Hello, and thank you for listening to the Microbinfeed podcast. Here, we will be
discussing topics in microbial bioinformatics. We hope that we can give you some
insights, tips, and tricks along the way. There is so much information we all
know from working in the field, but nobody writes it down. There is no manual,
and it's assumed you'll pick it up. We hope to fill in a few of these gaps. My
co-hosts are Dr. Nabil Ali Khan and Dr. Andrew Page. I am Dr. Lee Katz. Both
Andrew and Nabil work in the Quadram Institute in Norwich, UK, where they work
on microbes in food and the impact on human health. I work at Centers for
Disease Control and Prevention and am an adjunct member at the University of
Georgia in the U.S. Over the past week, there's been a huge change in how
scientists consume social media, and that's hit our fields with a tidal wave,
more or less, and our very own Nabil is at the very center of this. So, Nabil,
can you tell us what's gone on here? Okay, so what Andrew's alluding to is the
migration of academics on Twitter to a different platform called Mastodon, and
there's some interesting nuances and changes in that that I think we'll talk
about as we go to the episode, but before we get started, I'll probably just
explain what Mastodon actually is for people who don't know. So Mastodon is a
free and open web application that runs on a web server. So it's like a self-
hosted blog or maybe you've been to a bulletin board, a forum webpage or
something like that. It's basically the same thing. It's just an application
running on someone's computer somewhere. And the layout of the site is it's
designed very much for microblogging. So small, short, condensed posts from its
user base, very much like Twitter. There is, what makes it different though, to
other sort of insular things that you might have run into like web forums and
things like that, is that the different servers that people set up can actually
integrate and communicate with each other. So if you're a user on one, you can
look up a user on another and you can follow them or reply to them or read their
content. So that's, but that's basically it. That's all that's happened. That's
all that Mastodon is in a nutshell. I missed a step. Why are people migrating
from Twitter? So I think this was on the backend of Elon Musk buying some
fraction. I don't know what, how the ownership works, but he's bought some
fraction of it and he's started to make radical changes. People don't seem to
like Elon Musk for a variety of reasons. They seem to believe that Twitter prior
to Elon Musk buying it was somehow maintaining free speech and some sort of
something to do with democracy. I have no idea. And they fear that that's in
jeopardy because of Elon Musk's involvement. I don't know. For me, it's a
website where you post cat pictures and trash talk people, trash talk
celebrities. It's a private entity. It has no, it's not a universal human right
to be on Twitter, to have your freedom of speech protected by Twitter. Like,
like that's, it's just a webpage. It's just a website anyone can make. And if
you're a bioinformatician and a programmer, you're quite capable of making your
own version of Twitter. There's nothing magic about it. I've been using Twitter
since 2007 and I've found it very, very useful over the years for sharing papers
with other scientists, particularly, and, you know, for actually interacting
with the community. And I've made a lot of friends actually in real life through
Twitter, which is a bit mad when you think about it, because it is just a, a
random social network. Sunnabeal, where did this all come from? I know yourself
and Duncan McCannell have set up your own server. Yeah. So on the back of all of
this furor over Twitter, different people, including myself and Duncan started
setting up, started playing around with Mastodon and we decided to set up our
own instance. It was, I think Duncan said, you know, like, this is a thing, like
people are playing with this. We should have a go. He bought a domain. I set up
a virtual server and we just put up an instance of Mastodon in a couple of days.
It's Mastodon.science, so M-S-T-D-N.science. That's the, that's the web URL. And
we just decided to play around with it. We said, okay, well, this could be a
technology. We always like tinkering with things. And we thought it'll be a
place where some bioinformaticians, microbial genomics people, some
microbiologists who are quite tech savvy would come and hang out and post some
messages and share some memes and have a good time. And probably like 50 people,
a hundred people, you know, like nothing great, but just something to play
around with this technology. And in the back of my mind, when we were setting it
up, I was thinking, this is really interesting as an exercise because, you know,
playing around with it, seeing where this could go is you could use this. Let's
say you have a conference. Let's say you have a, an organization and you set up
Mastodon and you say, this is like a private social network, Twitter-like thing
for your participants, for your employees, for your, for, you know, you could
have it based around any type of organization or topic and just have a separate
space for that sort of walled space just for those people. So I would be
interested to hear if a research council uses it or a university uses it or an
interest group. And our thing was an interest group around microbial genomics.
