• Home
  • Help
  • Search
  • Register
  • Login
  • Home
  • Help
  • Search
ASync.pre ASync.pre ASync.pre Forum software future: Async(-.pre)

Forum software future: Async(-.pre)
smuggler
#1
12-01-2020, 12:19 AM
Part of Async.pre's mission is to create a new "communication tool" (aka, forum-like-ish-esque) that preserves privacy, motivates good communication practices, is resistant against takeover and shutdown, and is "secure" in general.

@frankbraun is dreaming of a forum software that "can also run in the terminal of an OpenBSD laptop".

I'll start with a few random thoughts of how that might look like, in the hopes that other's contribute ideas and at some point we might even implement it.

# The state of communication tools:
Roughly there are five kinds of community communication systems today:
- Social media. Which has an unchecked audience (too big and too unknown) and that relies on using platforms that are outside of the community's actual control.
- Instant messenger tools. Great of chatting, horror for long term discussions where newcomers should have a way to read themselves in and there's a persistent body of previous posts in a structured interface. Hard for long posts. Too fast.
- Web/BBS forums. The software quality simply sucks, and none of them allows end2end encryption within groups (or even p2p). We went through a dozen different software products that were horrific each and every one of them. Hosting the least bad one still makes me wake up with nightmares.
- Threaded messaging (email and the likes): Does not cater towards structured, persistent corpi of messages but is at best a collection of threats. There's no common visual identity.
- IRC: No retention, no longer posts possible, too fast.

They also fall in two dominant client categories: Either only web-based or only app-based.

# The Changes of user behavior
In "the olden days" people would join a community like a BBS, forum, mailing list, IRC channel and spend enough meaningful time there to actually be in the loop. This simply doesn't happen anymore. We use mobile devices much more than stationary ones when it comes to non-work communication and computing. And we split our attention over dozens of places and platforms. Online communities are largely dying - especially for people who have more than one topic of interest. The fragmentation of time spent simply prevents people from investing what is necessary to create real online relationships. At best, we find singular friends online, but we don't "become part" anymore. I think this is potentially keeping every potential community from ever becoming deep and big enough to have any significant impact anymore - both personally and on a "society scale". We're swamped by so many voices that we can only focus on very few that already have proven meaningful. Leaders (in attention retention) are now becoming thought leaders.
This is both dangerous and a waste of our cooperative powers. We need actual online communities where new people can integrate and learn, but also participate and CONTRIBUTE. For that, we cannot assume that people invest their attention in a self-driven way but instead have to beg for attention like all other comms tech these days (notifications).

Furthermore the speed of the post-response cycle today is incredibly high. Social media and messengers are essentially synchronous. Add a big audience and they often do not allow in depths discussion over long periods of time anymore. This just reinforces the issue of attention-sharing. But everybody is used to it today, and potentially needs it even due to dopamine addiction. Breaking this can't be simply asking for change of user behavior but realistically requires support by the technology used. Slow down everything, and notify only when required to keep base attention alive.

# The idea of Async
Forums are places where long-form posts are long-term retained so they can become the common substrate for a community. This implies that the post-response cycle can be much slower, and potentially even artificially
slowed down. This much slower interaction also helps with integrating privacy-preserving functionality: Post-events on the network layer can be widely separated from posts visibly appearing in the forum. Combined with some basic encryption this already enables much more private (and therefor open?) communication.

Furthermore, dropping all the dark-UI stuff that essentially makes people into addicts would be a nice feature. Less images, less animation, less endless scrolls, less gestures/swiping/scrolling. Focus on efficient use instead
of high engagement. Slim everything down, focus on simple taps (or keyboard shortcuts). Don't invite memes. Have minimum post-lengths. Limit metadata. Limit interaction and distraction (likes/dislikes, stars, hearts, emojis).

Only ask for attention to maintain minimum engagement instead of maximizing presence: Allow people to subscribe to topics/categories/forums. Let them choose how often they want to be notified (per N posts, per N hours, only after N hours of not logging in, etc). Notifications contain necessary information for user decision and lead to subscribed content immediately. Allow quick temporary muting of notifications.

