Youtube mates with epub

[haertig]

Originally Posted by HarryT

Video is streamed, not downloaded as a whole and then played.

I get your point. But this is not 100% true.

The sibling to caching a webpages static contents is buffering a streams contents. Depending on the length of the video and the storage available on the local end (either memory or disk), the entire video may be downloaded (buffered) and then streamed from the local copy. Depending on local settings, cache may persist after the application exits. Buffering seldom does, but it could. Actually, the buffer probably still exists (until it's overwritten), it's just that you don't have any easy way to access it outside of the application that created it. Additionally, the memory/disk where the buffer is stored has no doubt been marked as "free", so you are at the mercy of the OS when it comes to reusing this free area, thus overwriting the buffer. You would normally have to get into computer forensics, memory dumps, etc., in order to access this buffer outside of the application Not a skill typically possessed by normal users.

This is mostly technical talk though - the intent for a video stream is most probably to not be available locally as a standalone. Likewise the intent of a webpage cache is probably not to make content available locally standalone either, it's just much easier to find a cache than it is to find a buffer.

The problem for online providers is that if you make something available online, you have effectively made it available for offline use as well, given even a minimally technically competent end user. As a provider you may not intend for this to happen, but there's not much you can do to prevent it. You can lobby to have laws created to prevent it, but for 99% of the cases you'd never even know a copy had happened, let alone be able to enforce a law against it. The good news for online providers is that probably only 0.001% of their end users could be considered technically competent. So their existing attempts to prevent local copies are generally quite effective, and they don't have to go after web browser developers in regards to caching/buffering. That would not be a popular thing for them to do, since it would significantly effect performance and efficiency ... something that even non-technical end users would notice and complain about.