prebuilt-transcription VS react-window

If you are like spoilers and want to see what we’re building today, you can jump straight into the prebuilt-transcription code and also try a live demo

As we have seen in the previous section we can implement virtual scrolling from scratch. In this section, we'll discuss a third-party tool called react-window that allows us to implement virtual scrolling in a much easier way.

If you have been using React for awhile, you may have heard of the infamous virtualisation library react-window or it's predecessor react-virtualized

React Window is used to render long lists. Imagine you have a list of 1,000 items. Only ten are visible simultaneously, but your code tries to render all 1,000 items simultaneously.

You can try react-window by Brian Vaughn

I'm trying to use react-window and react-nice-scroll together but unfortunately it didn't work at all. As soon as I wrap my component with from react-nice-scroll everything disappears.

This is what you should look into. (Virtualization)

Take a look at react-window. But just my perspective; if you haven't needed to look into what virtualization is, it's probably because you haven't had a need for it yet. The one implication (which you'll see in another comment form me in this thread) is that it removes CTRL+F search functionality for items outside of the "window".

Virtualization can be achieved using third-party libraries like react-window or react-virtualized. These libraries provide a way to render only the visible data and load more data as needed, resulting in faster application performance.

If the API data is not paginated (i.e. you’re getting all your data in one go), then you may not need pagination. If rendering performance is an issue, a virtualized list may be more appropriate. If the response is too big and is slow to fetch it’s another issue entirely, since if the API doesn’t support offsets / pages you would need to invent a criteria to paginate your data on, or fetch it on the server using React Server Components which gives you implement a cache layer and potentially allows to only send a chunk of that data to the client, but you see how that can get complex fast.

Oh, right. I totally forgot to mention that– but the idea of "less rendering" in this case seems less like a Suspense concern and more like a windowing concern. I've written a few libraries for that stuff (react-window and react-virtualized) although there are others that may fit your needs better. Their main focus is limiting what you're rendering to more or less only what's on the screen at any given point. Combine that with memoized filtering and I would imagine you're set.