[fc417fc802]

> All frame extraction happens client-side via canvas – no server processing, no pre-generated thumbnails.

Doesn't that mean the client has to grab a bunch of extra data when it first opens the page, at least if the user calls up the seek feature? Since you effectively have to grab various frames from all throughout the video to generate the initial batch. It seems like it would make more sense to have server side thumbnails here as long as they're reasonably sparse and low quality.

Although I admit that one line client side integration is quite compelling.

[haasiy]

Exactly. I view this cache similarly to how a browser (or Google Image Search) caches thumbnails locally. Since I'm only storing small Canvas elements, the memory footprint is much smaller than the video itself. To keep it sustainable, I'm planning to implement a trigger to clear the cache whenever the video source changes, ensuring the client's memory stays fresh.