XMPP still exists. Google dropped support for it, that’s definitely not killing it. Google drops support for projects all the time by the way, it’s kind of their thing.
Google dropping support for XMPP is what put it one foot in the grave. They abused the protocol to gain the lion’s share of users for Google Talk, and then cut off any resistance that remained. It exists still, technically, but when’s the last time you heard about or used it? I only know about it because EVE Online players used it for large group text communication before Discord became a thing.
XMPP still exists in the same way that critically endangered animals still exist: barely and by the adamant will of some dedicated few.
XMPP wasn’t even remotely popular until Google integrated with it, I tried Jabber back in the day lol. Google brought the users it lost, you can’t argue this was an attempt to kill it. At worst it’s the same as before Google integrated.
That’s the problem though. If XMPP had grew organically then it would fare much better. With how it happened, XMPP’s growth was mostly because of Google, and that put a lot of pressure to other servers and the protocol’s development to cater to them, because they had the majority of the users in their platform.
It is absurd to think XMPP would have gained traction without Google. And it is an objectively shitty protocol, so Google dropping it was the right move. It is kind of weird to see people holding up Google dropping XMPP as some horrifying example of embrace, extend, extinguish, when anyone that’s actually developed software with the protocol wants it to die in a burning fire.
How convoluted the protocol is doesn’t really matter as long as someone creates an easy tool to spin up your own server.
I think the XMPP comparison stills stands: Google was able to steer how the protocol developed, or which version of the protocol people used because they had the majority of the users and other servers wanted to still be able to interact with them.
Suppose that Facebook joins the fediverse and most large instances federate with them. All is great, then Facebook starts to make demands to other instances in order to keep federating with them, e.g. no posts about protests. Because a large share of ActivityPub activity will be on Threads, naive users would prefer instances that federate with it, so instance mods will be incentivized to comply with Facebook’s demands to attract new users and maintain their current one and… you see where this is going. The only way to deal with this is to deny Facebook this kind of leverage in the first place, either by blocking them instantly or at their first mishap or demand.
so instance mods will be incentivized to comply with Facebook’s demands to attract new users and maintain their current one
This is where your argument falls apart. Why? There is no incentive for instance mods to want to grow their instances exponentially.
If Facebook’s ActivityPub grows to be incompatible with the existing implementation, who cares? So what if you run a Mastodon instance and aren’t getting millions of new users a day?
This is much ado about nothing. While there is a shared platform, enjoy the ride, and if they don’t want to play by your rules anymore, there’s no harm to anyone in saying goodbye and staying your course.
It’s true that instances don’t need to grow exponentially (or at all), but most mods/admins want to maintain their community and not see it dwindle down to nothing. People used to interacting with instances run by Facebook or other corporations (which most of their friends or family will use) might get upset if the federation link with them gets severed. If they do, they’ll either pressure the instance admin to comply with the corporations and federate with them again, or switch to the corporations’ instances. Both of these scenarios are bad for the future of the fediverse.
I’m not sure how things are going to go with Meta and federation and EEE could happen and definitely see some of the concerns, but the way people are just pointing to that XMPP article in every thread as some slam dunk argument I think is overstating it. It’s one example and there are lots of other considerations around it and different context that make it so it’s not something that can really be directly mapped onto this situation.
Things may go south with Meta and federation but the constant pointing to XMPP is not really making a solid argument IMO.
I think it’s all besides the point anyway. Some servers will federate with Meta and any other big companies that enter the Fediverse. Some wont. Meta is big enough not to care, and the big Masto servers are also going to do what they want to do and allow federation. And if there’s desire from Mastodon users to connect with Threads and follow accounts there people will move to servers that allow that. And then there may be communities that aren’t federated with Meta that are also great and strong. We’ll see how it plays out, but small Masto/Lemmy servers choosing to not Federate I don’t think will have much impact broadly speaking on how this goes. But by the same token if servers don’t want to federate with Meta that’s totally cool too and I respect that as well. We’ll have some parts of the Fediverse in the future that connect with the big platforms and some that don’t. That’s the path we’re on now either way — some will federate, some won’t — and people can choose which part they want to be part of.
Personally I think the Fediverse and ActivityPub will be more resilient than XMPP and will be durable against EEE. Especially if other players like Tumblr and Wordpress jump in that will strengthen interoperable ActivityPub even more. If people want to not federate with Meta that’s cool and I definitely see some good points around it (but not so much the much heralded XMPP article) but I think the Fediverse will be fine either way and ActivityPub’s future is looking stronger than ever.
