Unlike others, I’d expect a signicifant decline of posts as not only many users but also loads of communities would be lost. That’s why from my perspective users and communities should be evenly distributed across instances.
On top of that, there should be a feature to move entire accounts or communities to other instances. That way a community including all its members could just be migrated before a major shutdown.
Similarly, I think it would be a huge disturbance for the email system and possibly the entire internet world-wide if Gmail went down next month even though there are in theory plenty of alternative providers. Or supermarkets. If the IT of Walmart, Visa/MasterCard, Amazon AWS, Microsoft etc. have an outage it always has huge impact.
Lemmy as a whole isn’t that big and far from being critical infrastructure but we all want it to grow we should bear in mind that huge central services are always more risky than many small services.
That’s an amazing idea from my perspective which would really help the fediverse in general. Your proposal would solve not only the issue of instances disappearing completely but also temporary outages. For example Feddit.de as one of the biggest German instances had several technical issues lately. And as instances are typically administered by volunteers during their freetime that’s totally fine, accepted and expected. Such a failover concept would heavily reduce the pressure on their shoulders!
Unlike others, I’d expect a signicifant decline of posts as not only many users but also loads of communities would be lost. That’s why from my perspective users and communities should be evenly distributed across instances.
On top of that, there should be a feature to move entire accounts or communities to other instances. That way a community including all its members could just be migrated before a major shutdown.
Similarly, I think it would be a huge disturbance for the email system and possibly the entire internet world-wide if Gmail went down next month even though there are in theory plenty of alternative providers. Or supermarkets. If the IT of Walmart, Visa/MasterCard, Amazon AWS, Microsoft etc. have an outage it always has huge impact.
Lemmy as a whole isn’t that big and far from being critical infrastructure but we all want it to grow we should bear in mind that huge central services are always more risky than many small services.
It’d be cool if you could specify a backup mirror community, so there could be an immediate failsafe.
That’s an amazing idea from my perspective which would really help the fediverse in general. Your proposal would solve not only the issue of instances disappearing completely but also temporary outages. For example Feddit.de as one of the biggest German instances had several technical issues lately. And as instances are typically administered by volunteers during their freetime that’s totally fine, accepted and expected. Such a failover concept would heavily reduce the pressure on their shoulders!