When to use a snapshot

Hi Raymond,

As you mentioned, we always recommend the refresh row API call for a single row refresh.

In multiple rows refresh case, if the number of rows existing in your datasource is small (<100-200), you can retrieve the primary key value of the current row(to keep the position in the dataset), re-query the grid and then use api_datasource.setCurrentRowByPK API to set the last current row as current again and keep your old position. But remember that when you re-query your datasource, your dataset may be changed since some rows may be added, deleted or updated since your last query. However in snapshot case, you always work with the same dataset as queried at a given point of time.

In multiple rows refresh case, if the number of rows existing in your datasource is huge and the number of the edited rows is small, you can use api_datasource.getQuery API to loop edited rows and refresh them one by one using refresh API.

In our experience, the best use of the snapshot is when you need a savepoint similar sturucture, which is a very rare case.

To cancel the changes, we often use refresh API or the re-query options (or api_datasource.clear API if your grid will no longer be visible to the user).

Hope this helps,
Ibrahim