It is disingenuous to claim that decentralization does not offer privacy advantage due to public posts being... public. It is obvious that someone does not have an expectation of privacy for...
It is disingenuous to claim that decentralization does not offer privacy advantage due to public posts being... public. It is obvious that someone does not have an expectation of privacy for something they post publicly. Decentralization does have privacy advantages that are completely ignored here such as the lack of 3rd party trackers.
Siloing metadata into your chosen home servers prevents large actors from amalgamating multiple sources of metadata that you may not want connected.
Having choices in which server you want allows you freedom of choice in privacy policies that you don't get in an ecosystem of vendor lock-in.
No fediverse sites to my knowledge have advertisement based business models. The donation based models that the author decries are actually a privacy friendly model desired by the users of the ecosystem.
While I have accounts on a few instances, I'm not too familiar with Mastodon, because I dislike short-form social media in general, and public-post social media. However, I feel like they do bring...
While I have accounts on a few instances, I'm not too familiar with Mastodon, because I dislike short-form social media in general, and public-post social media. However, I feel like they do bring up what seem like potentially valid points, and claiming that they are being disingenuous seems overly harsh.
It is disingenuous to claim that decentralization does not offer privacy advantage due to public posts being... public. It is obvious that someone does not have an expectation of privacy for something they post publicly. Decentralization does have privacy advantages that are completely ignored here such as the lack of 3rd party tracker.
I think the concern here is less with posts, and more with metadata.
The authors point out that followers and following lists are public by default, and can be used to build social graphs. Unless I'm not aware of newer developments, the I think way following is implemented means that even selecting the hard-to-find option to hide these lists on your own profile will not avoid the problem of your social graph being constructed in an automated way, as there's no way for you to prevent yourself from being listed in the follower and following lists of other users. It does not appear that there is any way to follow a user without being linked to that user publicly.
Siloing metadata into your chosen home servers prevents large actors from amalgamating multiple sources of metadata that you may not want connected.
How? Wouldn't large actors just pull all the publicly available metadata from all home servers? Also...
Decentralization does have privacy advantages that are completely ignored here such as the lack of 3rd party trackers.
How does decentralization avoid 3rd party trackers? It avoids the addition of a single 3rd party tracker that can track everything, but there would seem to be nothing preventing a server from adding them. You also need to worry about privacy policies of every server you interact with. What metadata are they collecting, and what are they doing with it? If they don't comply with GDPR, how can you even know? What would prevent a large actor from just buying up data from malicious admins of a large number of servers, and combining these? While this data set would have gaps, it could have significantly more depth than what some other service might have if some servers are set up with significantly more tracking.
The article makes a declarative statement It is true that it is not a all encompassing privacy feature for all threat models, however it is false to assert that it is not beneficial for some...
The article makes a declarative statement
Decentralization still isn’t a privacy feature
It is true that it is not a all encompassing privacy feature for all threat models, however it is false to assert that it is not beneficial for some people's use cases.
There is a considerable amount of metadata tracked by social media sites now that is not publically available. Everything from how long you looked at a post, how many times you visited a profile, what you typed but did not send, your outgoing clicks, your mouse hovers that you did not click... it goes on.
None of this can be data-mined from mastodon and is not even tracked with the official build.
If the information that is publicly available on mastodon is too much for you personally then it is fair that you may not want to use Mastodon. However that isn't the argument made in the article small, disclaimer withstanding, and is even less representative of the fediverse in general.
How does decentralization avoid 3rd party trackers? It avoids the addition of a single 3rd party tracker that can track everything
You have answered your own question. Any random blog likely already has a facebook tracker, google analytics etc, however I doubt that you will ever see a blog with trackers for the thousands of different fediverse servers.
As to the possibility that any given fediverse server could choose to add anti-features to their site; you are free to use a different server that does not do this, in fact servers that do not do this constitute the vast majority of all fediverse servers now and probably into the future provided that donation models continue to be viable.
What would prevent a large actor from just buying up data from malicious admins of a large number of servers
Impossible? No. But the more servers there are the more difficult this is to do. Who do you trust more to not sell your data? The VC funded startup?, or a server maintained by privacy advocates, or a community group... or even yourself?
In any event these potential threats you raise are hypotheticals in the fediverse while already being standard practice in centralised websites. This is already a notable privacy improvement you can obtain right now, even if it could theoretically disappear one day.
This is kind of a weird article, and really not about the Fediverse. This (company? magazine publisher?) briefly mentions two "issues" with the Mastodon service - inadequate post migration...
This is kind of a weird article, and really not about the Fediverse. This (company? magazine publisher?) briefly mentions two "issues" with the Mastodon service - inadequate post migration capabilities and that boosts and favorites are cleared upon edits1 - but spend the remaining 90% of the article complaining about societal problems inherent to any form of social media or online communication itself.
1 And, how exactly else do they think this would work? If I boost someone else's post, and the poster edits it to something malicious, should it just stay in my likes or on my timeline?
