Medium Embraces Mastodon
blog.medium.com
blog.medium.com
I would love to see an integration where the discussion on a Medium post is viewable both as comments at the bottom of the page and as Fediverse posts. (This is what I was pushing in https://www.jefftk.com/p/mastodon-replies-as-comments)
Definitely a possibility. Where we ended up landing is that Mastodon is culturally short form. From prior experience just syndicating into a short form ecosystem is kind of lame even if the ActivityPub protocol supports it. So if we have Medium authors on this part of the fediverse then we want to encourage them to be fully there.
The medium-length fediverse, i.e. blogs and articles, is still emerging and for us is also a much more complicated product question. But definitely open to it.
This is essentially how blogs are promoted on Twitter and Mastodon already (though people often create longer threads) so it's a "paving the cowpaths" strategy.
Since you can do it already, the only difference would be having an official short post representing the article, which would be a focal point for discussion about the article.
For better or worse. Maybe having a focal point for article discussion is bad?
This is still somewhat significant because if you see comments from @someone@me.dm you'll know it's the person who goes by the same name as a Medium author, who you might know from elsewhere. This would be more effective if the Mastodon instance were at something like fed.medium.com though.
Mastodon represents a short form corner of the fediverse. To participate in this corner via a medium-length tool would mean participation from people who are blindly syndicating at worst, and having a disconnected experience at best. It'd be disrespectful to the community there.
Instead, Medium authors should be fully present as short form participants, even if all they want to do is share a link.
To speak in the town square you need to be present in the town square. In the case of Mastodon, that means reading and writing from a short-form app.
The point of adding ActivityPub (both ways) would be to be part of a much, much smaller blog-length corner of the fediverse. It's interesting, but is fundamentally a different goal.
Archive.org has their own instance @ https://mastodon.archive.org
https://hachyderm.io/@PeterBronez/109552198943104609
Current list is:
The Internet Archive Open Street Map Open Source Initiative Mozilla (planned) Association for Computing Machinery (ACM) Medium
Looking forward to seeing how this plays out as more entities take this approach.
Worth noting that fediverse itself isn't -- my Pleroma server has a 5000 character limit, other servers I've seen have more. There's no reason why they couldn't have every medium post available in fediverse other than profit motivation... look at writefreely as an example for a blog platform that incorporates fediverse, or, uhhh, Wordpress.
Well, it's still better that they're wading in the waters than not wading at all, even if they're not willing to dive in :-)
If Medium publishes everything to the Fediverse, nobody will pay for Medium anymore.
Free articles (marked as such by the authors) should still be postable, though.
That said, I don't know if the average Mastodon user or app expects blog posts with 10k words and embedded media. I would expect such rich content, but the "using Mastodon as a Twitter alternative" may not. In the end, it's up to Medium to guess what their user base expects from them.
I believe Mastodon already does this in web interface.
With RSS the person that provided your reader was able to do the same. And you're still able to run your own instance, just like you've been able to run your own RSS reader.
Not comparable, and honestly I find it dishonest that you even try.
RSS can be just a client on your machine, no need for server running in the cloud. That's far less effort involved. Also I wouldn't be surprised if some instances start to close down traffic from "random one man instances" at some point.
Especially that moderation and "for free" usually ends up with power trips being actual payment to the mods
> as long as there are other servers you can feasibly switch to if you don't agree with their definition of "spam".
Not that simple. If server A doesn't like X (not even abuse, might just not like topics that are talked about there), server B doesn't like Y and server C doesn't like Z your only choice is your own instance and that's far more effort than most want to go thru
Maybe it's kind of there in how instances interact, or the post visibility features, or some of the forks like Hometown, but Reader had that right-feeling combo of simplicity, personalization, and built-in network.
The technology to do this existed circa 2000 but I’d never seen anyone give it a serious try. One reason for that I realized is that, for a centralized service, you can get better recommendations based on other people’s preferences.
Assuming you have a huge user base, users will perceive a collaborative filtering interface learns better. My current classifier takes 20 or 50 judgements to start to understand that I like news about the NFL but I don’t want to see articles about that other kind of football.
The content-based model produces a good relevance score (“do I care about the topic?”) but it has no indication of quality apart from that. (Consider how Google is ranking both on query-document match and a quality score.). One input to my system is computer science papers on arXiv and it shows me more papers on ML than I can possibly read, never mind seriously comprehend the abstract in all cases. The relevance score does a great job of vanquishing theoretical CS papers I can’t stand but I wish I saw the 30 best ML papers and not the 30 most tropey. It is gratifying when it shows me papers about how to build this kind of classifier!