This is pure speculation at best, but since we’re speculating I strongly disagree. The internet overall didn’t care about open source software in the early 00s, and most people still don’t today. Corporate freeware that can spend more on a polished product is going to win over the general population every time.
Talking about any alternative scenario is always speculation, but I believe the “How to kill decentralized networks” post that’s been going around lately puts it nicely:
One thing is sure: if Google had not joined, XMPP would not be worse than it is today.
You missed rest of my comment. You, and this article, are speculating on made up assumptions, and frankly silly assumptions. Open source software is almost never more popular than freeware counterparts. Saying “oh maybe it would’ve been this time” is ridiculous.
Can you explain how Google helped XMPP even in the slightest way? Because that’s what I’m arguing against.
The only thing I can come up with is the increased popularity, which is shaky because tech-naive users wouldn’t know or care about Google Talk’s underlying protocol. Also, considering the rest of what Google did with XMPP, like making it hard for their servers to be interoperable with others, or their slow adoption of new features, it’s clear to me that Google getting involved was a net negative for XMPP. I don’t think I’m assuming anything to arrive on that conclusion.
I never argued that Google helped XMPP, I’m arguing that it isn’t applicable to the “extend, embrace, extinguish” crap that people keep parroting like it’s an actual playbook used by tech companies and not just some silly nonsense created by some middle manager at Microsoft 30 years ago lol. The users Google brought they took, at worst it was net neutral.
like making it hard for their servers to be interoperable with others
Because they forked their own deviations of XMPP to work with the updates made to Google Talk. It’s original state was left untouched and by no means “extinguished”. This is just another example of corporate freeware winning over open sourced because of a more polished product.
their slow adoption of new features
I assume you mean Jingle which they adopted in 2007? Why would slow adoption of XMPP features into Google Talk affect non Google Talk XMPP users? They were always free to use XMPP without Google Talk, just as we’re free to stay on Lemmy/kbin/Mastadon without Threads.
enthusiast dev here, can vouch, having to make a XMPP library for myself for a bot I ran, I HATE the protocol with a burning passion, it’s weird and not how you would expect it to be. I’m sure the complexity of the standard didn’t help against its downfall.
That being said, fully think that it will be harmful in the longrun of Activity Pub for Meta to be jumping in. but there will be some enthusiasts that still use it regardless.
XMPP still exists. Google dropped support for it, that’s definitely not killing it. Google drops support for projects all the time by the way, it’s kind of their thing.
Google dropping support for XMPP is what put it one foot in the grave. They abused the protocol to gain the lion’s share of users for Google Talk, and then cut off any resistance that remained. It exists still, technically, but when’s the last time you heard about or used it? I only know about it because EVE Online players used it for large group text communication before Discord became a thing.
XMPP still exists in the same way that critically endangered animals still exist: barely and by the adamant will of some dedicated few.
XMPP wasn’t even remotely popular until Google integrated with it, I tried Jabber back in the day lol. Google brought the users it lost, you can’t argue this was an attempt to kill it. At worst it’s the same as before Google integrated.
That’s the problem though. If XMPP had grew organically then it would fare much better. With how it happened, XMPP’s growth was mostly because of Google, and that put a lot of pressure to other servers and the protocol’s development to cater to them, because they had the majority of the users in their platform.
It is absurd to think XMPP would have gained traction without Google. And it is an objectively shitty protocol, so Google dropping it was the right move. It is kind of weird to see people holding up Google dropping XMPP as some horrifying example of embrace, extend, extinguish, when anyone that’s actually developed software with the protocol wants it to die in a burning fire.
How convoluted the protocol is doesn’t really matter as long as someone creates an easy tool to spin up your own server.
I think the XMPP comparison stills stands: Google was able to steer how the protocol developed, or which version of the protocol people used because they had the majority of the users and other servers wanted to still be able to interact with them.
Suppose that Facebook joins the fediverse and most large instances federate with them. All is great, then Facebook starts to make demands to other instances in order to keep federating with them, e.g. no posts about protests. Because a large share of ActivityPub activity will be on Threads, naive users would prefer instances that federate with it, so instance mods will be incentivized to comply with Facebook’s demands to attract new users and maintain their current one and… you see where this is going. The only way to deal with this is to deny Facebook this kind of leverage in the first place, either by blocking them instantly or at their first mishap or demand.
This is where your argument falls apart. Why? There is no incentive for instance mods to want to grow their instances exponentially.
