I often introduce a “hide” column in my spreadsheets, by which I mean a boolean column together with a filter on that column. If you select a cell in the hide column, it has the effect of hiding the corresponding row.
I have found that in Grist, whether the row is actually hidden in response to the selection action is non-deterministic. Often it works, but other times the row stays in place, and I have to reload the page to force a full recompute.
I’m not sure, but what I think is going on is that there’s a race condition between the update of the underlying table data and the filter computation. Is this a known issue? I haven’t been able to find any discussions of it online, but maybe I’m missing something.
Do you have any other columns with filters turned on?
If you edit a cell in a way that would normally cause the row to be filtered out, the row stays visible so it doesn’t disappear while you’re editing it. The one exception is boolean column filters, which can cause row to disappear.
What you’re probably seeing is:
If toggling the Hide boolean makes the row disappear, the row hasn’t been “exempted” yet from filters—no other edits have been made that would have caused a filter to hide it.
If toggling the Hide boolean doesn’t change whether the row is visible, the row was likely “exempted” from filters because of an earlier edit in a column with an active filter.
Hope that helps
Here’s the GitHub discussion where this filtering behavior is explained, including why boolean columns are treated differently than other columns.
