Відео доповіді

There is generally a good understanding on how to render large (say, 100K items) datasets using virtual lists, …if they remain largely static. But what if new entries are being added or updated at a rate of hundreds per second? And what if the user should be able to filter and sort them freely? How can we stay responsive in such scenarios? In this talk we discuss how Flipper introduced map-reduce inspired FSRW transformations to handle such scenarios gracefully. By applying the techniques introduced in this talk Flipper frame rates increased at least 10-fold and we hope to open source this approach soon.