That's why we set it up. And we weren't thinking it was going to be very big,
but then something happened. There was some sort of stampede of people who
wanted to get out of Twitter and try this out and find a separate space and
connect with other people. And it wasn't just us. There were a number of other
instances were set up. There's one set up called genomic.social, which is just
for genomic science. There's one called coevo.social, which is for ecology and
evolution. There's existing ones like vettyscience.org or scholar.social or
scicom.xyz. And all of these instances, new and old, suddenly got a barrage of
people trying to join up. I mean, at the moment we have about 2000 users that
have signed up for just the instance I and Duncan set up, there's 2000 for that.
But there's probably maybe several other thousand users that have joined the
other instances around science. And a lot of scientists have just started to dip
their toe into this and see, is this like a thing? Should we be playing around
with it? And that struck me as really odd. You know, like, OK, well, let's just
ride the wave and see what happens. So we just keep throwing more resources to
the server. It's a virtual service. You can dynamically resize it. So I mean,
you don't have me going to a cupboard somewhere and swapping out disks. I don't
have to worry about that sort of thing. But it's dynamic. We can resize it well
enough and things like that. And it's fairly manageable to just accommodate more
and more people. So all right, fine. Let's see what happens and see whether it
could work. It might not work. It might just crash and burn. Probably everyone
will get bored or maybe there will be a core group of people that keep using it.
I have no idea. But let's just try it out. What's the worst that could happen?
So what are you up to on your virtual server? Like what are your resources? And
how many users is that holding? So we've got almost two, at the time of
recording, we've got almost 2000. It's a, oh, I've forgotten. It's a four gig, I
think it's a four gig memory and four vCPUs, virtual machine that's running the
most of the content, like static things, like all the pictures are being put on
a object store bucket somewhere. It's not very intensive. I mean, the most
expensive thing is actually getting a mail server that works. Sending emails to
people has become incredibly complicated, it seems, because of the amount of
spam out there. So there's a lot of controls to try and stop spammers. And that
actually makes it difficult for an honest person like myself, who just wants to
send an email to someone saying, hey, if you want to reset your password, just
click this link or, hey, you've actually been confirmed for the website that you
yourself signed up for. That's actually the most complicated part. But that's
the resources that's going on in the background, really. It's all fairly light
and easy. So how much is that going to cost? You know, like, is there a rough
guide to how many thousand users and how much it's going to cost every month? It
depends. It depends because it's not just the user count. It's actually the,
because it's like a network that's integrating with other instances. It's also
the amount of interactions and following they do for other users elsewhere,
because you're spending this resource of trying to cache what other users are
connected to. So it's not like a strict linear thing. I will say it's about,
it's about a hundred pounds per thousand user to a point. Wow. That's quite a
lot. And I guess that's not something that you realize. So when you use a free
social network, particularly because, you know, they're making money off
advertising and off your data. But that's, that's quite a substantial amount.
Yeah. It is a substantial amount for one person to carry just like by the by,
like just so everyone else can have fun. Yeah. I should mention the platform
platform doesn't actually give you any system of.  adding in ads, even if you
wanted to, not that I would, so it's all ad-free, but someone has to pay the
price for running the server and yeah, it costs money. Yeah. So what is your
model of, of paying for it long-term? I have no idea. Some people have been
very, very kind to just chip in some money, which I really, really do
appreciate. They know who they are, but yeah, I don't know. Most people for
other instances that are longstanding, there are these sort of Patreon-like
monthly subscription that sort of work on a donation basis. That seems to be
what people tend to use. There is one, there are a few of these, which are more
like very transparent and the design specifically for NGOs and charity sort of
endeavors, like foundational kind of things, which are a bit more formal, but
look very good in terms of transparency and things like that, so there are
platforms to, to raise the money. So I'm not too worried about it. Well, all the
financial concerns aside, I mean, this is really exciting and I've enjoyed it a
lot. Do you have certain people that you were excited and surprised that they
came on, or do you have certain toots, as they call them, that you liked? So I
don't watch the signups for other instances, but I have noticed at least in our
tiny, so this is one instance out of many where I'm seeing this just for, just
for our own, but we've had a Nova Laureate join. You've had a lot of journals
suddenly show up. I was surprised about that. I was surprised about the number
of scientists from other disciplines who sort of hopped on it as well. And
there's very interesting posts that I would not have seen on Twitter that I'm
starting to see more, more of, so things from things, people from pharmacology
or immunology or people from even history, things like that, who are rocking up
and making accounts and posting content. And it is actually interesting to read
what they're up to. I was never really subscribed to that in Twitter. Yeah. I'm
sure you would never kick anyone off. No. I know that you've had some technical
challenges, like I know the other day you had some kind of issue with pictures.
