Mastodon Migration
Starter Packs, Follow Packs and all that...
So, let's talk about #StarterPacks, or as this account has implement in a kind of Mastodon fiddly way #FollowPacks (https://mastodon.online/@mastodonmigration/113498826844243920) Directory here >>> @mastodonmigration.wordpress.com
Apparently, there are a number of other efforts here that promise to make a cleaner more integrated solution for following groups of accounts.
What are some lessons learned so far and what should we be thinking about?
1/n
Mastodon Migration
First the positives. People really like them. Particularly due to the absence of an algorithm (thankfully!), packs help newcomers get going and veterans find people in they may have missed. So all good, right?
Not so fast. Like any new thing that has promise, there is also the need to be careful. Dynamite is a wonderful tool for road building, but...
One of the good things about Mastodon is we try to be deliberate about dangers, and this an area we should pay attention.
2/n
Mastodon Migration
So what are the concerns? And shout out to @thenexusofprivacy for flagging some of these:
1. Opt-In vs. Opt-Out vs. Tough Luck
2. Who curates the list and what are the criteria?
3. If the pack includes bridged accounts to Threads and/or Bluesky does following it expose users data to these companies?
4. Can the pack list be used as a vector for abuse?
5. Can packs themselves include 'hidden' toxic accounts so people are fooled into following them.
3/n
Mastodon Migration
1. Opt-In vs. Opt-Out vs. Tough Luck
This is a tough one. Clearly Opt-In is safest, but would it be such an impediment that it would kill the feature before it can even get going? Opt-Out seems to be a minimum requirement and it is the technique that #FollowPacks has adopted. Tough Luck should simple not be OK.
Would suggest when a pack is published, and again in any directory, users should be advised that they can ask to be removed and given a clear means to request removal.
4/n
Mastodon Migration
2. Who curates the list and what are the criteria?
Packs floating around without any accountability should not be tolerated. The creator of the pack should stay attached to it via a recourse address, and be responsible for curing any problems that may come up. The criteria they employed in selecting the members of the pack should be clearly specified when announcing the pack and again in any directory along with the recourse address.
5/n
Mastodon Migration
3. If the pack includes bridged accounts to Threads and/or Bluesky does following it expose users data to these companies?
Some might not care, but others might not want to bridge to companies like Meta whose privacy policy says they can use it for targeting ads and training AI.
Any pack that includes addresses outside the Fediverse should include a notice eg. "NOTICE: Following these Threads accounts could lead to your information being shared with Meta as per their privacy policy."
6/n
Mastodon Migration
4. Can the pack list be used as a vector for abuse?
This is a tough one. Certainly these packs have the potential to be used by those who would target a particular group of people. The entire purpose of the pack is to help people find others with similar interests. But rogue actors can also use packs to find people to abuse.
It is hard to say how to combat this other than to continue to strengthen our communities and moderation activities, but need to acknowledge the threat.
Ideas?
7/n
Mastodon Migration
5. Can packs themselves include 'hidden' toxic accounts so people are fooled into following them.
This is now a known issue on #Bluesky, and should be taken really seriously. The proliferation of packs without accountability can easily lead to 'malware' or 'trojan' packs that contain accounts that expose unwitting users to follow people they would never want to deal with.
This underscores that the above mentioned accountability and recourse requirements for any pack.
8/n
Mastodon Migration
So, can we have nice things, or are we so concerned about these kinds of curation and privacy issues that we are unable to move forward?
Would argue that we can move forward while being aware of things that can go wrong and concerned about safety in the implementation of both technological and social progress. If we anticipate things that can go wrong and try to be intentional about our progress, it will be much better than dealing with all the fallout when things go off the rails.
9/n
Mastodon Migration
Thanks for sticking with this. Have learned a lot in the last few days implementing these fledgling #FollowPack lists. If you want to give them a try check out this post: https://mastodon.online/@mastodonmigration/113498826844243920 and the directory on this account's 'FediBlog' @mastodonmigration.wordpress.com .
Any thoughts or concerns relating to this implementation or packs in general? Suggestions for packs or want to collaborate to make one? Please feel free to reply here.
10/10
aGate to: peace and rest
@mastodonmigration about hidden toxic accounts specifically : maybe following a pack could take into account your own instance moderation rules, limitations, restrictions, etc.
For instance : if you follow a pack with a hidden but known bad actor that is blocked or suspended on your server or by your mutuals, maybe it's just okay to drop it and not follow it.
Similarly: if you follow a pack with a hidden but unknown bad actor, we could keep a track that people followed through a pack were followed automatically, and simply stop the follow on behalf of pack subscribers without their manual intervention, when the actor in question reveals himself and starts getting blocked more widely.
Another way to put it would be: why not have follow packs merged with blocklists to create and update semi-automatically the pack for all the people using it?
GNU social JP is a social network, courtesy of GNU social JP管理人. It runs on GNU social, version 2.0.2-dev, available under the GNU Affero General Public License.
All GNU social JP content and data are available under the Creative Commons Attribution 3.0 license.