Rabe_Duck
Chris TrottierTwitter is killing free API access.
Will most devs pay for access? No thanks.
If what happened to 3rd party clients proves true, I have a feeling many devs will move their efforts to the Fediverse.
Killing free API access is an incredibly terrible business decision from Twitter.
1. It was once free
2. It’s main competitor (Mastodon) doesn’t just offer a free API, it’s open source
3. It’s also competing with ActivtyPub
Elon Musk is motivating the Fediverse ecosystem!
What Elon Musk doesn’t realize is that there’s a whole lot of devs who used Twitter’s API, not because of the network effect, but because it was a free API to build neat products upon.
Now you’ve removed the biggest motivation for actually using that API.
People still ask, “What if Twitter joins the Fediverse?”
A better question: with all the effort they’re putting in to destroy free Twitter API access, why would they want to join the Fediverse?
Elon Musk bought Twitter for the network effect. We know that.
Twitter’s actual technology is unimpressive, and the reason Mastodon is able to actually be a viable alternative is because it actually reached feature parity in *most* of the ways that matter.
But a network effect is only important insofar as people desire to use the tech.
What current Twitter doesn’t appreciate is that a free API is a mechanism that allows small projects to leap frog larger corporations.
Former Twitter knew that. It’s how they became Twitter.
Elon Musk is probably thinking, “What can Masodon do? They have a handful of employees.”
And that’s his undoing.
Other devs will build on Mastodon because the API is free—and Twitter’s isn’t anymore.
Frank Bennett 🇯🇵@atomicpoet The Twitter API has been used by disaster-notification bots, such as Japan's NERV system. If Twitter aspired to serve a public service role, it seems that's been abandoned.
Samir Al-Battran@atomicpoet
You don't need a lot of developers if you have an open API
You don't need a lot of infrastructure engineers if your platform is decentralized
You don't need BDRs if you don't sell ads
And don't ignore the collective wisdom of the developer community supporting you!
You don't need a lot of developers if you have an open API
You don't need a lot of infrastructure engineers if your platform is decentralized
You don't need BDRs if you don't sell ads
And don't ignore the collective wisdom of the developer community supporting you!
George Roberts 🇵🇸@atomicpoet I loved former twitter with the old Adobe Air Tweetdeck. Aesthetic and function were what I thought would be a "vision of the good future". There was a brief while when the F***book api was open and could be read through Tweetdeck. When FB shut down its apis and algorithmised the timeline I stuck with twitter. one of millions, I suspect. My anecdotal account supports your assertion.
Not a Spring Onion@atomicpoet
"What if Elon Musk buys the Fediverse?"
"I can't give you any meaningful answer except that your question is based on wrong preconditions."
Miredly@atomicpoet a half-finished art project built on that api got me the best job I've ever had, really sad to see them do us dirty like that
edgeoforever 🍎@atomicpoet tbh, he removed the network effect for the rest of us too.
Connected Places@atomicpoet By an large this is a pretty good thing and boost for the #fediverse. The one worry I have is that people will treat the Twitter API and Mastodon API as the same, without regard for how people here feel about scrapers. Think we will revisit the search/scraping/consent debate pretty regularly in the coming weeks and months.
Mikko Alasaarela@atomicpoet I agree. This is not even the first time Twitter is cutting off its APIs, so it has now betrayed its developers twice. I don't think there will be a third time, so the days of Twitter as a developer platform may be over.
Dean@atomicpoet that’s what he meant when he said it was for saving freedom of speech. Did he meant to kill Twitter all along?
maddie (MOVED!) 
@atomicpoet elon doesn't take mastodon seriously, he thinks it's not a threat despite it and the fediverse growing
Matthew 
(he/him)@mjdxp @atomicpoet I think he did blocks link to 3rd party spical media temporary which did include both Mastodon and Instagram. He reverse the change after the blacklash
luckysitsinbackwe're just lucky HE'S so fucking stoopid, too !!
John GordonThey will probably defer the cutoff but it’s fun to extrapolate on what this means.
Let’s assume Musk doesn’t care about Twitter the product. He bought the customer base. Those who remain are his addicts and he can ride them hard.
He cares about his Mars ambitions and Tesla and he wants the GOP in power (in return for space money and crushing his irritants).
