I actually think that the overwhelming majority of blocking should be done by individuals curating their own timelines. I am sensitive to the effect on the Fediverse as a whole, especially as we've already been through this.
Even the original #bifurcation (when the largest instance at the time, Identica, severed communication with #StatusNet / #GNUsocial & #OStatus and switched to the #Pump.io protocol and software) and the subsequent #ActivityPub - #OStatus split have caused untold breakage. I've seen AP-side devs, admins, users patting themselves on the back while commiserating about brokenness that is built into the protocol itself or at least its common implementations.
I have also seen people telling other people to create "alts" on various instances, so that their posts can reach all of their intended contacts. Not for resilience against instance shutdowns or separating by posts and recipients by topics and interests (which is what groups and Diaspora style Aspects / GPlus style Circles are for), but because #blockwars prevents posts and members from one instance to be seen on certain others.
For the record, I think that instance governance is something that Mastodon should include in its instances.social instance-picker, along with instances' topical foci. People should have a way to see what they're agreeing to (and what the alternatives are) before the sign up.
In other words, it isn't my way or the highway so much as it is making it possible to know what one is getting into. I am certain that there are (or were) instances with democratically chosen rules. I also believe that we're not doing the people who use an instance any favor by not making it possible for them to contribute to the financing and administration of the instance. If you're paying all the costs and doing all the work to maintain and moderate the instance, it is difficult to let an election institute a policy that you disagree with. (I've started to really disagree with the idea of individuals hosting public instances wholly out of their own financial and time resources. Besides the "truck factor", it is much easier to keep an instance going if everything was already handled by a team and at least partly member supported.)
On the other hand, if the instance encourages those in its membership who can do so to participate in keeping it going, then it is perfectly reasonable to expect the admin team to carry out the decisions voted by the membership. I do realize that not everyone can contribute funds, nor can everyone do the technical labor ... but as @simsa04 will remember, things like writing documentation, contributing in discussions about improving the software, designing and implementing themes, and even marketing-type tasks such as creating a logo and a favicon or promoting the instance to people outside the #Fediverse are beneficial.
@dragnucs I disagree. Once your federated network survives one severe #bifurcation (the #Pump.io split), you know it can survive others. The #ActivityPub addition is still somewhere in the process, but I doubt that defederating from the #OStstus part of the network would speed it up. @gargron @heluecht
the scale of the companies that are being built, the services being built, the wealth that is being created is phenomenal. Chinese Internet is a greater percentage of the GDP of China, which is a big number, than the same percentage of the US, which is also a big number [https://boingboing.net/2018/10/10/july-18-leak.html] #bifurcation
To be fair, the !StatusNet network that Evan was running became unmanageable -- it cost too much to run, and I think Evan was funding most of it himself (although there were/are a number of paying #StatusNet customers). Evan developed #PumpIO to reduce the number of servers needed to run a federated network, and purposely kept the UI to a minimum to encourage federation. Sadly, that didn't work. Porting identi.ca from StatusNet to PumpIO was intended to introduce people to PumpIO as well as reduce Evan's costs. That partly worked; identi.ca is alive and well as a community, although much reduced from its glory days around 2013. But the #bifurcation did spawn a large number of new StatusNet / !GNUsocial instances, so that was a good thing too. But you're right in that PumpIO never gained widespread traction, the proof of which is in its lack of continued development. In that respect #GNUsocial and !OStatus are more successful than PumpIO
@clacke True. It hasn't been updated for the post #bifurcation fediverse, where larger character limits are the norm instead of the exception (and where some instances use #Markdown or #Textile to enable formatted posts).