After all this I have a lot more respect for collaborative filtering despite how dangerous it is socially.
I don't think Fediverse should necessarily replace RSS, but there's something nice about being able to retweet / like posts in a somewhat agnostic context, just as there is having the option to "follow" feed sources in RSS readers if you want that.
There are a number of Twitter feeds I subscribe to in my RSS reader using nitter.
Ken Sheriff posts a blog post every few weeks. I’d really like to get notified of his blog post the instant he posts and and be the first to post it to HN. To get that kind of immediacy I’d have to poll his blog at an unreasonable rate. His blog is just one of 100-1000 independent blogs that I’d like to follow.
Maybe drop this pointless karma farming attitude, and you’ll realize polling every ten minutes is totally fine.
My current system runs on a 24 hour cadence of training models and getting judgements from me so I don’t need faster than 24 hours updating.
Mastodon (or rather ActivityPub) fixes this by pushing out updates.
What other instances do you know of with a longer character limit? Can we see how such posts look on other instances with shorter limits?
"Hey, see the stupid thing that happened to Eli Lilly? You know how you can actually just have your own email instead of gmail in your company if you pay for it? Also your own website? This is that, for what Twitter does."
Seems positively stupid for companies to NOT go in this direction.
> it's just like email, foo@gmail.com becomes @foo@gmail.com, that's it!
Is pretty easy to sell. The issue there is that in any "real" email host (ie not in your closet) you have no problem[1] emailing another host. Yet with Mastodon it's rather dependent on the specific of platforms blocking rules. Which isn't to say that this is an indigestible concept for the average person; just that it's a rough edge, it needs some work, some transparency.
Likewise how federation works is also a bit obtuse. The speed of nodes federating to each other feels inconsistent on a UX, and imo you should basically _never_ miss pieces of a conversation. Or at least if you do, it should be very clear that there is a blocked comment, or a yet-to-be-synced comment, etc. Right not it can just look like someone is replying to themselves or etc.
Really though it's all quite solvable with some UX it seems. There's some tech issues too of course, but all together this feels quite a feasible "twitter but like email" tool. I'm quite happy with it, and i'm even happier to ignore Twitter.
[1]: generally speaking of course. It's expected to just work.
Spam filtering and intense host allowlists and denylists are just as bad in email. In many cases it is actually worse because the number of "real" email hosts has shrunken to a staggeringly small number (most of which are major corporations) in some situations and self-hosting can be quite fraught with hidden traps and accidental denylist penalties and an increasingly baroque maze of security standards like SPF and DKIM and… From that perspective Mastodon is currently far more open in federation than today's email.
As much as anything "email is open" and "Fediverse is subject to blocks" is a marketing problem more than anything. Email as an internet foundation tech has decades of accidental propaganda versus reality. Mastodon has a lot of loud critiques from people that either don't want it to survive or have some other political game they are playing.
> Likewise how federation works is also a bit obtuse. The speed of nodes federating to each other feels inconsistent on a UX, and imo you should basically _never_ miss pieces of a conversation. Or at least if you do, it should be very clear that there is a blocked comment, or a yet-to-be-synced comment, etc. Right not it can just look like someone is replying to themselves or etc.
This is rarely a "speed" thing and in email analogy terms much more a "BCC" problem. Your instance was not BCCed on the other content. I remember ages of email when mailing lists were far more plentiful when these sorts of "BCC issues" were quite common. If your mailing list wasn't BCCed on a reply but then the reply of the reply BCCs it, there's an obvious missing piece in the conversation.
Mastodon UX could absolutely improve, but in this case the instance generally has no idea there is a "yet-to-be-synced comment" because there isn't a "yet-to-be-synced comment", the instance never saw that comment because it wasn't BCCed on it in the first place. (Though yes, this analogy breaks down a small bit because some of the "mailing list" posts come from RSS-like feed syndication, but there also there's no easy way to "guess" what might be in a feed the instance hasn't synced.)
Also, some of these BCC issues are intentional, for the same sorts of privacy reasons BCC itself exists in email. Mastodon offers some privacy settings for how far messages travel and some users don't want to "BCC the world" with some of their posts. There's never going to be a "complete view" of many conversations because people want some privacy in their conversations and Mastodon allows that.
Agree entirely. I've argued (in recent threads) that Mastodon, just like Email, is not meaningfully Federated because ignoring this problem is what caused Email to be not meaningfully Federated. "Meaningfully Federated" as in, Email encourages centralization.