Separate forum-operators from content-responsibility: Whoever hosts the technical infrastructure does not have to have insight into all content. This also limits the impact of security events. End2end encryption of groups/topics/threats in the forum. The backend only has to enforce access rights - not know contents.

Allow the sharing of files, without creating a visual enticement. No thumbnails and previews. Do not support the inclusion of remote content (URLs automatically fetched, embedded videos etc). Instead, upload the content and turn it into the community's property and shared memory.

Give people a choice to use interfaces that suit them. The same backend can serve all of mobile app UI, web UI, and text UI (+whatever might come). WebUI is necessary for easy onboarding, App UI might be the choice of some, text UI is clearly the preferred choice of some of the founders here (@frankbraun, but also me).

Make sure that content does not disappear easily. Every community member should be able to easily retain an offline version of the complete forum. And potentially restart hosting it when the original host goes down. Or fork it and start a new community. Forums are community memory and as such potentially incredibly valuable. Don't risk that value to chance.

Now, these were just my most pressing thoughts on the topic and I hope for commentary and improvement. While re-reading it, I can't shed the feeling that it's a monastic like perspective. Long term, long post, community memory, preservation, focus, "purity", offline capability, slowness. Makes me think of Anthem by NS, or the Old Kartheuser, or the monastaries in Meteora or the St Katherines in the Sinai. I'm getting old....
Guy Fawkes
#2
12-01-2020, 09:26 PM
There has been a thread on the CPunk mail list, spearheaded by James (Jim) Dalton Bell, regarding what he has identified as modifications or deletions mainly to his historic posts dating back to the 1990s. Jim is the author of the paper, Assassination Politics (AP), promoting the use anonymous prediction markets to eliminate corrupt or otherwise criminal politicians, which got him in lots of hot water with the US Department of Justice. Indirectly, as a result of his AP paper, he spent 12 years in federal prison. He speculates that some of the alternations appear to have been intentional others a result of lost archives or bit rot.
frankbraun
#3
12-01-2020, 09:31 PM
(12-01-2020, 09:26 PM)Guy Fawkes Wrote: There has been a thread on the CPunk mail list, spearheaded by James (Jim) Dalton Bell, regarding what he has identified as modifications or deletions mainly to his historic posts dating back to the 1990s. Jim is the author of the paper, Assassination Politics (AP), promoting the use anonymous prediction markets to eliminate corrupt or otherwise criminal politicians, which got him in lots of hot water with the US Department of Justice. Indirectly, as a result of his AP paper, he spent 12 years in federal prison. He speculates that some of the alternations appear to have been intentional others a result of lost archives or bit rot.

Not sure if this is true, but important point about unmodifiable content. Post should be cryptographically secured against modifications.
frankbraun
#4
12-01-2020, 10:40 PM
Thank you Smuggler for writing this extensive outline.

I agree with you regarding the analysis of different communication
platforms. I think Web/BBS forums is closest to what we need, but
nothing available comes even close.

In what follows I wrote down some of my thoughts regarding an ideal
system, which should still be minimal enough to allow easy
implementation for bootstrapping.

Must have features
------------------

- persistent cryptographical identities for users
- private groups / sub-forums
- allowing for moderation
- multi-device capable
- long-term readability of content
- should be extensible for communication through mix-networks

I believe giving users the ability to create private groups /
sub-forums, but also allowing sub-forums to be heavily moderated is how
one solves the freedom of speech/quality dichotomy. With moderation one
can achieve high quality discussions (Hacker News is an excellent
example of this). Giving every user the ability to create his own
private sub-forum solves the freedom of speech problem moderation
produces. A private forum is basically like private property, the
owner(s) decide the rules.

Nice to have features
---------------------

While I would discourage heavy formatting of content, it would be nice
to allow some consistent formatting to make longer posts readable.
Therefore I would include something like Markdown formatting which can
be rendered nicely on the web and in mobile apps, while also allowing to
show it mostly "as is" in the terminal.

It would be great if editing of posts is possible (mistakes happen after
all), but in that case the entire history should be available.

Non-features
------------

The following should **not** be included:

- liking of posts
- follower counts (not entirely sure about that one)
- profile pictures
- embedded pictures of animated gifs (God forbid!)

