In our company we are facing problems with huge datasets on client side. So far there wasn't any good solution with MIT license, we've decided to create our own:
Check it our here RevoGrid.
The biggest challenge was to create multiple viewport renderers which will work fast. Except of usual virtualization we are using VNode DOM, which updates only changed parts, but we went much farther:
- We are monitoring nodes per scroll and move nodes from edges, not entire screen redraw. With this we achieved amazing performance.
Now we are going to add more features and grow up.
Thank you for reading, hope to get some feedback from you guys.