Comment on YouTube tests server-side ads to make your coveted blocker obsolete
dan@upvote.au 3 months agoLet’s go full guerilla: Plugin that lets you select the first and the last frame of an ad, thus allows to report the beginning and length to a synced database. When that frame is found in the buffer, skip X frames ahead.
This would fit in well with SponsorBlock, which already does the same thing for different parts of videos (eg sponsored segments, intro and outro animations, non music segments in music videos, etc).
I suspect YouTube will find ways around this, like running ads of differing lengths, add random amounts of padding at the start of the video or between ads, etc.
The challenge is that videos will have a varying amount or type of ads based on the client’s country/demographic and simply on the timing of ad campaigns.
Not baking-in ads was the advantage of Youtube and other streaming platforms over the likes of traditional TV. That’s why they were client-side in the first place. I wonder how much the extra effort, bandwidth, and processing will cost Youtube to achieve server-side ads. Would be funny if it simply ended up being too expensive for them.
They have a LOT of compute power… They could have several baked in ads per geographical area / demographic and only store them on servers in / close to the relevant country. There’s definitely associated costs but I wonder if it’d amortize well given their viewer count.
Let the games begin.
ArchRecord@lemm.ee 3 months ago
It actually already did break sponsorblock for a bit because user submissions would include the wrong timestamps, due to the ads changing the duration of the video.
This would be hard to implement, but I personally would be happy to donate more to fund the development costs for such features. Adblocking is the largest consumer boycott in history and I won’t let a corporation try to crush it again.