I'm inclined to (very slightly) agree, but what I find hilarious is how their issues with the Fediverse are pretty much most of the reasons why I'm less and less active on Reddit, their networking...
I'm inclined to (very slightly) agree, but what I find hilarious is how their issues with the Fediverse are pretty much most of the reasons why I'm less and less active on Reddit, their networking platform of choice, anymore.
To an extent this is true, and is something that the mastodon developers are mindful of. One example is that Mastodon deliberately does not allow "quote tweets" to avoid the twitter behavior of...
To an extent this is true, and is something that the mastodon developers are mindful of.
One example is that Mastodon deliberately does not allow "quote tweets" to avoid the twitter behavior of dog piling and dunking on others.
Another example would be that not migrating people's responses to edited content discourages trolling behavior of saying something agreeable then modifying to something objectionable in order to create the appearance of said people agreeing with something objectionable. Mastodon does not consider enabling this behavior a worth while cost to allow preservation of metrics in the way the author wants.
The societal behaviors that the author is bothered by are that people are able to criticize them and the services they recommend and that there is not a culture of reverence for engagement metrics. The author simply is inherently at odds with the values of mastodon users.
It's not any social media or online communication though. Twitter made some specific design decisions that make these tendencies worse. Mastodon lifted those paradigms wholesale, so it replicates...
but spend the remaining 90% of the article complaining about societal problems inherent to any form of social media or online communication itself.
It's not any social media or online communication though. Twitter made some specific design decisions that make these tendencies worse. Mastodon lifted those paradigms wholesale, so it replicates many of the same problems. Any social media platform that is focused on broadcasting marketing slogans as the primary interaction paradigm is going to run into these problems. It's toxic to community building and creates an environment conducive to groupthink, bullying, and disinformation. On a site like Twitter, there are enough journalists to somewhat militate against it by acting as a (very inadequate) safety valve against viral disinfo. But Mastodon doesn't have the scale to do that and, in fact, it's designed not to have that scale.
Facebook and Reddit have the same problems for similar reasons. I think any alternative model of social media is going to have to function very differently from how social media, as we think of it, functions today. I expect it would need to look a lot more like the old world of Xanga and LiveJournal and a lot less like timelines designed to virally spread atomized bits of content.
There are some technical differences. Boosts and retweets don’t work quite the same. I assume the designers of Mastodon and ActivityPub were not trying to repeat all of Twitter’s mistakes, but...
There are some technical differences. Boosts and retweets don’t work quite the same. I assume the designers of Mastodon and ActivityPub were not trying to repeat all of Twitter’s mistakes, but rather to improve on Twitter.
But the question is whether the changes are enough to avoid Twitter’s problems. The linked article says it’s not enough.
This is the problem. The Fediverse is a connected system of social media platforms, but it is just that, social media platforms. Every site has the same issues, be it Facebook, Reddit or even Tildes.
If you want to be an active part of the Fediverse, be prepared for the same problems as on any other social network. We decided to leave the Fediverse for now.
This is the problem. The Fediverse is a connected system of social media platforms, but it is just that, social media platforms. Every site has the same issues, be it Facebook, Reddit or even Tildes.
I'm glad this was posted. My extremely prototype community framework addresses most of these shortcomings. The catch is that self-hosting is required, but I offer "hatchling" hosting for friends...
I'm glad this was posted. My extremely prototype community framework addresses most of these shortcomings.
The catch is that self-hosting is required, but I offer "hatchling" hosting for friends initially.
It is disingenuous to claim that decentralization does not offer privacy advantage due to public posts being... public. It is obvious that someone does not have an expectation of privacy for something they post publicly. Decentralization does have privacy advantages that are completely ignored here such as the lack of 3rd party trackers.
Siloing metadata into your chosen home servers prevents large actors from amalgamating multiple sources of metadata that you may not want connected.
Having choices in which server you want allows you freedom of choice in privacy policies that you don't get in an ecosystem of vendor lock-in.
No fediverse sites to my knowledge have advertisement based business models. The donation based models that the author decries are actually a privacy friendly model desired by the users of the ecosystem.
While I have accounts on a few instances, I'm not too familiar with Mastodon, because I dislike short-form social media in general, and public-post social media. However, I feel like they do bring up what seem like potentially valid points, and claiming that they are being disingenuous seems overly harsh.
I think the concern here is less with posts, and more with metadata.
The authors point out that followers and following lists are public by default, and can be used to build social graphs. Unless I'm not aware of newer developments, the I think way following is implemented means that even selecting the hard-to-find option to hide these lists on your own profile will not avoid the problem of your social graph being constructed in an automated way, as there's no way for you to prevent yourself from being listed in the follower and following lists of other users. It does not appear that there is any way to follow a user without being linked to that user publicly.
How? Wouldn't large actors just pull all the publicly available metadata from all home servers? Also...