Are these easy to deal with? Yeah, it's easy to deal with. You just have to get
your head around. The problem is, is that there's so much, because the user base
is growing very quickly, it's about, it was like two or 300 people a day jumping
on. And then each of those would, would spawn a whole bunch of activities. So
you had to be quite, what was happening in the background sometimes was that in
the evenings, a lot of people would, would hop on and then basically the web
server would have run out of threads, it wasn't able to keep up. It's, it's set
up quite, it's, it's quite robust in the backend. It's not going to crash. It's
designed in a way that, well, I'll explain the implementation and explain what,
then you'll understand why things get stuck. So the server is written in a mix
of JavaScript and Ruby. It's basically run as a web application with the
Postgres QL as your database backend, and it has a Redis cache, which is sort of
a in-memory cache of things sitting up front so that it can quickly serve
content to people. This, if you're someone who's done web, this is, there's
nothing bizarre or strange about it. It's all very standard stuff. There is a
background process called sidekick, which sits there as a demon that takes on
all of the sort of background jobs and processes it as well. So you've got that
other background process running as well, but that's, that's it. You have your
web server process, your sidekick, which is your background process and your
database process running at the same time. You said it was a hundred pounds per
thousand users, give or take, and that might explain it too. Like there's a,
there's a process happening all the time or a few processes. The web server
itself has processed about 1.5 million jobs in a week. It's doing a lot. What I
didn't realize at first is because, because I see like us as users posting and I
think, okay, well, there's not that much traffic, it's sort of chugging along.
Okay. Then I, then you look at the backend and you realize, oh, wait a minute.
That's, this is the problem with this is that it's, it's trying to replicate all
of this content and pull content from all of the other instances. So it's
spawning a lot of jobs, even if your user count doesn't change very much. So are
you limiting which instances it communicates to, like just to like a little
science one, or are you taking in everything? No, I'm taking in everything at
the moment because a lot of people who are scientists have jumped on to more
generic instances as well. So they're on, so the biggest instance is called
mastodon.social. So if you hop on that, that's just sort of, but that's a free
fall of anybody, anybody who's anybody who's just on it. And quite a few people
in our community have jumped onto that and they've kept the account there. And,
you know, I don't see a reason to drop, to block domains. I have started
blocking domains, but this is not because of trying to limit the requests or the
access, but because those, those, those instances, those servers are serving
content that you don't want people to see. Oh, I don't think people would
appreciate it. This is, this is the wild west of the internet. If, if you've
been on the internet for a long time, you'll remember there is a lot of gray
spaces out there that aren't safe and you need to be a little bit careful and,
you know, because this is a decentralized network, they're also sort of, there's
also, they're also sitting there in the background. I mean, you can find bot
accounts and spam accounts and all of that on, on anything like Twitter or
Facebook or whatever as well, but it's sort of more upfront, it's more upfront
for me at least, because I'm seeing it as a, as an admin, I have to moderate it.