I would go with UTF-8 by default to allow all languages but either
filter or strongly discourage emojis. Emojis do not really lead to
better discussions and they are hard to display correctly in the
terminal. Normal ASCII smileys ought to be enough. Or I'm simply too old
:)

Platforms
---------

(in order of importance)

- terminal
- mobile apps
- web app

I think the terminal route is the only way to make this software
_extremely_ secure, because then it can be run on OpenBSD and doesn't
need the entire crazy web-browser stack (and not even X).

Terminal is also what is the "coolest" in the sense that it is something
unique, more BBS like. Smuggler and I also discussed "alternative
terminals", but that's something for another post ;)

Afterwards mobile apps are the most important one in terms of usability,
that's simply what most people use.

Web app (or desktop with GUI) would be nice, but I consider it not high
priority.

Centralized or decentralized?
-----------------------------

A big design decision is if the system should be centralized or
decentralized. In the centralized version each user should at least be
able to restart hosting the project (as Smuggler mentioned in his post,
also the way it is done in Fossil, see below).

In a more decentralized system the content is pulled from multiple
users. I would prefer a solution like this, but it would increase the
complexity of a solution a lot (Scuttlebutt is an example of this, see
below).

Maybe a good compromise would be a centralized version with restart
ability that is upgradeable to a fully decentralized one.

If we are too ambitious with this we'll never get it done.

Interesting projects
--------------------

Some projects I find an interesting inspiration for `Async`:

### Fossil

Fossil is a simple, high-reliability, distributed software configuration
management system that recently added a very simple forum. Really
fascinating is there file format described here:

https://www.fossil-scm.org/home/doc/trun...ormat.wiki

### Scuttlebut

Scuttlebutt is a decentralized secure gossip platform. The most
interesting (for me) is the protocol/data structure they use which is
documented here:

https://ssbc.github.io/scuttlebutt-protocol-guide/

### The WELL

The WELL is one of the oldest virtual communities in continuous
operation. It shrinked _a lot_, but it has some good lessons to teach in
terms of "terminal user interfaces" (it can be used very efficiently).

Data structure
--------------

In conclusion, I think the most important thing we would have to design
is the underlying data structure of a future `Async`. Everything else
basically follows from there, a lot of other stuff is then basically
"just" synching the data and the UI layer.

I believe many successful projects are so powerful because they got the
data structure right from the start (Git is a great example of this)
which then was leveraged into very powerful tooling on top of it.

The data structure should allow a decentralized system, but could be
bootstrapped in a more centralized fashion.
Voltralog
#5
13-01-2020, 10:00 AM
Great thoughts.

I strongly agree with using Markdown for formatting, very simple to implement but also remains human readable even when not formatted.

The terminal approach does seem valuable. If you care about having less technically adept people in the community, then you would need to ensure it weren't the only platform. But if you didn't care then that could be a selection mechanism itself. How would images work in terminal? I imagine for collaborating on physical projects image sharing would be valuable.

>While re-reading it, I can't shed the feeling that it's a monastic like perspective. Long term, long post, community memory, preservation, focus, "purity", offline capability, slowness.

At first glance, the monastic approach of guarding knowledge over long time periods is now obselete because of the internet. But the sheer amount of information just creates a new problem, the fracturing of focus that Smuggler talks about. I'm interested to see if creating an environment that somewhat replicates the monastic approach will help individuals regain that focus.

We at least have an empirical test to perform. The community either finds the platform valuable, or it fizzles. I wonder what the founders consider success? How large and active do you hope this community becomes? I've seen plenty of forums die, but I've also seen many become victims of their own success, the community growing too large and unfocused. It seems like the focus on moderation and the ability to create sub-forums would help prevent that outcome. Though perhaps the fact that we are such a small minority of people to begin with is already a sufficiently limiting factor!
frankbraun
#6
13-01-2020, 12:25 PM
Thanks for your input @Voltralog. I forgot to mention files, but Smuggler mentioned it:

> Allow the sharing of files, without creating a visual enticement. No thumbnails and previews.

That's how I envision images. They would be possible as attachments, for the terminal version one might think of adding a shortcut to call an external image viewer. In my opinion the important part is that images are not shown inline, in order to keep the system text focused.

