The protocol has a notion of moderation accounts, sometimes called "labelers" because they can apply labels to your account. Users generally [1] subscribe to the moderation that they want to see. They can then set those labels to make the application to "show", "warn" (collapse by default, click to expand), or "hide" based on the labels.
1: People using the bluesky app cannot unsubscribe from the bluesky moderation service, but that is a policy choice of the bluesky app, not a protocol level choice, other clients can do as they choose.
I'm no expert in atproto but it seems to be far worse than that. Apparently, Bluesky applies country-specific labellers automatically based on your IP address.
If you register a Bluesky account from Germany, your account is assigned the German moderation labeler with no option to opt out. As soon as I noticed, I created a new account using a US IP address. This fixed it.
While signing up for moderation sounds very attractive to me, Bluesky's whole layered moderation approach seems designed to maximize algorithmic censorship.
Compare that to, for example, a mastodon instance (or a forum like HN) where you participate because you align with the general moderation approach cultivated by and within the community.
Yes, it is specific to Bluesky. It is not dependent on the official app, though: Both the app and the website are subject to this kind of censorship.
There seem to be loopholes [1] through some third-party apps, but that does not change much about the overall problem. Defaults matter a lot.
I have a hard time understanding why the Bluesky authors came up with an idea like that in the first place. Not to mention actually creating it. The internet (or social media) probably does not benefit from even more granular censorship tools for governments all over the world.
I'm not sure if this is some kind of tactic, but both of these claims seem trivially wrong.
Bluesky is available in both main app stores and on the web. There is no merit in denying this. There are often differences in content restrictions between what app stores enforce and what service providers enable, but one cannot really say, "Just don't use the app," if by "app" they mean every common way of accessing the service.
Also, I think the protocol does not matter much if the user experience of almost every user is entirely controlled by a single provider. In such a case, the protocol might as well be entirely absent.
It is difficult to tell whether there is some substance here, since you do not explain (or source) anything.
The protocol has a notion of moderation accounts, sometimes called "labelers" because they can apply labels to your account. Users generally [1] subscribe to the moderation that they want to see. They can then set those labels to make the application to "show", "warn" (collapse by default, click to expand), or "hide" based on the labels.
1: People using the bluesky app cannot unsubscribe from the bluesky moderation service, but that is a policy choice of the bluesky app, not a protocol level choice, other clients can do as they choose.