Nah you’re on the mark, it was more like the older Internet where you were just a username and didn’t have your entire identity attached to your account. And yeah, for better or worse Tumblr was known to have a very insular community with weird interests and trends. That’s how we got Goncharov.
Two other things about Tumblr that no other site has managed to capture are longer format posts and being able to reblog multi-user conversations. Everything else prevents you from making longer posts outside of screenshotting a notes app.
Word from insiders communicating openly about it was that it wasn’t going to happen for the same reasons behind this soft end of life: there wasn’t profit in it and they needed to find profitability.
Word was that the project to federate was internally put in a “too hard let’s try later” basket. Of course none of that was made public. But this life support status is really just the other shoe from the broken federation promise.
Interestingly, I suspect there’s a conversation to be had around how difficult it is to federate with the fediverse and AP, whether from the ground up or as an add on. I’ve seen conversations about how it really is difficult work. If it were easier, maybe things like tumblr would have done it earlier and the fediverse and it’s various platforms would be richer and more successful.
The good thing is that the DIY people don’t need to reinvent the wheel - you can just fork an existing project and shape it into what you’re interested in. There’s are already some bare bones implementations around that lend themselves well to building things on top of.
Of course not arguing that making it simple isn’t a good thing - but the reality is that interoperability between platforms like this is a complex challenge, and even though nothing is perfect ActivityPub is doing a pretty neat job making it a reality.
The growing options to just fork is an interesting and good point. Quite a few languages and stacks are now represented in some project somewhere.
That being said, we’re probably fast approaching the point where some form of generic modular software for the lower level stuff would provide a productivity boost to the ecosystem. My understanding is that there’s some work starting on that front for testing and maybe a sort of reference implementation.
Do you mean being attacked by outsiders? Reddit was down a significant amount of time and they had tons of resources. We’re doing okay and everyone powers through.
No, just the work of building the federation part of a platform.
Look at both mastodon and Lemmy, both have a huge blind spot for a major part of the ActivityPub protocol. Mastodon doesn’t manage groups well and Lemmy doesn’t manage users well, as content creating actors that is. This is obviously by design to a large extent, but the difficulty of expanding the bounds of a platform to include more of the spec seems to clearly be part of the problem.
Part of what problem? It would be nice if we could all communicate easily but it’s not really a problem. Just like twitter and facebook not really communicating isn’t a problem. If they decide to blend them, cool. If not, that’s cool too.
That platforms struggle to interact with each other more than the promise of the fediverse implies.
Segregated platforms is a hang over from competitive big social. It’s all just text messages in the end as webpages are all just HTML. Part of the promise of the fediverse is to break down these capitalistic walls. If it’s too hard to engineer though, then the promise may need some rethinking.
That’s not good, especially as they wanted to join the Fediverse in the future.
I mean there’s Pixelfed which is similar, the communities could just migrate to that.
I feel like it’s a different format and use case?
Pixelfed feels more like Instagram, which is with real profiles with names attached sharing photo updates about their lives with friends and family
Tumblr feels like it’s more for niche communities and semi-anonymous posting. Although I personally haven’t used it as much so I might be off there
Nah you’re on the mark, it was more like the older Internet where you were just a username and didn’t have your entire identity attached to your account. And yeah, for better or worse Tumblr was known to have a very insular community with weird interests and trends. That’s how we got Goncharov.
Two other things about Tumblr that no other site has managed to capture are longer format posts and being able to reblog multi-user conversations. Everything else prevents you from making longer posts outside of screenshotting a notes app.
Didn’t know this existed. Good to know.
Word from insiders communicating openly about it was that it wasn’t going to happen for the same reasons behind this soft end of life: there wasn’t profit in it and they needed to find profitability.
Word was that the project to federate was internally put in a “too hard let’s try later” basket. Of course none of that was made public. But this life support status is really just the other shoe from the broken federation promise.
Interestingly, I suspect there’s a conversation to be had around how difficult it is to federate with the fediverse and AP, whether from the ground up or as an add on. I’ve seen conversations about how it really is difficult work. If it were easier, maybe things like tumblr would have done it earlier and the fediverse and it’s various platforms would be richer and more successful.
Meh, it’s harder for a for-profit company but not hard for the DIY people. I’m cool with that, it’s rich enough and getting richer.
Well from what I’ve gathered it’s tough work for DIY people too.
The good thing is that the DIY people don’t need to reinvent the wheel - you can just fork an existing project and shape it into what you’re interested in. There’s are already some bare bones implementations around that lend themselves well to building things on top of.
Of course not arguing that making it simple isn’t a good thing - but the reality is that interoperability between platforms like this is a complex challenge, and even though nothing is perfect ActivityPub is doing a pretty neat job making it a reality.
The growing options to just fork is an interesting and good point. Quite a few languages and stacks are now represented in some project somewhere.
That being said, we’re probably fast approaching the point where some form of generic modular software for the lower level stuff would provide a productivity boost to the ecosystem. My understanding is that there’s some work starting on that front for testing and maybe a sort of reference implementation.
Do you mean being attacked by outsiders? Reddit was down a significant amount of time and they had tons of resources. We’re doing okay and everyone powers through.
No, just the work of building the federation part of a platform.
Look at both mastodon and Lemmy, both have a huge blind spot for a major part of the ActivityPub protocol. Mastodon doesn’t manage groups well and Lemmy doesn’t manage users well, as content creating actors that is. This is obviously by design to a large extent, but the difficulty of expanding the bounds of a platform to include more of the spec seems to clearly be part of the problem.
Part of what problem? It would be nice if we could all communicate easily but it’s not really a problem. Just like twitter and facebook not really communicating isn’t a problem. If they decide to blend them, cool. If not, that’s cool too.
That platforms struggle to interact with each other more than the promise of the fediverse implies.
Segregated platforms is a hang over from competitive big social. It’s all just text messages in the end as webpages are all just HTML. Part of the promise of the fediverse is to break down these capitalistic walls. If it’s too hard to engineer though, then the promise may need some rethinking.