Also a bizarre move considering they are so hard on for ads because Youtube is incredibly expensive to maintain. You are going to pay for comically large servers by… burning more computer time inserting ads into streams? Google screwed adbuyers before by not actually showing the ads where they were supposed to, Idk why they are trying so hard instead of doing more of that.
Not really. If you know what you’re doing video stream manipulation is easy. They might just do the Webm concat thing (two webm files put together is just a single webm)
Not just that, these big tech companies have some genuine black magic fuckery with their tech it’s unbelievable. I about lost it when I discovered that Google Drive files are not all copied 1:1 in their datastores. Same runs of 0s and 1s in your file and other files in anybody’s Google Drive files are stored only once, and so when you retrieve the file, it actually recreates your uploaded file with a bunch of pointers in sequence to all these different data locations
Sounds pretty resource intensive at scale… I’ll be impressed if they pull it off.
Also a bizarre move considering they are so hard on for ads because Youtube is incredibly expensive to maintain. You are going to pay for comically large servers by… burning more computer time inserting ads into streams? Google screwed adbuyers before by not actually showing the ads where they were supposed to, Idk why they are trying so hard instead of doing more of that.
I was of the impression that a farcically low number of people even used adblockers in the first place.
As of 2021 it was roughly a third of people using them. No idea what it is now.
32.8% of people worldwide use them at least sometimes. China stays winning the world’s highest adblocker use (total and per capita) at 40%
Not really. If you know what you’re doing video stream manipulation is easy. They might just do the Webm concat thing (two webm files put together is just a single webm)
Not just that, these big tech companies have some genuine black magic fuckery with their tech it’s unbelievable. I about lost it when I discovered that Google Drive files are not all copied 1:1 in their datastores. Same runs of 0s and 1s in your file and other files in anybody’s Google Drive files are stored only once, and so when you retrieve the file, it actually recreates your uploaded file with a bunch of pointers in sequence to all these different data locations
Yeah why would they store files directly? And its probably more fault tolerant.
Yeah, fuck Google forever of course, but it sounds like a pretty smart way to perform deduplication.
deleted by creator
chunkservers, chunkpointers, chunkhandles, pulls up bfi risk chart for cats the chunkchart