Reading up on the issue, I understand that openblas is superior to blas in performance. So if I have blas I should replace it with blas-openblas. Am I correct in this understanding?

I see several comments talking about this being a wrong decision, or Beehaw needing to change its attitude etc. I think these opinions come from a misunderstanding of the fundamentals of federation. Federation is not about all the instances coming together to cater to our needs. It’s about each instance doing its own thing, and communities will form around the ones that cater to them. In other words, we don’t need Beehaw to budge on its decision, we need to build the community we want without Beehaw, while Beehaw caters to the users who aren’t in this with us.

Meh I don’t use it enough to warrant the effort of benchmarking. I just wanted a really general “if you don’t know what you want, your should probably use X”, and it appears that I should just use blas-openblas.

I’m wondering if a fundamental separation of users from content would be the best way to go. Apart from simplicity, I’m not sure I see any benefit to the current model of instances holding both our identities and the content we want to browse.

Of course these games are not going to last forever, but every day that subs are flooded with John Oliver is an extra day for people to learn about the drama and to consider moving to another platform. For subs that were forced open, it was either this or already going back to normal.

From what I understand, bubblewrap is supposed be configured by passing flags from the command line. It seems that the way to “configure” bubblewrap is to create wrapper scripts. For example make /usr/local/bin with the following contents

#!/usr/bin/bash
bwrap --flags-and "arguments" steam

As it’s not very practical to figure out a good sandbox from scratch for each and every program you use, you probably want to find scripts from other users or tools that build on top of bubblewrap and are bundled with profiles. The wiki article has examples of both.

I’m not sure federation is that important on sites that aren’t built around socializing. I think it is sufficient for a wiki to provide a good export mechanism so that it can be archived or mirrored by others.

Well, realistically there is a good chance that this will turn out just fine business-wise. They don’t care if they lose some engagement or if the quality goes to shit. It’s all good, as long as it makes some money.

In my opinion, this sort of model should be considered anti-competitive. It has become apparent that these services operate on a model where they offer a service that is too good to be true in order to kill the competition, and then they switch to their actual profitable business plan. If you think about it, peertube is a much more sensible economical model with its federation and p2p streaming. But nobody has ever cared about it because huge tech giants offer hosting & bandwith “for free”. The evil part of youtube is not the ads, its the fact that it allowed us to bypass them long enough for the entire planet to become dependent on it.

Ok, but what would the reputation score be based on that can’t be manipulated or faked?

I’m assuming this is a joke based on the Futurama references you used, but just to be clear for everyone: this won’t work because it simply moves the problem one step further. How do you prevent bots from upvoting other bots to build a reputation?