy very-long old-school spreadsheet formula full of IF()s, AND()s, OR()s, and IS*()s (vs eg Google Sheets REGEXMATCH().

tl;dr What is the timeframe for adding field/column entry validation? I might be an Airtable fanboy if that appears.

As it stands, it is so-close-and-yet-so-far. If I give someone a multitab Google Sheet, I know I’ll get usable per-field-data out of it, but relationships will be awkward, and a lot of lookup()s etc for the cross-tab checks. With Airtable, I’ll get good, easy, relationships, with a great UI, but all sorts of GI→GO problems with the non-relational attribute values.


NB: 15 mins to hit the wall, but 30 mins to write this feedback, should hopefully convey that I’m hoping validations is ‘coming soon’ because there are other aspects of the product that I love and have been dreaming of finding in a ‘spreadsheet’ forever!

1 Reply 1

Yes, field validation is a missing feature from Airtable. Probably the best you can do is create a formula field that results in some warning message (or some warning emojis) if a field fails your validation. You could even group and/or color records based on that field, to make those records stand out even more. You could also create a new view that shows you the records which have failed validation.

And you are correct that advanced formula functions are missing, too. However, the recent addition of the JavaScript scripting block can probably make some of those things — if not all of them — possible.

In general, I would recommend thinking of Airtable as a “very basic” database program combined with a “very basic” spreadsheet. It definitely won’t meet everyone’s needs. For a more advanced database program, I would recommend FileMaker, which is considered one of the the leaders in the database industry.