How does decentralization avoid 3rd party trackers? It avoids the addition of a single 3rd party tracker that can track everything, but there would seem to be nothing preventing a server from adding them. You also need to worry about privacy policies of every server you interact with. What metadata are they collecting, and what are they doing with it? If they don't comply with GDPR, how can you even know? What would prevent a large actor from just buying up data from malicious admins of a large number of servers, and combining these? While this data set would have gaps, it could have significantly more depth than what some other service might have if some servers are set up with significantly more tracking.
The article makes a declarative statement
It is true that it is not a all encompassing privacy feature for all threat models, however it is false to assert that it is not beneficial for some people's use cases.
There is a considerable amount of metadata tracked by social media sites now that is not publically available. Everything from how long you looked at a post, how many times you visited a profile, what you typed but did not send, your outgoing clicks, your mouse hovers that you did not click... it goes on.
None of this can be data-mined from mastodon and is not even tracked with the official build.
If the information that is publicly available on mastodon is too much for you personally then it is fair that you may not want to use Mastodon. However that isn't the argument made in the article small, disclaimer withstanding, and is even less representative of the fediverse in general.
You have answered your own question. Any random blog likely already has a facebook tracker, google analytics etc, however I doubt that you will ever see a blog with trackers for the thousands of different fediverse servers.
As to the possibility that any given fediverse server could choose to add anti-features to their site; you are free to use a different server that does not do this, in fact servers that do not do this constitute the vast majority of all fediverse servers now and probably into the future provided that donation models continue to be viable.
Impossible? No. But the more servers there are the more difficult this is to do. Who do you trust more to not sell your data? The VC funded startup?, or a server maintained by privacy advocates, or a community group... or even yourself?
In any event these potential threats you raise are hypotheticals in the fediverse while already being standard practice in centralised websites. This is already a notable privacy improvement you can obtain right now, even if it could theoretically disappear one day.
This is kind of a weird article, and really not about the Fediverse. This (company? magazine publisher?) briefly mentions two "issues" with the Mastodon service - inadequate post migration capabilities and that boosts and favorites are cleared upon edits1 - but spend the remaining 90% of the article complaining about societal problems inherent to any form of social media or online communication itself.
1 And, how exactly else do they think this would work? If I boost someone else's post, and the poster edits it to something malicious, should it just stay in my likes or on my timeline?
Societal problems and medium architecture are deeply intertwined.
The way UI and software are architected shapes the flow of discussion.
I'm inclined to (very slightly) agree, but what I find hilarious is how their issues with the Fediverse are pretty much most of the reasons why I'm less and less active on Reddit, their networking platform of choice, anymore.
To an extent this is true, and is something that the mastodon developers are mindful of.
One example is that Mastodon deliberately does not allow "quote tweets" to avoid the twitter behavior of dog piling and dunking on others.
Another example would be that not migrating people's responses to edited content discourages trolling behavior of saying something agreeable then modifying to something objectionable in order to create the appearance of said people agreeing with something objectionable. Mastodon does not consider enabling this behavior a worth while cost to allow preservation of metrics in the way the author wants.
The societal behaviors that the author is bothered by are that people are able to criticize them and the services they recommend and that there is not a culture of reverence for engagement metrics. The author simply is inherently at odds with the values of mastodon users.
It's not any social media or online communication though. Twitter made some specific design decisions that make these tendencies worse. Mastodon lifted those paradigms wholesale, so it replicates many of the same problems. Any social media platform that is focused on broadcasting marketing slogans as the primary interaction paradigm is going to run into these problems. It's toxic to community building and creates an environment conducive to groupthink, bullying, and disinformation. On a site like Twitter, there are enough journalists to somewhat militate against it by acting as a (very inadequate) safety valve against viral disinfo. But Mastodon doesn't have the scale to do that and, in fact, it's designed not to have that scale.
Facebook and Reddit have the same problems for similar reasons. I think any alternative model of social media is going to have to function very differently from how social media, as we think of it, functions today. I expect it would need to look a lot more like the old world of Xanga and LiveJournal and a lot less like timelines designed to virally spread atomized bits of content.
There are some technical differences. Boosts and retweets don’t work quite the same. I assume the designers of Mastodon and ActivityPub were not trying to repeat all of Twitter’s mistakes, but rather to improve on Twitter.
But the question is whether the changes are enough to avoid Twitter’s problems. The linked article says it’s not enough.
Yeah. I don’t think just improving on Twitter is enough. I think the idea is foundationally faulty.
This is the problem. The Fediverse is a connected system of social media platforms, but it is just that, social media platforms. Every site has the same issues, be it Facebook, Reddit or even Tildes.
I'm glad this was posted. My extremely prototype community framework addresses most of these shortcomings.
The catch is that self-hosting is required, but I offer "hatchling" hosting for friends initially.
Could you elaborate or point me to some code? I am also interested in these problems.
Yes, feel free to check out the links in my profile. :)
Erm.. thanx, and also eww? :)