So I've blocked a few pretty dodgy, pretty dodgy servers. Yeah. So people are, I
think you said that we're migrating, people are migrating over from Twitter. Do
you think that people are using it as a total replacement? Some people seem to
have totally deleted their Twitter account and jumped over to this. I think
we've got to unpack that. There's been a lot of bad will towards Twitter as a
platform for a long time. A lot of people I think are quite tired. If you were
someone like, like for instance, me, like I'm quite fed up with Twitter, if, if
you were someone who joined Twitter circa 2007, 2009, because you were a
professional, someone in tech, someone in science, Twitter was a very, very,
very different place. It was just the people you could find, you could follow,
those people would retweet and you would see their content. That was it. There
wasn't any promoted content. There wasn't really any ads actually embedded in it
that much. And you would just see the scientists. So you would hop on and you'd
see your, you'd, you'd just see like scientists saying things. And then it
became what it is, which is more of a news thing or a gossip channel, and that
was more, started to get away from why we joined it in the first place. So,
because on the back of that, I think a lot of people just said, yep, okay. This
is a sign, there's somewhere else to go. I'm going to kill my Twitter account
and I'm going to leave now because I've been fed up for years and I'm out. I
don't think there's a single, this is like a singular event that's, that's
driven people away. This is a sort of straw, final straw. Emma, why did you jump
from Twitter over to Mastodon? For me, it's a little bit of hedging. It's, I'm
not sure what direction Twitter might go, but the idea of having something that
can't be fully controlled by one person that is more federated, that if one part
of it becomes bad and you can shift that weight onto other parts, that seemed
like a really good idea right now. So I'm still not really sure what's going to
happen. I don't have any plans to leave here at the moment, but it seems like a
good idea to have a backup. And it looks like you're actually in Hingston now. I
know that background. I am indeed. Yes. What, what is the background? How do you
recognize it? Yeah, that's pretty impressive. I'm just in like a corner. I
worked in Hingston for a decade, so I've been there many times. The bricks are
very individualized. We went over a lot of the, of the inner workings of
Mastodon and how Nobile's running it and the costs and compared to Twitter and
everything. And so it's really nice to hear your opinion. I think Emma raises an
interesting point though, that there are plenty of other social media networks
people can go to, right? We could have all gone to like LinkedIn or something or
whatever. There's plenty of places you can go. But I'm interested in this, this,
this is not like, Emma's not the first person to say that it's interesting that
we've changed the architecture of the, of the platform, it's a decentralized
platform. The power is in, you know, if like, let's say you get kicked off an
instance, you can create your own instance and you can start participating in
the network again. Like you're not, or if you don't like what an instance is
doing, you can go to another one. So the power is, there's a power shift in the
content you want to see and who you want to interact with that's not available
in a centrally controlled platform, and I'm curious if that's for everyone, if
that's why that has suddenly become really important or what, what changed all
of a sudden? Cause that's always been there. That's how the internet always
worked. The internet has always worked like that. I don't know if the internet's
always worked. Like it's, it's meant to be decentralized. Right. So, but, but
it's gone towards, yeah, like, like this aggregation of, of these sites, like I
don't go to a zillion sites. Like I used to in the nineties, I go to like the,
the five or 10 sites on my phone. It's weird. No, I think it's really strange. I
mean, what is the internet? It's a bunch of.  of other people's computers
talking to each other. I have like on the front screen of my phone, I have like
Reddit, I have Twitter, but now I have Mastodon and I think that's great. Yeah,
I read the Mastodon feed before I read anything else at the moment. It might be
because I'm just trying to check if the servers crashed or not. I mean, I wonder
if part of it is just that maybe we did need larger things to convince us that
these ideas were useful and possible and accessible because I simply was not
part of anything like Mastodon before Twitter or I mean, until 10 days ago. And
I'm not sure that it's something I would have necessarily gotten into before I,
before Twitter, which is very accessible, very easy to use, very well known,
kind of showed me what can be done with something like Microplug. So maybe these
bigger services have just done the hard work of introducing these ideas and how
these can be useful and how they can even be revolutionary to a broader audience
that we can then take that back. And we have the, I don't know, the interest to
make it work in a more decentralized way, which does take a little more work. So
you'd have to kind of know, is it's worth it? And maybe that's, I don't know,
maybe that's how it will all go, that we learn that these things are cool. And
then we decide actually we're willing to put the work in to make them ours.
Yeah, I totally agree with that. I think if you weren't raised in IRC channels,
you probably just, yeah, this would be your first time seeing that part of the
internet. It's a delightful part of the internet and I miss it a lot, actually.
