Fediverse/As site of research
🌊🏄
role of creative practitioner in the context of the fediverse
Main phd question: How is online federation configured in alternative social media?
configuration: co-shaping, how it shapes and is being shaped
both a concept (longer term ideas around federation) and design material (in tech context)
two approaches in phd:
- geneaology: how ideas and practice have evolved over time
- practice based research (configuring)
- reflection on post.lurk.org involvement
- and two participatory design projects (2 new platforms based on the fediverse)
social media as product, which looks seamless and coherent from the outside
social media as ecosystem of different parts
USERBASE
ADMIN
SERVER
ADMIN
APP
DATA
DO
a social network as a smaller interconnection of parts, rather than a single huge central concentration.
currently:
12 million accounts
1 million active acounts
(means active once a month)
27.000 servers/instances in total
fedidb.com
for an overview of different types of software you can find
post.lurk.org
hcommons.social
tldr.nettime.org
social.edu.nl <– experiment by the dutch ICT association for the higher learning
SERVER
SOFTWARE
APPS
while Mastodon is the most well known microblogging app, there are different kind of project, like Lemmy, Pixelfed, Misskey.io.
more variety in different apps, such as phanpy.social, which comes with a summary function which is configurable (specify a time frame)
lurk.org
intentionally stay small
started as a way to give people an interesting place to hang out
managable moderation work
staying small both to create a space that stays interesting, and also to keep the moderation work manageable
“fun - manageable - interesting” <– nice!
lurk.org/TOS.txt
“interesting safe welcoming” intentionally vague, as this means something else to different groups
on which terms this is offered to people, an agreement.
in relationships of hosting there is a lot of trust involved.
these practices are not unique to alternative social media
but what is unique is… that these documents only apply on the local server (aha! this is our question)
you host the data from your own userbase, but also host data from federated servers (making copies of that)
serverblock/defederations
Qs:
-What's the workload -When does moderation happen -Can you expand more about the trust part -What's day to day of doing this -how long does it take to moderate who is choosing the moderation team -how often does the moderation happen when does it happen? -what do we do if you are in convo with someone form another instance? -how does the process work, do you talk to all 6 members of the moderation? -how do you chose who moderates -referrals? -does the moderation team do server maintenance as well? -what happens if there is a conversation w someone from another instance, which is problematic, whose TOS do you use? -Do you update the code of conduct?
easy answer: it’s all up to you..
it depends…
LURK started with 2 people.. now 6
people who are moderators are people they know personally, people offer help on lurk but the core team does not need it because they
need of trust relationship
there is not a lot of work, because people on lurk.org know each other or kind-of-know each other, that really helps
this makes it an exclusive community, which sucks because it’s hard to come in, but it allows to make it happening
it does not cost much time until something happens
in reality, it depends a lot on context:
what can take quite some time: ex of previous experience someone posting about hurting themselves, at these moments you need to understand your positioning. How to deal with these extremes has been learned on how to deal with this in the hard way
Timezones, if mods are in EU things might go south in America for example, you wake up to n. unread horrible messages.
Being available is not plannable, but this does not mean that you are available 24/7
You can only act on your local instance, when you as a user or as an instance block someone, they don’t disappear, you don’t interact with them.
local concerns
If everyone is on the same place, the center can delete it
Moderation being 6 peeps, trust comes up again, you don’t roundtable all the time because there is trust underneath.
if anyobe can make an account and join the situation with moderation might scale, inviting friends or friends of friends is a way to prevent things.
There is a shared blocklist of other instances as well.
another thing to prevent too much moderation work is shared blocklists (like fediblock)
The TOS gets updated constantly.
With a lot of this you cannot plan or imagine from the start all of this casistics. An existing code of conduct can help you navigate this, but there’s always different specifics. The main mode is to “learn as you go”
post.lurk.org for Roel as practice based research
engaged in this environment, not to study lurk, but to learn more through encounters
lumbung.space
on invitation by indonesian collective ruangrupa, during documenta15 in 2021/2022
festival of artistic self-organisation, one of which was to prototype an online platform
functioning as a publishing outlet and archive
a collection of applications that could be tried
that could all publish to the same front page (using RSS i think… just saying)
to make this an infrastructure that other groups can also use..
there was a collaboration with co-op cloud
https://autonomic.zone/blog/2023/02/co-op-cloud/
news.lowtechmagazine
forum for paid subscribers of the magazine
heavy ui redesign
was never published, because the moderation tools were not good enough
In essence, alternative social media allow you to take it, try it, modify it, weird it pulling apart those aparently seamless layers hidden by centralised social media.