Given that what we will have his addicts do? He will want them to use his payments scheme …
DaltonRoad@atomicpoet
sorry, but you say this like his other decisions make sense---
Twitter wasn't profitable before he bought it, they were surviving on cash flow from ad sales and that's the only thing allowing them to pay rent, pay employees and pay hosting fees... what does musk do week one? He alienates the advertisers and there goes the cash flow.
Granted his investor's goal has always been the death of twitter but still... he's not even trying to pretend that's not what he's doing.
Adam Dalliance@atomicpoet as soon as my robot can't forward tweets to rss i stop reading any Twitter at all. Sounds like we're down to the last week!
lamp
@atomicpoet remember when elon said he was gonna open source a bunch of twitter's components?
Tomi the Slav and 1024 others@atomicpoet Oh. This is the last no-no for me.
I still have a business account there and some simple bots that I wrote just out of curiosity and learning.
I'm switching them off and will rewrite them for Mastodon.
I still have a business account there and some simple bots that I wrote just out of curiosity and learning.
I'm switching them off and will rewrite them for Mastodon.
One year ago I started a little experiment with my #Twitter account (for my company). TLDR: it ended badly.
I wrote 3 python scripts for Twitter:
1. auto-post tweets 3 times a day from excel (I wrote approx 500 tweets in advance - business related),
2. follow interesting people (predefined keywords) and
3. unfollow inactive accounts
Scripts ran fine, but the result was:
I was followed mostly by shady bots, Asian girls with links to p0rn sites (also bots).
Summary: make a bot, get a bot.
Fabio Manganiello@atomicpoet let's not forget that the Fediverse still has entry barriers and a learning curve for those who aren't tech-savvy.
This would be the perfect moment for us to slay whatever is left of Twitter and convince everyone to move to the Fedi. Elon's sheer imbecility is literally providing with assists on a daily basis - from banning journalists, to promising zero censorship against the fascist, to actively working to compromise the stability of the platform, to turning Twitter's working environment into a toxic mix of chauvinism and stakanovism, to closing the APIs and killing 3rd-party clients, to the mess with the new verification program...
And, indeed, every time Elon decides to set a new bar for what being a true idiot means, Mastodon and the Fediverse get a new ripple of users.
But those ripples don't stay long. Looking at the stats of new users, one can easily overlap the peaks with Elon's acts of stupidity, but it's also true that not many of those who join in those ripples stick around.
It means that, no matter how bad other platforms are, most of the people still have trouble, for some reason, using a free and decentralized alternative that isn't run by any sociopath billionaire.
@blacklight This is where a dev ecosystem comes into play. More hands on deck, a better user experience will be created.
This is a continuing work in progress.
Paul Chambers🚧We see this in the current dash to create user apps.
Also, there are projects like @benbrown's single user, stand-alone NodeJS web application which requires no server dependencies, #activitypub server #Shuttlecraft, which I envision could be deployed in mass, on a user-by-user signup level. Users signup at a single sign-up page and their own single user server is created, that they own. Kind of a #Fediverse Geocities but server creation
@paul @atomicpoet @benbrown I love the idea of more apps and more decentralization efforts. But that goes at odds with how social networks work (on a sociological, not on a technological, level).
Even on Twitter, only a minority of the users used to use alternative clients. Sure, they were also those who spent a lot of time on the platform and cared the most about its features. But, statistically, they were still a minority (and probably not even those that the platform could monetize the most). Most of the people were happy with the default app, even if it sucked.
Does Twitter killing their free API and 3rd-party clients mean that those who used Twitterrific, TweetDeck or Twidere will probably move to the Fedi? Most likely yes. But, just to be clear, we're talking of a number of users within the hundreds of thousands, not more. Most of the Twitter's userbase didn't know/didn't care about APIs and alternative clients.
As I previously stated, social networks abide to the rules of (human) network effects: people tend to gravitate towards the solution with the lowest friction points, especially if that's also what most of those in their circle already use. No matter how good the alternatives are.
From this perspective, more decentralization + more user choice = more fragmentation + more friction coming from the initial cognitive burden required for adoption. We should not forget that we live in an age where consumers are used to somebody else making choices for them, and any additional burden of choice on the user is likely to decrease the chances of adoption.
