Page History

...

The reason to keep `materialized`,  is that the window store requires a concrete retention time, window size and number of rolling segments to properly setup. On the application side the stream job could not infer the windowed topic retention or window size, so these are required information from the user. 

The reason to keep `consumed` , is that in the table() API, the reason   we could omit `consumed` or `materialized` is that replace `materialized` serde with `consumed` serde because they could share the `keySerde` and `valueSerde` during state store and node construction. This in In the windowed table context, however, is this will not truework, since we are using both windowed keySerde and raw keySerde at the same time. So both structs are required.

...