There’s obviously a lot of talk about Tumblr being shut down, which does not appear to be happening, but do you think the re-organization will be enough to keep Tumblr running indefinitely? Or is this maybe the start of an eventual shut down of Tumblr? https://www.darnellclayton.com/post/733595566933704704
@darnell Are there any updates on the status of the ActivityPub integration for Tumblr?
@ajsadauskas Yes! Matt Mullenweg answered on #Tumblr: https://www.tumblr.com/photomatt/733546273183006720/youve-mentioned-in-the-past-that-adding-some
Matt wants Tumblr to embrace the #Fediverse. On #Twitter Matt also wanted #Matrix integration for Tumblr as well.
However, I have read from Tumblr employees (current & former) blog posts that #ActivityPub support is not actively being worked on: https://darnell.tv/2023/10/01/tumblr-fediverse-too-high/
Matt did like the idea of adding #JetPack though (which could enable auto posting to #Mastodon): https://www.tumblr.com/photomatt/733557666342977536/there-are-a-lot-of-jetpack-features-that-do-what
@darnell Here's hoping it still goes ahead.
The biggest benefits of federation come with scale. As more platforms federate, and there's more users on platforms that have federated, the greater the benefits are.
If the Fediverse is just Mastodon and a few smaller platforms in its orbit (PeerTube, Pixelfed, etc), then the value of federation is relatively low.
Add Tumblr, and suddenly there's an easy-to-use gateway to the Fediverse for newcomers. Which, in turn, helps the other platforms to grow, which means more people and content to interact with on Tumblr.
@ajsadauskas from everything I've seen I get the impression that everyone WANTS Fediverse integration, but when Tumblr started working on the implementation they quickly found that Fediverse was designed in a way that is technically too expensive or unworkable for them.
The biggest benefits of federation come with scale, but unfortunately it was designed in such a way that the resources required to operate it scale even faster.
So as it scales the cost blew up faster than the benefits, if that makes sense.
These are fundamental design decisions made long ago that can't really be changed at this point.
Everyone agrees that integration would be good. It's just that it ended up being too expensive to actually implement.