More seriously, there’s nothing in particular that requires complete centralisation for success, though a future billion person ActivityPub network probably relies mostly on a few giant hosts (a la gmail) rather than millions of small ones.
The lesson we should have learned about alternative platforms is that they will attract power users and enthusiasts. These people will be very vocal about why everyone will switch to the next big whatever. A lot of enthusiasts and power users will switch, and a big flood of people comes over to the alternative platform like the tide coming in.
But normal users don't really give a crap about Twitter politics. The people who don't think about Twitter as a community or wider thing and just want to retweet funny cat pictures or whatever. They'll never move. It is a lot of effort for something they just don't care about.
And the enthusiasts? Well, now they're dealing with a fragmented community, where some of their contacts are on one service and some are on the other but a majority are on both because let's be honest, all the new Mastodon folks aren't burning their Twitter bridges behind them.
They start out posting everything on Mastodon. Then they realize they're leaving lots of people out, and they start posting everything on both platforms. Then slowly, one post at a time, the laziness sets in and before you know it they're posting exclusively on Twitter and wondering whatever happened to that promising Mastodon thing. The tide goes back out.
Your mistake is thinking Twitter users were normal users to begin with.
Twitter users have always been weird. There's a reason the user base never really grew after hitting a certain plateau. It's always catered to chronically online, media savvy, news obsessed types.
They are by definition enthusiasts.
> The people who don't think about Twitter as a community or wider thing and just want to retweet funny cat pictures or whatever.
You're describing Facebook and Instagram, not Twitter.
> They start out posting everything on Mastodon. Then they realize they're leaving lots of people out, and they start posting everything on both platforms. Then slowly, one post at a time, the laziness sets in and before you know it they're posting exclusively on Twitter
People weren't even doing that with Twitter.
The people who actually made Twitter successful--the celebrities, news sources, "influencers, etc--have always had a presence on multiple social media platforms, whether it was Facebook, Instagram, Snapchat, Tiktok, etc.
If Mastodon is successful, it's just gonna be another one of those platforms.
And hence, this article.
i believe that could be a middleground : a mastodon server hosting service.
Twitter makes it needlessly painful to scroll down these days, but historically while I never got a twitter account I could still check e.g. my public utility outage notices on Twitter if I had to. Facebook is harder or impossible these days to view any content without signing in. Where does mastodon lie?can content creators choose open policies?
It's easy for me to view and scroll his feed despite not having an account on his platform. Also easy for me to find his posts via Google. And (in case it's not obvious in your browser) it's also easy for me to find the automatically generated RSS feed of his posts.
YMMV.
https://www.tbray.org/ongoing/When/202x/2022/12/30/Mastodon-...
But if it's open, then you most probably even have access to an RSS feed of a user's public posts, so it can even be checked from outside the browser (and the specific app)
https://mastodon.social/@Gargron.rss
Of course it's all up to the instance owners.
I am hoping that companies will do a more "natural" implementation of ActivityPub. By simply making their existing services talk the protocol.
Then you could follow Medium authors via ActivitiyPub, comment on their articles and exchange messages with them.
ActivityPub, which is the thing that actually powers the fediverse, is an open standard that anyone can implement on their own.
I can understand the tendency towards cynicism, but in this case it's probably misplaced.
It won't matter what anyone else does with the code or protocol after that, any more than attempts to compete against Google, Twitter or any of the browser makers matter now. The big players will simply take what they want and make it theirs, or else destroy it at their leisure, the way they did with the web, javascript and any number of ostensibly free and open platforms and protocols.
Open platforms and protocols are incredibly powerful and resistance to balkanization. Heck, I find it really ironic you mention "the web, javascript and any number of ostensibly free and open platforms and protocols" as examples of things that companies supposedly made theirs or destroyed, given the web itself is more open and standardized than it every was. If anything, Microsoft's failure to push proprietary solutions/extensions into the web (HTML extensions, JScript, ActiveX, Silverlight, etc) is a great example of how resilient the web truly is.
Yes, they're used as a substrate for building walled garden services, but how do you interact with those services? Using open protocols implemented by interoperable clients and servers.
Regardless, I'm not sure what your point is. Email is still very much an open ecosystem with plenty of interoperability and competition. I myself use a custom domain with a hosted email service that isn't being run by Google or Microsoft (I opted for mailbox.org).