one of the layers (presented in the drawings).. can be enough to focus on
USERBASE/ADMIN/SERVER > etnographic research
some instances might have very democratic approaches
in the case of LURK, calling themselves a benevolent dictatorship, and that’s how it is, as it is the way to keep it manageable, it’s intentionally non-member based, to make it work
computational research
using scripts and research to collect data
can COC’s be compared for example? are they all really so different??
design research
exploring different possibilities in apps
how to make these more useful, interesting??
what is missing??
practice based research
engaging with a particular practice
artistic research
where is the creative potential here?
for example: write different apps to connect to mastodon
an interface that allows you to only post at night
one that only accepts images with a specific size
Questions
How does LURK organize the server work?
Does post.lurk.org consume a lot of space??
Lurk is more than mastodon. At mastodon, it is mainly Roel. Lurk also runs email systems, backup systems, …. and the tasks are split across people’s skills and availability
In terms of space..
To run Mastodon, you need different things
- processing power
- memory
- storage First two depend on how many people and how often use
700 accounts, 300 active was running on a 12 year old computer
Issue of storage is one of the harder ones… the design of the software works by making copies.. copies of posts from other servers.. everything that you see on your server is copied from other servers.
Everything piles up constantly… since at lurk they don t have infinite…
LURK deletes all the remote media older then 4 days, not the posts, but the media. Posts are kept longer, 2 weeks/1 month.
How a big a server is, is connected to the community of it, as the workload is also connected to the revenue model.
customization questions.. what are customizations that are done at post.lurk.org?
allow more characters, default is 500
hometown
custom emojis get suggested by users
It is possible to modify open source software, but then how to maintain these is questions for the people who do it, at lurk they make it so that there is less work to do as possible.
artistic work based on the fediverse?
Darius Kazemi, maker of Hometown, made a Mastodon instance in which you can only use the letter e –> dolphin.town
On twitter, there used to be a lot of artistic bot practices.. when Twitter closed their API down, a lot of these things disappeared
art bots
why did you choose to go for an academic context to do this research?
i was looking for paid work to continue my practice within
any other positive or negative aspects of working in this context?
longer moment of time to focus one topic
how do you think it shapes the research topic in itself, and who it is directed to?
first it was surprising to find all the things written about design but never reach the people who practice this
it’s less freewheeling as an art and design practice can be..
but intentionally made space for working practically
in terms of accountability…
the work needs to respond to funders.. where practice is still hard to be acknowledged as such
and also respond to the people you work or have worked with, like the people who i worked with to make those prototypes with
What research do you see around you you think it is interesting to look at?
Do you see research happening on the fediverse?
Yes on many different levels.
The communities and groups running instances.. this is research in itself
through specific interventions (tocs, software modification etc.)
there is quite some collective push back against Threads (made by meta)
figuring out as they go could be considered research in itself
another type of research: useful additional tools that are being made
When twitter was bought by Elon Musk, some people build tools to find people’s @s to automatically search and followed them away from Twitter, onto the fediverse
automatic follow gadget, to find all the followers on Twitter that also used their mastodon handle as username
the software itself does not really give you much different… so it repeats a lot of the same patterns so the changes are made by people in the end.
rethinking social media as a federated open web (!)
another type: alternative social media scholarship, working on this for longer, not only fediverse related
mostly coming from communication studies and journalism, not so much design
More funny/weird specific things (gadgets, bots, scripts)?
examples of bots/scripts, taking RSS feeds as an input and posting to Mastodon as output
Treed (mastodon client dzwdz.github.io) makes different pages, for days, against the feed. This shows how much can you gain with just a few tweaks. You cannot post with this, you basically get asked permission to read your mastodon feed. (python or js mastodon API)
phanpy project (again) but they are more “make social media better?” not “question social media”
Do you know anything about Glitch-soc?
Glitch-soc, it’s a fork made back in the day done by people who wanted a fork to try different things, very internet culture based. THey use styling similar to old forum posts.
When the project was still new, around 2016, there was much more work on forks around Mastodon, to try different things, and one of them was Glitch
one of the main people behind Glitch is now working at Mastodon actually, it worked as a short cut to a paid job ;)
Glitch supports local-only posts
and has support for forum-style short cuts..
Mastodon does not want to support local-only posts
This is the bibliography of alternative social media scholarly work: https://www.socialmediaalternatives.org/bib.html
connected to this is a zotero group, which you can use to search
for example if you’re looking for work around content moderation
keywords: #data-ownership #self-organization #agreements #learn-as-you-go #benevolent dictatorship #tweaks #interventions