@daniel As discussed above, the issue here is not whether some instances temporarily blocked mastodon.social while the attack was going on. The implication was that people on mastodon.social were at risk for being "defederated" and that they should change instances to avoid this. This was a misleading because peoples accounts on mastodon.social are not at risk of being isolated.

1/

@daniel If what had been presented is that instances under attack may be temporarily defederate, and that mastodon.social has recently experienced hacker attacks, which resulted in a few instances blocking them for a short period of time while the attack was being addressed that would not have been as frightening.

In fact, that would be a good description of the system working as intended. Isolating the area under attack while the problem was addressed.

2/

@daniel The implication is that mastodon.social is particularly vulnerable to such attacks because it is big, and it is too big to be successfully moderated. This simply does not make sense. Any instance with open admissions can easily be subject to such attacks and while it is certainly difficult to moderate larger instances it is just a matter of allocating sufficient resources. In this case the attack was dealt with pretty swiftly, so it is not a good example of a breakdown.

3/

@daniel Any instance, big, medium or small that is targeted in this manner will need to do the same thing. It will need to address the attack and while it is doing so, it may be defederated. Hopefully, that defederation will only be temporary as was the case with the latest mastodon.social incident. One could say that smaller instances have the risk that if they don't respond quickly enough it could trigger more permanent isolations. These are all issues we should be discussing.

4/