@anderspuck @SuiGeneris3722 @randahl a better solution is showing edit history with diffs the way Wikipedia does it.
On the other hand, people now leave typos deliberately uncorrected to prove they are human and not AI. Except now AIs deliberately make typos to mimic this.
@fazalmajid Yes and no. 90 percent of people would never dive into the edit history, and also many people will never discover that something they've boosted has been changed.
This means that anyone with a cat video that's gone viral can change the text and get reach with whatever message they want.
At least Mastodon will send a notification to anyone who interacted with a post if it is edited. That said, the protocol does not enforce this -- you could create an ActivityPub implementation that didn't notify boosters of edits made after their boost. Their subscribers wouldn't necessarily see the edit, either.
@anderspuck @fazalmajid
@osma @fazalmajid Exactly. And big channels will often suffer from notification overflow, so it's easy to miss that there was an important one among them.
Notifications are relatively easy to filter though, if you'd be concerned about missing a notification of an edit to a boosted post. Mastodon's own apps are nowhere near the state of the art for this (or any other usability feature).
@anderspuck @fazalmajid
It would also be trivial to set up an automation that retracts the boost from any post which is edited post-boost.
@anderspuck @fazalmajid
@osma Yes, it can work well for most people, but if you have an account with thousands of notifications every day, it gets hard to keep up. I think it's fine to just set a time limit for people to correct typos and then lock the post. That also solves the issue.
That's the nice thing about not defining the policy on the network level but instead provide notifications. Each actor may decide their own.
@anderspuck
@osma Not sure I understand. What i mean is that if a post is boosted on Bluesky by someone like Mike Masnick, then it reaches his 200k+ subscribers. It's not reasonable to expect him to stay constantly alert to changes, if one of his reposts is suddenly turned into a crypto ad.
No, it wouldn't be. But it would be easy for him to automate it, to retract boosts of any post edited over 15 minutes after his boosting. That is, if Bluesky supported edits (it apparently does support displaying edits bridged from Fedi, btw!)
@anderspuck
This is an interesting discussion. Back before Mastodon had edits the concern was raised of people getting boosted and then substantially changing the post. Edit history and notification resulted from these concerns.
In the years since editing was provided on Mastodon, have not seen even one single accusation of editing being used to change the meaning in this way. None. Zero. So seems like this is really not an issue with the present safeguards.
I have neither. That's not to say it could not happen, of course: fedi is small and the attempts to game the network have not yet been very established. Boost-gaming is probably not the weakest area though - easy to mitigate, compared to many other issues related to weak moderation tools. Reply spamming, for example, is more prevalent.
@mastodonmigration @anderspuck
Agreed. Feel like it really comes down to a matter of integrity. If a poster were to do that it would be noticed and they would be called to account. Also, it is just not the way people work, and the utility is fantastic. Other things are more realistic concerns.
@osma I've seen people playing with beta versions of some AT Proto app trying out edits in Bluesky, so I'd be surprised if it doesn't show up soon.
I found out entirely by accident that my toot edits changed the bridged posts, too. Don't know exactly when it started.
@anderspuck