Again, I'm not advocating to do things differently. I'm here exactly because of how things are done here. I'm just saying that we should be realistic about what to expect: people won't suddenly flood to Mastodon or Pleroma because they have more choice and more apps. Facebook and Twitter have managed to get away with literally anything (including things that would have taken any other business down within days in the past) because people are addicted to their products, and because they have benefited from social network effects for more than a decade. It's very hard to convince enough people to jump the gap if the alternative comes with a higher initial cognitive burden for the user.
I consider myself happy with the current number of people on the Fedi. I don't think that, given the way we do things, it makes even sense to aim for a more exponential-like growth going from here. I don't even want to think of scaling up my instance for a 10x-100x increase in federated traffic. A greater number of apps is likely to greatly improve the user experience, but only for a minority of the overall market of potential social media users. And actually I'm perfectly fine with it :)
@atomicpoet btw I'm a HUGE fan of APIs and 3rd-party clients. But we should also be careful about the fragmentation effect - something that has already been affecting Linux since its birth.
Walk for a moment in the shoes of the average Joe - not the Twitter geek, not the IT enthusiast.
Joe wants to share stuff with his friends on some social platform. He learns about Twitter. He creates an account, logs in, starts posting. End.
Now Joe learns about the Fedi. And people start asking him - which platform? Mastodon? Pleroma? Diaspora? Which server? Which client? And, by the time Joe is asked the third question, he's already walked away.
As an engineer and an open-source enthusiast, I often believe that all people want and need is more choice and a more customized experience. But every time people surprise me by choosing the solution that gives them the least amount of choice and power, as long as it's something that has close to zero entry barriers.
@blacklight Sorry, I don't buy "fragmentation destroyed Linux".
Fragmentation has allowed Linux to dominate EVERYTHING.
That is: the router, the TV, the smartphone, the car, the refrigerator.
Some always say, "But what about desktop?"
Windows now has Linux, and Chrome OS is basically Linux.
In fact "average Joe" (not a good persona because no one is average) uses Linux more than any other OS.
Does he know it's Linux? Nope.
But so what?
@blacklight Devs generally have to stop being myopic about the Fediverse.
Multiple protocols and platforms and clients should exist.
This is bigger than a "Twitter replacement".
This could be bigger than the web itself if people had a little more imagination.
Think about how ActivityPub can apply to creative collaboration, gaming, commerce, and leisure.
Endless possibilities, and people already want to re-centralize.
@atomicpoet I've spent most of my life advocating for open protocols, no need for further convincing there :) backend decentralization, multiple clients and shared protocols are the foundation of the Internet.
But when I try to look at things from the perspective of average Joe, it seems to me that a social network abides to different rules, and the successful platforms eventually are those that gravitate towards centralization - because of network effects.
Average Joe doesn't care to have a choice of 30 possible apps to interact with social media. Nor about RSS feeds, ActivityPub, relays, instances, directories of users and so on. Average Joe wants to download the app that all of his friends and family members are using, register, log in, and immediately be able to communicate with them (like Facebook/Twitter). Or log in and directly be exposed to a continuous feed of self-curating content designed to be addictive, without further user interaction (like TikTok/Snapchat and, lately, Instagram).
Everything else is friction, and it creates a funnel where only very few among those who start eventually decide to stick around. The barriers to take users away from products they are already addicted to (and all of their friends are also addicted to) are high, no matter how good the alternative is. That's why these companies can literally get away with everything without ever losing too much of their user base.
Again, I'm not advocating for doing things in a more centralized way - quite the opposite: the reason why I'm here is because I can tinker with the platform and the source code to customize my experience. And I'm not even advocating to lower the barriers so more users can come in - quite the opposite: more users means heavier burden on our instances, more moderation efforts, and more chances for things to get corrupt as more admins will need more money/resources to run their thing.
But then we also have to be realistic: our requirements and expectations from social media put us in a position of minority. I think that 10M users on the Fedi is a good trade-off, and aiming for higher volumes may not even be what this thing was supposed to do (on a social/cultural/economical level, not on a purely technical level).
@blacklight Average Joe is not realistic.
No social network has been built for Average Joe in mind.
Drill down to every social network that gained traction, it was made with a specific persona in mind.
Example: Facebook was made for Ivy League students that wanted to check on the relational status of friends / crushes.
Certainly, Facebook expanded from this persona—but that's not how it started.