The spam problem means that it can be tough to ensure your server is being treated as a member in good standing--you need to implement things like SPF/DKIM, you need to ensure you don't land on any blackhole lists, etc--but so long as you do those things, you system will absolutely interoperate with any others.
There may be open standards, but Google and MS only want to interoperate with other big players. They don't want to bother with small fry.
And the whole point is that if you don't agree with how an instance is being run or moderated, you can migrate to a different instance or stand up your own.
Now, it is notable that a lot of the fediverse is concentrated in certain large instances (e.g. mastodon.social). Bu that kind of concentration is not itself an evil so long as people have the freedom to move around. Closed off, walled gardens are the problem.
Regardless, which is better: A federation where, yeah, there might be a bit of drama over who blocks whom, but each instance can set its own policy? Or a single private entity controlling the entire platform?
You cannot solve the problem you're highlighting with technology. What you're describing is the human consequences of people exercising their freedom of association.
But what you can do is place that freedom in the hands of the community, and I struggle to see how that is anything but a definite improvement over the current status quo.
For example, people generally don’t care about data ownership, and that’s one of the main reasons Mastodon is designed the way it is, and causes to problems it does.
No, it hasn't. Where did you get this misinformation? That flawed Guardian article that only covered a small period in December when the entire web was less active?
Even if you believe their flawed numbers, Mastodon has over a 70% retention rate, which is fantastic (2M MAU down from 2.8M, though, again, their analysis is deeply flawed for a number of reasons). Way back in the early days of Twitter, the media was claiming their retention rate was a mere 15%.
> people generally don’t care about data ownership, and that’s one of the main reasons Mastodon is designed the way it is
No, it isn't.
The main reason Mastodon is designed the way it is is to ensure no one entity controls the platform, therefore preventing any one single person from censoring or promoting a specific agenda.
That's a huge problem for many people, in particular the many journalists Elon banned from Twitter and never reinstated.
As for Mastodon's numbers, you can check for yourself, and it isn't good[0]. December 12th has 2,163,975 active users for a total of 5,474,864 Mastodon accounts, and on January 12th, Mastodon has 1,731,986 active users for a total of 5,940,022 people who created an account. That’s a 20% decrease in active users in a month.
That's objectively bad, and it's unsurprising. It can't scale, and solves problems most people don't care about.
Mastodon is not the future.
LOL. 80% retention is bad, now.
I cannot think of a more comically skewed view.
Most tech startups would absolutely cheer at an 80% retention rate.
I'm done with this conversation. I don't personally understand your motivations for wanting to undermine Mastodon, but it's obvious you're engaging in (bad) spin to push a preconceived narrative.
But according to you, disliking Mastodon as a Twitter replacement can only be possible if I'm trying to execute some kind of psyops against Hacker News. I can't just dislike the idea of decentralization to enable hate groups or a technology that hasn't risen to the demands of its new users. It has to be some kind of vast conspiracy, some kind of manipulation, for someone to disagree with your point of view...
Get a grip. I backed my statement up with facts, and you respond with conspiracy. I think that makes my point all the stronger.
And Mastodon is a pig. As software I have to say I am not that impressed with it at all. I appreciate it for what it is, but it is not quality code.
It's still rough around the edges, but is actively developed and has an engaged and welcoming community.
Making a change impacting user experience in a product like Medium is a months-long/multiple comitees process.
I don't think he ever intended to purchase Medium, but he did tweet about purchasing Substack at some point. But then again it's Elon, he tweets a lot of things he never seriously intends on doing and he just killed Revue (more or less Medium/Substack competitor).
Medium's entire model has been centralizing blogs and writing to accrue profits in a centralized c corp.
In this case, Medium is standing up their own invite-only Mastodon server. Users on that server will be able to interact with users on other servers and vice versa.
I wish we would stop thinking about this from siloed / walled off perspectives. The real game will start when cms devs put the protocol in as a first class citizen.
I see a lot of posts but it's very rare that there's a reply or conversation in even 1% of them
Still love RSS :D
Less than 1% of tweets have replies or conversations. Most of what people have to say just isn't interesting.
The interaction we used to see on Twitter is largely dead as well. Few people with followings in the thousands will engage with someone outside their small circle. Many block for trivial reasons. It's not what it was a decade ago.
Don’t confuse:
"I’m discovering it now" with "it’s new and will not last long"
and
"Because I’m new, I don’t have the code thus very few engagement" with "there’s no engagement there".
You don't think mastadon (sp) will be around for long. How much longer do you think it has? Mastodon began 6 years ago and has seen nothing but upward growth since.