It's good that people feel it, engaging with it and seeing the value there. And
it is limited. It does have the quintessential problem. The first thing people
complain about with Mastodon is the fact that it's very hard to find other
people and search for things and locate content. And it's like, well, yeah,
because on the old days of the internet, it was a skill to locate the good
memes, to locate the good jokes, to locate the good website, the informative
stuff, to be able to use a search. I don't know if people still get taught how
to use a search engine in school, like how to use Boolean operators when you use
a search engine, when you can sign up for a library training, get trained to use
this at the library. But that was like a thing, like you had to know how to find
the thing you wanted. And someone who was very, very good at it was actually
very valuable. And that's where data aggregation is sort of this fundamental
problem. And people might not remember that that's where websites like Buzzfeed
or Reddit or Dig or Huffington Post or whatever got famous and got rich because
they were able to do this very, very well. So I migrated over from Twitter quite
easily and brought a lot of my followers and they were all on Mastodon as well,
or a fair few of them. There's like websites out there where you can just kind
of put in your Twitter stuff and it logs in, you put in your Mastodon handle, it
logs in, and then it gives you a CSV file, which you just import and it
magically finds, it doesn't find everyone, but it finds a few people to get you
started, which is quite handy, particularly people who put their Mastodon handle
in their profile somewhere. I find it nice to start from scratch. I make a fresh
choice of who I want to follow. I've brought over my dozens of users. I've been
pretty selective. I mean, being on the admin side, you're sort of the dungeon
master. You wind up having to troll through all the stuff anyway. You don't want
to read all the backstories, but it's just sort of there. You have to kind of go
through it. You see all the emails and people signing up and you kind of go,
okay, fair enough. Is this the end of Twitter or is it just a temporary flash in
the pan and we'll all go back to Twitter in a few weeks, and we'll forget about
Mastodon? Probably, I don't know. We'll see. I think it's introduced a lot of
people to a different way of approaching social media. So it's a useful
exercise. That sort of education is always helpful. I think it will change.
There are mechanical things about how the Mastodon itself, but even others that
use this sort of federated decentralized network, because there are actually
other tools, other apps that will interact with the content. Realistically,
there's a whole bunch of platforms that are stitched together with just like
with API, RSS, API, JSON strings all over the place that are all talking to each
other. But the problem is with that is that there's a good latency involved.
Decentralized networks are always slow because you can't, for instance, you
can't guarantee that the other thing that you're interacting with, you say, hey,
can I have this post? And it doesn't respond. You just have to say, oh, well, I
guess they're busy. And then you wait. And then, so then you go back and you try
it again. So you have this latency with decentralized networks. And so that
mechanically, it can't replace Twitter because Twitter is designed as a central
fast resource. For instance, they could index all of their tweets and they could
give you a really good search engine. You can't do that in a decentralized thing
because you can't necessarily scrape all the data off every single server. You
don't have the control over the content to really be able to do that. So it
won't replace Twitter as a one-for-one replacement but it might fit certain use
cases that people are looking to Twitter for, which is a bunch of academics
slapping around and posting and complaining about papers, basically. I think
it's perfectly legitimate for that. Any final words from Emma? I don't know. I'm
maybe a little more optimistic. Certainly for me, the vibe that I'm getting from
Mastodon right now is one that I really like. I feel like it's more like Twitter
when I first got active on Twitter. And I think part of that is maybe just
numbers. There's few people there, but I also just think that everyone there has
kind of been... It's re-leveled the playing field a little bit. People are not
necessarily feeling as driven by trying to get into the algorithms that of
course dictate Twitter. But also it's Mastodon.science, it's what I'm on. And so
it's a lot of scientists and that's really nice. It's kind of back to being
about talking about science and science-related things. And that's what really
got me falling in love with Twitter way, way, way back when. So I think it could
really go both ways. I think Twitter itself, that's not... Whether that fails or
not is not gonna have anything to do with whether or not Mastodon really
succeeds. I don't think that's gonna be how that works, but whether Twitter ends
up falling will of course have influence on whether something like Mastodon
becomes even more successful. I think mostly Twitter's fate is gonna just rely
on what crazy ideas they come up with and continue implementing. But there's
been some pretty notable movement that I think that there was an article that
someone posted recently that was about this kind of losing that final barrier of
trust. And then everyone's just like, yeah, that's it, we're out. And I think
that may be happening. So yeah, maybe that will lead to more people deciding
that sure, Mastodon is not the same. It will never be quite the same, but maybe
in a way that's good and better. That's fantastic. I think we'll have to end on
that. So we were talking about Mastodon and thank you very much to Emma Hodcroft
for her special guest appearance. And we hope that you will check it out in the
future and maybe decide for yourself. Thank you so much for listening to us at
home. If you like this podcast, please subscribe and rate us on iTunes, Spotify,
SoundCloud, or the platform of your choice. Follow us on Twitter at MicroBinfy.
And if you don't like this podcast, please don't do anything. This podcast was
recorded by the Microbial Bioinformatics Group. The opinions expressed here are
our own and do not necessarily reflect the views of CDC or the Quadram
Institute.