I have been wondering for some time whether #selfhosting and #decentralisation also have an advantage over centralised services in terms of the CO2 emissions. This is a very important question if we are to believe the predictions that we will face an enormous growth in demand as #digitalisation progresses.
Are there any studies, findings or evidence to support this?
@tobi I'm sorry that I don't have any studies to back this up, but personally I think its the other way around. Decentralized Networks, if used in the same way and from as many people as the big centralized ones, will likely be more power consuming. The main reason would be IMO the more efficient infrastructure of the big centralized Networks, like the giant data centers of Facebook, Google or Amazon.
We discuss this in German language also here - so you may join the club:
@tobi most servers reside in the same datacenters as the centralised hardware. Why should there be a (big) difference? And if there is one, why should there be a positive one? Big datacenters can massivly optimize there energy consumtion, which you can not do with only a few servers.
@tobi what maybe makes a difference are adblockers or adfree content. But this has per se nothing to do with centralisazuon.
@tobi @tobi There is a distinction to make between federated network, which still required a server and P2P networks which don't and are decentralized.
So I think that P2P networks can reduce the energy needs for public content hosting. It will be interesting to compare the Youtube and peertube comsumption for hosting the same content.
@Nachbarschaft Yes and no. In peertube, you upload a video on a server. The server stores videos and metadata information. But the server is also a #webtorrent peer, and all viewers of a video become also a peer of this video. So the more a video is watched, the more the video is shared. It is the reason why peertube server can be small.
One of the first Mastodon instances, there is no specific topic we're into, just enjoy your time!