The reason Mastodon (and the wider Fediverse that it is part of) doesn't go down is because all its servers are totally independent. One Fedi server might go down, but that doesn't affect the other servers at all.
That's one of the many advantages of decentralisation, that there is no central point of failure.
Centralised networks go down because they depend on one central organisation's infrastructure.
reshared this
always tired
in reply to Fedi.Tips • • •Fedi.Tips
in reply to always tired • • •@project1enigma
One server might go down, but the whole network doesn't go down.
I'm not sure what you mean by losing an identity? Servers that go down usually come back up later and you can sign in as normal.
always tired
in reply to Fedi.Tips • • •Fedi.Tips
in reply to always tired • • •Ah yeah that's true, but that's why it's best to use servers that promise to give three months warning of a closedown (all the servers listed at fedi.garden and joinmastodon.org/servers promise three months warning).
Sudden closedowns are a risk for anything on the internet to be honest, which is why it's best to encourage a diverse range of online providers.
Fedi.Garden – An easy way to join Mastodon and the Fediverse
FediGarden (Fedi.Garden - An easy way to join Mastodon and the Fediverse)Carlos Solís
in reply to always tired • • •Brain Bug
in reply to Fedi.Tips • • •Leti Uchiha
in reply to Fedi.Tips • • •Fedi.Tips
in reply to Leti Uchiha • • •Svavar the Neurospicy
in reply to Fedi.Tips • • •Fedi.Tips
in reply to Svavar the Neurospicy • • •@svavar
Yeah, that's a fair point. If many servers rely on a single provider's centralised service that is a form of centralisation.
Svavar the Neurospicy
in reply to Fedi.Tips • • •JoanlyFans 😷
in reply to Fedi.Tips • • •Max - Poliverso 🇪🇺🇮🇹
in reply to Fedi.Tips • •@Fedi.Tips
The fact Facebook is one social network, owned by one company, doesn't mean everything is hosted on one single server that could go down and block the whole infrastructure.
So I wouldn't say that decentralisation is better because there are no single points of failure.
Dr. Juande Santander-Vela likes this.
Nate Gullion 🏳️🌈
in reply to Max - Poliverso 🇪🇺🇮🇹 • • •Max - Poliverso 🇪🇺🇮🇹
in reply to Nate Gullion 🏳️🌈 • •@Nate Gullion 🏳️🌈
And I'm simply pointing out that centralised infrastructure could be designed (and the big ones actually are designed) using redundancy so as not to have any single point of failure.
If Meta central routing is build on a redundant architecture it doesn't have any single point of failure.
Cat 🐈🥗 (D.Burch)
in reply to Fedi.Tips • • •