Same is true for Mastodon. It succeeds because the persona of who it was made for is real.
@atomicpoet I'm not saying that the Fedi should be built for Average Joe. If that was the case, I'd probably leave this place.
But, if we want something to grow to big scales, we should have Average Joe in mind, because the average person out there is neither the guy with a degree in science or engineering, nor the privacy enthusiast, nor the social activist.
True, Facebook wasn't built for Average Joe, but it's been increasingly adjusted to the point that it could appeal also to Average Joe. To the point that my parents in their 60s can easily register, use it and contact anybody they can without much technical assistance. I tried to also explain to them how they could do the same on Mastodon, but they already got lost when I tried to explain them what an instance is.
In order to get to the Average Joe, platforms like Facebook scavenged all the user data they could put their hands on, ran tons of A/B experiments to make their interface as intuitive as possible, and purposefully designed the platform to be addictive so users would come back to it as often as possible.
It's hard for us to get to the same point without some trade-offs that many people who either built or run this place would have a hard time accepting. And it's even harder because your idea would be to penetrate a market that is already heavily consolidated, with high entrance barriers, and with users who are already so addicted to the few available products that they keep using them even when those businesses are literally selling their private lives to anyone without their consent.
And, again, maybe it's better if things stay like this. I'm perfectly ok if only a minority of users who are aligned with the values and trade-offs of the Fediverse come onboard, rather than compromising too many of our values to try to make a bigger dent on Facebook's or Twitter's user base.
The Fedi is likely to grow, but it's going to be a modest linear growth rather than an exponential explosion. People will keep flocking here when Elon or Mark mess up things too much, but it's likely to be in ripples of a few thousands at a time, not millions.
@blacklight The success of Facebook isn’t that they built an app for a collective “average Joe”. It’s that they first concentrated on one persona, then expanded it to several personas, building mechanisms to keep each of these personas in their walled gardens.
However, the growth of the Fediverse will be in developing unique apps for unique personas. Not a walled garden. A flourishing ecosystem.
Schoolteachers will have their own Fediverse app. Accountants too.
@atomicpoet I wish this could be true, while keeping at the same time the growth of the Fediverse sustainable without too many compromises.
Personally I'm quite disillusioned, having been burned years ago by the way the Web 3.0 project (the real one, the semantic Web, not the crypto-scam) failed. That idea had a lot in common with the Fediverse, and many of those behind it later started working on ActivityPub too. Even that initiative was based on open protocols, decentralization and machine-to-machine markup exchange (not only markup made for the sake of human eyes).
It ticked all the boxes. Technologically speaking, it was literally the ultimate solution for how the Web should work.
Then venture capitalists realized that they couldn't make much money out of open decentralized protocols, big tech realized that the whole idea was actually a threat to their hegemony over data, the technological "progress" went where the money went, and the whole idea lost momentum.
I see the same risks for the Fediverse. Besides the barriers on the user adoption side, there are barriers on the economic side too.
Money goes where investors can make more money, where there's a high change of locking in customers, and where there's a compelling business model. There aren't many of these things in decentralized networks and open protocols. And we will need more money if we want to scale up the Fediverse - to add more servers, to pay the bills, to spend more time on moderation and maintenance etc. Volunteers like me aren't that willing to take on more duties and costs while being rewarded only with sporadic donations. So, even if we manage to jump the entry barriers and expand the user base to hundreds of millions, who's going to run the instances in charge of ingesting millions of posts per day into their federated timelines? Who is going to run and maintain the Fediverse platforms and apps for teachers, accountants and students?
I only see this happening in a world where instances become more isolated, because participating in large relays with thousands of other high-volume instances will mean pumping more content than today's instances can chew.
Or if governments step in to fund our platforms, because the private sector works on a completely different set of incentives. And that's something that is already being discussed in the EU (assuming that their current Mastodon experiment won't be just a hay fire).
However, when it comes to independence, government-sponsored initiatives and government-funded platforms come with similar problems as those that you would have by relying on the money on somebody waiting to make a profit. So I really don't see a solution on the horizon that would allows us to scale up exponentially without having to accept uncomfortable trade-offs.