If Facebook’s ActivityPub grows to be incompatible with the existing implementation, who cares? So what if you run a Mastodon instance and aren’t getting millions of new users a day?
This is much ado about nothing. While there is a shared platform, enjoy the ride, and if they don’t want to play by your rules anymore, there’s no harm to anyone in saying goodbye and staying your course.
It’s true that instances don’t need to grow exponentially (or at all), but most mods/admins want to maintain their community and not see it dwindle down to nothing. People used to interacting with instances run by Facebook or other corporations (which most of their friends or family will use) might get upset if the federation link with them gets severed. If they do, they’ll either pressure the instance admin to comply with the corporations and federate with them again, or switch to the corporations’ instances. Both of these scenarios are bad for the future of the fediverse.
I’m not sure how things are going to go with Meta and federation and EEE could happen and definitely see some of the concerns, but the way people are just pointing to that XMPP article in every thread as some slam dunk argument I think is overstating it. It’s one example and there are lots of other considerations around it and different context that make it so it’s not something that can really be directly mapped onto this situation.
Things may go south with Meta and federation but the constant pointing to XMPP is not really making a solid argument IMO.
I think it’s all besides the point anyway. Some servers will federate with Meta and any other big companies that enter the Fediverse. Some wont. Meta is big enough not to care, and the big Masto servers are also going to do what they want to do and allow federation. And if there’s desire from Mastodon users to connect with Threads and follow accounts there people will move to servers that allow that. And then there may be communities that aren’t federated with Meta that are also great and strong. We’ll see how it plays out, but small Masto/Lemmy servers choosing to not Federate I don’t think will have much impact broadly speaking on how this goes. But by the same token if servers don’t want to federate with Meta that’s totally cool too and I respect that as well. We’ll have some parts of the Fediverse in the future that connect with the big platforms and some that don’t. That’s the path we’re on now either way — some will federate, some won’t — and people can choose which part they want to be part of.
Personally I think the Fediverse and ActivityPub will be more resilient than XMPP and will be durable against EEE. Especially if other players like Tumblr and Wordpress jump in that will strengthen interoperable ActivityPub even more. If people want to not federate with Meta that’s cool and I definitely see some good points around it (but not so much the much heralded XMPP article) but I think the Fediverse will be fine either way and ActivityPub’s future is looking stronger than ever.
This is pure speculation at best, but since we’re speculating I strongly disagree. The internet overall didn’t care about open source software in the early 00s, and most people still don’t today. Corporate freeware that can spend more on a polished product is going to win over the general population every time.
Talking about any alternative scenario is always speculation, but I believe the “How to kill decentralized networks” post that’s been going around lately puts it nicely:
You missed rest of my comment. You, and this article, are speculating on made up assumptions, and frankly silly assumptions. Open source software is almost never more popular than freeware counterparts. Saying “oh maybe it would’ve been this time” is ridiculous.
Can you explain how Google helped XMPP even in the slightest way? Because that’s what I’m arguing against.
The only thing I can come up with is the increased popularity, which is shaky because tech-naive users wouldn’t know or care about Google Talk’s underlying protocol. Also, considering the rest of what Google did with XMPP, like making it hard for their servers to be interoperable with others, or their slow adoption of new features, it’s clear to me that Google getting involved was a net negative for XMPP. I don’t think I’m assuming anything to arrive on that conclusion.
I never argued that Google helped XMPP, I’m arguing that it isn’t applicable to the “extend, embrace, extinguish” crap that people keep parroting like it’s an actual playbook used by tech companies and not just some silly nonsense created by some middle manager at Microsoft 30 years ago lol. The users Google brought they took, at worst it was net neutral.
Because they forked their own deviations of XMPP to work with the updates made to Google Talk. It’s original state was left untouched and by no means “extinguished”. This is just another example of corporate freeware winning over open sourced because of a more polished product.
I assume you mean Jingle which they adopted in 2007? Why would slow adoption of XMPP features into Google Talk affect non Google Talk XMPP users? They were always free to use XMPP without Google Talk, just as we’re free to stay on Lemmy/kbin/Mastadon without Threads.
App I work on, we’re replacing XMPP with messages over push/rest/websocket. XMPP is not fun to use compared to newer stuff.
enthusiast dev here, can vouch, having to make a XMPP library for myself for a bot I ran, I HATE the protocol with a burning passion, it’s weird and not how you would expect it to be. I’m sure the complexity of the standard didn’t help against its downfall. That being said, fully think that it will be harmful in the longrun of Activity Pub for Meta to be jumping in. but there will be some enthusiasts that still use it regardless.