That's a bit odd since grid has an internal variable to disable the prefetch.
I understand that the prefetch is there to increase the grid performance, but this is also preventing us from implementing our own cache system. I already mentioned that the grid is triggering many unnecessary requests in this post. Dino confirmed the problem was created by the prefetches. In large systems, we can't denied the global performance issue caused by those request.
An other options (and probably the best one) would be to specify if the grid request is a fetch or a prefetch in the readOptions parameter passed to the transport.read. That way, we would have the choice to create a normal request, implement our own logic or to simply ignore the request (if the success callback can handle an empty result).
Please, take a moment to consider providing us flexible prefetch options. I'm not developing for Telerik, but after looking at the source code, I think small changes in the grid would allow way more flexibility in the grid behavior.
answered on 26 Jan 2015, 08:11 AM
Thanks for the suggestions!
I will put them together and forward them to the people behind the virtualization functionality of the grid. In the mean time you can check our UserVoice section, where you can always post any suggestions that you have for our products, as all posts there are reviewed and considered for future implementation.
Join us on our journey to create the world's most complete HTML 5 UI Framework - download Kendo UI now!