I noticed this a few days ago trying to test edge cases with larger datasets in the table - if you have a large remote dataset, that has a small number of large groups for any given column (lets say 1000 rows with only 2 distinct values for any given column), on initial load the first group will render correctly but as soon as you collapse the first group, bringing the second group into the visible section of the table, all of the rows will be empty.
I'm assuming this is because the remote fetch only happens on scroll? Presumably as the first expanded group is the only group visible initially, the table caches those rows but nothing else. If you scroll down a few rows after collapsing the first group, the second groups rows abruptly render as the remote fetch is triggered.
I'll try and create a simple test harness to simulate this using the remote data example, but hopefully the explanation is descriptive enough in the first instance.
I noticed this a few days ago trying to test edge cases with larger datasets in the table - if you have a large remote dataset, that has a small number of large groups for any given column (lets say 1000 rows with only 2 distinct values for any given column), on initial load the first group will render correctly but as soon as you collapse the first group, bringing the second group into the visible section of the table, all of the rows will be empty.
I'm assuming this is because the remote fetch only happens on scroll? Presumably as the first expanded group is the only group visible initially, the table caches those rows but nothing else. If you scroll down a few rows after collapsing the first group, the second groups rows abruptly render as the remote fetch is triggered.
I'll try and create a simple test harness to simulate this using the remote data example, but hopefully the explanation is descriptive enough in the first instance.