The Fediverse has many solution for the technological problem. But only once we also solve the users problem ("how do we move millions of people away from products they're addicted to and onto an alternative with higher entry barriers?") and the financial problem ("how do we make sure that there are enough financial incentives in place for people to keep the lights on, even when the resources to keep them on go up a ramp?") we can think of *really* scaling things up here.
@blacklight Dragons can be slayed:
guy,in,a,hat@atomicpoet @blacklight this is to the point that Linux Desktop Ubuntu is 👍 great. I've used Linux since 2003 or so. It has improved steadily without fuckin everything up that you have worked for. #Windows / #Microsoft likes to reinvent the wheel every few years causing shitty migrations from hell. Linux has been pretty stable and I still have all docs and emails I have ever written in #openoffice #libreoffice. There have been a ton of hiccups along the way, but not bad.
@atomicpoet I should have clarified a bit better: by "destroying Linux" I didn't mean "Linux has remained a niche product". Of course it's not, of course it's everywhere - and, of course, the desktop experience is not even that bad.
But you also raised a good point: "does average Joe know that he's using Linux?"
Average Joe doesn't need to know that his phone, TV, car and dishwasher (and, of course, all of the servers that run the Internet) are running some flavour of Linux. He perceives them as completely different products with completely different purposes and, as long as he's provided with an intuitive interface, he doesn't even have to care. Barriers have already been lowered by the producer of the good or service
That's not the same for social media. When it comes to social media, average Joe has to take an *active* action - set up an account, figure out how things work in the least possible number of steps, and get used to a new interface. And that's where the Fedi still has many friction points because of the number of choices that average Joe needs to make before being able to be an active user.
(And btw, I'm not advocating to lower the barriers to let average Joe in. We may also decide that the technical trade-offs of the Fedi are exactly what keeps it good instead of being just another centralized platform. We may also decide that we don't want to dilute the quality of the content here by letting the flood gates open).
@blacklight By ActivityPub existing, the Fediverse is ultimately the "Linux of Social Networks".
Also, I reject your premise because Average Joe is a shitty persona.
How do I know that a social media app will die? It's made for Average Joe.
Scott M. Stolz@Fabio Manganiello @Chris Trottier My tactic for dealing with that is to create communities that people want to join, revolving around content and discussion boards, and giving them a fediverse account for free. They join for my website's community and content, and then perhaps later they realize they can talk to the entire fediverse. It is a way of onboarding people into the fediverse that don't even know what the fediverse is yet.
Matthew@atomicpoet Anybody know if this will affect Nitter, or does Nitter not use the API?
GunChleoc@picard From using Nitter, I'm already under the subjective impression that some throttling has been going on. Statuses look fine, but profiles only seem to give the latest tweet.
@crossposter has already given up due to the API throttling.
I would not be surprised if in the long run they will kill Nitter too, because it gives them no direct ad revenue.
@gunchleoc @atomicpoet Thanks. I mostly use Nitter for the RSS of accounts I want to read, so far that has been OK but like you say I do expect them to take action against it at some point as it's purely a cost on them.
I just checked an account on uk.unofficialbird.com and it did show more than the most recent post, so maybe the larger instances are getting throttled more or something.
@picard I'd expect what they can fetch will depend on the traffic they're getting and that everybody has the same rate limit, so a less busy instance would be able to fetch more @atomicpoet
@gunchleoc @atomicpoet That'd be my guess yeah. I suppose Nitter was good while it lasted, but it was probably never likely to be permanent...
ophiocephalic 🐍@picard @gunchleoc @atomicpoet According to the Nitter developer it won't be affected as it doesn't use the APIs "in the official way".
Ted Garrison@atomicpoet Finally figured out the only way to kill the bots?
Juraj Maťaše@atomicpoet It's funny to watch this slow suicide.
Zak8022 
@atomicpoet Given the way they handled third party apps a few weeks ago… I’m shocked they announced this ahead of time!
Kichurano wonder why i quit twitter.
LBS@Home@atomicpoet Yeah I'll be trying to post here more actively once that happens, can't rely on the crossposting app anymore.
ElOssiPolar
levampyre@atomicpoet How is birdsy still alive!?
Marcus
Crone Cold Kate Sherrod@dc5dm @atomicpoet I loved Plume so much!
Steve Lhomme@dc5dm I worked on it for 3 years, then it was sold and I moved to other things (and iOS). I was barely a Twitter user before that. I learned to love it. A sad day. RIP