I haven't sensed more anger from fedi towards #bluesky recently.
-
Yes. Like that too. Absolutely.
These are precisely the conversations we should be having. Fediverse decentralization has lots of room for improvement.
It's confounding that AT Protocol users are not demanding more diversification from Bluesky PBC. Blacksky is a great first step, but until a meaningful percentage are on independent nodes the entire network is at elevated risk.
We're heading into even more challenging times, and should be thinking more strategically.
@mastodonmigration @thisismissem Tbh I generally agree with the PoV I've read a few times on Bluesky, that making Bluesky/ATProto more decentralized isn't something that is the responsibility of Bluesky PBC, but rather the community… Kind of by definition they can't work on setting up servers which aren't theirs. Other people/orgs need to do that. What they can do is only try to make it easier, maybe change the incentives etc.
-
@mastodonmigration @thisismissem Tbh I generally agree with the PoV I've read a few times on Bluesky, that making Bluesky/ATProto more decentralized isn't something that is the responsibility of Bluesky PBC, but rather the community… Kind of by definition they can't work on setting up servers which aren't theirs. Other people/orgs need to do that. What they can do is only try to make it easier, maybe change the incentives etc.
@mastodonmigration @thisismissem So we can push them to make better docs, more configurable and easier to install distributions of the software, add the missing pieces and so on, but we can't really push Bluesky to create alternative non-Bluesky relays, appviews, PDS hosts etc. themselves, this is on us.
-
@mastodonmigration @mackuba right, but arguing about protocol superiority isn't servicing this cause for anyone, and it divides us when we should be working together to tackle these risks
!!!It can both be true that the protocols have interesting differences and the networks would be stronger if they were more decentralized.
Would argue that the decentralization issue is more acute with AT Protocol because of the overwhelming dominance of Bluesky PBC.
-
@mastodonmigration @thisismissem So we can push them to make better docs, more configurable and easier to install distributions of the software, add the missing pieces and so on, but we can't really push Bluesky to create alternative non-Bluesky relays, appviews, PDS hosts etc. themselves, this is on us.
Kind of hard disagree.
Having been involved in standards efforts over the years, some of which where initiated by single corporations. Can without a doubt say the ones that were successful were spearheaded by companies intent upon fostering independent adoption of the standard, and insuring they were not the sole dominant enterprise using it.
This is very hard to do, and more failed than were successful.
-
!!!
It can both be true that the protocols have interesting differences and the networks would be stronger if they were more decentralized.
Would argue that the decentralization issue is more acute with AT Protocol because of the overwhelming dominance of Bluesky PBC.
@mastodonmigration @thisismissem I think how I understand the key of @thisismissem's call is that this isn't so much about not discussing decentralization and each protocol's problems and so on and pretending there are no faults when there are, but about keeping the discussion civil, having respect for each other, discussing real facts, trying to be well informed and objective when doing so, avoiding [dm]isinformation, and generally not crossing the line from criticism into attacks/fighting
-
Kind of hard disagree.
Having been involved in standards efforts over the years, some of which where initiated by single corporations. Can without a doubt say the ones that were successful were spearheaded by companies intent upon fostering independent adoption of the standard, and insuring they were not the sole dominant enterprise using it.
This is very hard to do, and more failed than were successful.
What this would look like in.practice is all the things you mention. Better doc, distributions etc. But also economic support, promotions and clear metrics celebrating meaningful 'competitive' adoption of the standard.
The reason it is hard to do is because there are always people inside the company who view these adopters as competitors and since they have the lock on the information, they have an unfair advantage.
2/
-
What this would look like in.practice is all the things you mention. Better doc, distributions etc. But also economic support, promotions and clear metrics celebrating meaningful 'competitive' adoption of the standard.
The reason it is hard to do is because there are always people inside the company who view these adopters as competitors and since they have the lock on the information, they have an unfair advantage.
2/
So, what the companies need to do, at the earliest possible moment is spin it all out to an independent standards effort where they do not have any proprietary advantage. Then enforce internally that they follow the standard and do not have their own version. (This is kind of where Mastodon extensions to ActivityPub seems dicey.)
But, as said above, it is really hard. Lots of competing forces and agendas. Get a bunch of VCs in the mix and it's nigh impossible.
-
@mastodonmigration @thisismissem I think how I understand the key of @thisismissem's call is that this isn't so much about not discussing decentralization and each protocol's problems and so on and pretending there are no faults when there are, but about keeping the discussion civil, having respect for each other, discussing real facts, trying to be well informed and objective when doing so, avoiding [dm]isinformation, and generally not crossing the line from criticism into attacks/fighting
!!! The thing that make this tricky is the difficulty many have in distiguishing AT Protocol and Bluesky PBC. But in any case we do ourselves no favors by being uncivil with each other.
-
@mastodonmigration @mackuba as mentioned, there's nothing stopping regulators or prosecutors from going after mastodon or other activitypub software for non-compliance. It's probably quite easy too given most servers are hosted in Germany, a country with one of the oldest age verification laws on the books and a right wing politician with a fetish for regulation enforcing it. Seriously, this guy had an entire AI bot built to help prosecute adults in Germany who posted material against his laws
@thisismissem @mastodonmigration @mackuba most servers aren't hosted in Germany. This says 15.6% are: https://fedi.wrm.sr/
-
@thisismissem @mastodonmigration @mackuba most servers aren't hosted in Germany. This says 15.6% are: https://fedi.wrm.sr/
@ikuturso @mastodonmigration @mackuba I'm pretty sure deeper research showed that many instances that looked US based were actually using CDNs in front and hitting US based POPs
-
@ikuturso @mastodonmigration @mackuba I'm pretty sure deeper research showed that many instances that looked US based were actually using CDNs in front and hitting US based POPs
@ikuturso @mastodonmigration @mackuba for instance, Hachyderm has US based POPs but core infrastructure is all in Germany
-
@ikuturso @mastodonmigration @mackuba for instance, Hachyderm has US based POPs but core infrastructure is all in Germany
@thisismissem @mastodonmigration @mackuba I'm sure that is true to some extent but I also don't think all or even most of the US share is very likely to be in Germany.
If I had to guess a significant chunk (even most) is probably actually in the US and the rest likely more or less divided in similar shares as the overall numbers are among the other countries.
-
@thisismissem @mastodonmigration @mackuba I'm sure that is true to some extent but I also don't think all or even most of the US share is very likely to be in Germany.
If I had to guess a significant chunk (even most) is probably actually in the US and the rest likely more or less divided in similar shares as the overall numbers are among the other countries.
@ikuturso @mastodonmigration @mackuba if you're EU based doing a query from here to each server to determine ASNs might be an interesting research project, maybe you'll get different results?