This part from Smuggler needs further thinking I believe:

>Do not support the inclusion of remote content (URLs automatically fetched, embedded videos etc). Instead, upload the content and turn it into the community's property and shared memory.

I like the idea of not using URLs (definitely no auto-fetching), but at the same time it is very useful and uploading the content has it's own challenges. Not sure how this could work.

> We at least have an empirical test to perform. The community either finds the platform valuable, or it fizzles. I wonder what the founders consider success? How large and active do you hope this
> community becomes? I've seen plenty of forums die, but I've also seen many become victims of their own success, the community growing too large and unfocused. It seems like the focus on
> moderation and the ability to create sub-forums would help prevent that outcome. Though perhaps the fact that we are such a small minority of people to begin with is already a sufficiently limiting factor!

I would consider this forum a success if we can get fruitful discussions out of it.
A mega-success if it leads to the development of better software like we are discussing here.
I don't think it is about the numbers of users, it is more about the actual discussion. However, I believe that moderation is kind of crucial in the process.
smuggler
#7
13-01-2020, 09:02 PM (This post was last modified: 13-01-2020, 09:07 PM by smuggler.)
@Voltralog: As a founder my success definition is pretty clear:
1.) Async(.pre) becomes the site for >50 humans where they daily log in and communicate (read AND write) and where the topics are both focussed on X-Anarchy as well as personal, that is: Async(.pre) managed to become an actual virtual community.
2.) I am one of those people, meaning: Async becomes the community for me :)
3.) Bonus: Async(.pre) develops a platform/software/tool that both allows better communities to be maintained as well as creates an experimental space for new social protocols.
[hr]
@frankbraun: No URLs in Async....
My thought there is that users save the content, upload it to the forum, and add the original link. Or they just copy&paste the text.
Content on the net often becomes lost, or is hidden behind paywalls etc. furthermore remote content poses a security and privacy risk.
That makes remote content a _negative_ in a community forum context. Lost/hidden content becomes a gap in the reference in communication and thus disrupts it.
Yes, it requires a little bit of extra effort of the user, but it's a necessity IMHO.

Better even is if people re-possess foreign content: Write an abstract with the relevant sources present, instead of just posting something that is 90% noise, 9% signal and 1% relevant to the context.
Voltralog
#8
13-01-2020, 10:23 PM
That's helpful to hear what success looks like. Seems like we're off to a good start.

I admit I'm surprised at the no URL proposal. It's such a fundamental part of the web. I can understand the reasoning behind it. For the long form, deeply thoughtful pieces that you're looking for, I see the appeal of forcing people to put in the work to show sources, both the improve the content itself but also save the information for posterity's sake.

Perhaps a dual approach will emerge where serious threads will not have URLs but casual threads will. This assumes casual threads exist, I guess that depends on moderation.
smuggler
#9
14-01-2020, 12:24 PM
@Voltralog: It also depends on how much work one wants into developing such a software. For static pages (that is, text content) it is conceivable to have a browser plugin that exports an URL to PDF, signs and timestamps it, and uploads that to Async. Technically that is very doable, but integration work must be done by *someone* :)
I think for video/audio that will work in theory but be a bit too much to implement in practice.
OTOH, it's a thing that will be seen in time. Maybe a combination of technical support software and user behavior/culture will find a middle ground that is realistic to achieve.
smuggler
#10
14-01-2020, 01:51 PM
@frankbraun:
# Platforms
I think we must look at this mostly from the perspective of how people engage with communication today. That means that for regular users, mobile UI is the most important one, especially when combined with a notification system. Most people today are not very stationary most of the day.
For Newcomers the web UI is likely the easiest way to onboard them. Both web UI and mobile UI can be mostly exactly the same thing.
For Terminal/Text UI it might be sufficient by starting exposing the datastructure and see where that leads :)

# Centralization vs Decentralization
As long as all content can be easily mirrored, I think that centralization is more attractive. That way communities can isolate themselves well. And it makes things soooo much easier to implement :)
Recovery and evolution by forking the data to a new host also seems to be an interesting "political" feature to me.
« Next Oldest | Next Newest »
Pages (3): 1 2 3 Next »



  • Subscribe to this thread
Forum Jump:

© Designed by D&D - Powered by MyBB

Linear Mode
Threaded Mode