Doc Edward Morbius ⭕​Nov 18, 2021

@humanetech User accounts are a huge barrier period.

The time when you could just post to a forum (or Usenet) via an anonymous form or self-asserted identity ... had its appeals. I'd love to see some kind of client-side PKI-based identity assertion rather than all the current crap of needing accounts or having some centralised identity service (which was what G+ was originally intended to be, BTW, since everyone's waxing rhapsodic over it here...).

Use robust tech, hide that from the users who need to have that hidden from them. But enough with centralised identity.

(And yes, it should support pseudonymity.)

@galaxis @tindall

Show threadRicardo J. MéndezNov 18, 2021

@dredmorbius @humanetech @galaxis @[email protected] Fascinating reading this - it's exactly one of the key design principles of something that a friend and I are working on: designing for personas.

https://distributedc.org/#/post/zdpuAxUCDJSmGHsXKphXGkVqdEgGshhAu67r7Zyd3domWMF8N/799545d3-4778-53b8-91c8-3f5a5a6f5cb6

Would love to read what you think if you give distributed[C] a try. Not adding more details for now because it would be great to see what kind of questions come up on a first encounter.

(UI is still kind of raw - we are focusing on the core for now)

distributed[c]

Thoughts distributed to the web

Show threadDoc Edward Morbius ⭕​

@ricardojmendez First impression is: this is a tabula rasa.

To the extent that I can't for a good sense of what it is or is meant to be.

I've made some attempt to make sense of the social media / user-generated content space as part of the project of moving people off Google+, a/k/a the Plexodus. See:

https://social.antefriguserat.de/index.php/Social_Platforms

@humanetech @galaxis @tindall

#DistributedC #SocialNetwork #PlexodusWiki #SocialMedia #UserGeneratedContent

Social Platforms - PlexodusWiki

Nov 18, 2021 at 10:59pmWeb
Show threadDoc Edward Morbius ⭕​Nov 18, 2021

@ricardojmendez One article that isn't there, though it probably should be, would be titled "So you want to start a new social network".

It would probably draw heavily on my observations around the ill-fated "kinder, gentler Reddit", a/k/a Imzy:

https://old.reddit.com/r/dredmorbius/comments/500ysb/the_imzy_experience_well_that_escalated_quickly/ (one of a series of essays, but the key one).

TL;DR: new social networks basically face two problem phases: 1) growth, 2) the problems of having grown.

@humanetech @galaxis @tindall #UserGeneratedContent #SocialMedia #PlexodusWiki #SocialNetwork #DistributedC

The Imzy Experience: Well, that escalated quickly

As I mentioned at Ello a few days ago, [I've taken a look at Imzy](https://ello.co/dredmorbius/post/9_abtcjwsspjwtnth57yxw), billed as [a kinder,...

Show threadDoc Edward Morbius ⭕​Nov 18, 2021

@ricardojmendez For a mature network, it's Problem Set 2 that is most evident. Veterans of mature networks are aware of the problems encountered at scale, and design a new network to address those problems.

But to get to Problem Set 2, you've first got to address Problem Set 1: growing big enough to have those problems.

And that's really tough because a huge element is sheer dumb luck. I mean, one lesson of G+ was that even with Google's backing and what was quite probably a $10 billion+ budget, specifically engineering growth is really, really hard. G+ didn't fail entirely (it had 10--100m active users, at least at its peak), but it certainly had the perception of failing.

And as I've noted elsewhere in this very large thread, they actually addressed many of the issues of Problem Set 2.

@humanetech @galaxis @tindall #DistributedC #SocialNetwork #PlexodusWiki #SocialMedia #UserGeneratedContent

Show threadDoc Edward Morbius ⭕​Nov 18, 2021

@ricardojmendez A common theme of the most successful forums I've seen is that first and foremost they addressed some extant group or community's real needs. Usenet, BBSes, mailing lists, Slashdot, Reddit, Facebook, Twitter. Probably Instagram and TikTok.

And that extant group had at least some level of aspirational appeal that others would be drawn to.

Dates refer to formative periods, not the full reign of the mode

Usenet (1970s -- early 1990s) was mostly technically-inclined college students (largely graduate), faculty, and staff, as well as some tech/military sector companies and a few government departments.

BBSes and dial-up services (late 1980s -- mid 1990s) saw strong early growth amongst military families dealing with overseas stationing. I'd always sensed a somewhat stronger theme of this from that community, but it wasn't pointed out to me that these services explicitly marketed to military families until decades later. Cheap fast comms of a widly-distributed and large population.

The WELL (1980s) was a BBS catering specifically to personal computing enthusiasts around the existing Whole Earth community.

Mailing lists (1980s -- early 1990s) initially addressed programmer and technical-user needs to share development and peer-based technical support, as well as of distributed academic groups. (Email was largely limited to the same groups as Usenet and drew heavily from the same cohort.)

Slashdot (late 1990s) was free-software geeks looking for a news / discussion source which wasn't beholden to Microsoft and its desktop monopoly and anti-Linux FUD, as well as the burgeoning set of dot-com / Web1.0 companies and technologies.

Digg and later Reddit (early-to-mid 2000s) were something of refugee communities from early Slashdot, and both largely saw en-mass defections to themselves. Arguably that crown's now passed to HN, which similarly revolves around tech and startups much as Slashdot once had.

Facebook (2004--8) was Literally Harvard, the Literally Ivies, then Literally Selective Universities, then Literally College Students. It based growth out of a young (long-term habit-forming) and appealing (both to other users and advertisers --- educated & high income potential) group.

Twitter (2009--2014ish) seems to have largely appealed to media in both researching and publicising news stories, as well as interacting amongst themselves.

Instagram and TikTok (late 2010s / early 2020s) seem to appeal strongly to the celebrity community --- pop music and cinema largely. (It's still early, I'm out of touch, this may well be wrong.)
@humanetech @galaxis @tindall #UserGeneratedContent #SocialMedia #PlexodusWiki #SocialNetwork #DistributedC

Show threadDoc Edward Morbius ⭕​Nov 18, 2021

@ricardojmendez An alternative to cohort-based communities is mechanism-based or technical ideology. Imzy was one of these, I'd argue that Tildes (ex-reddit staff) is, and Diaspora, Mastodon, and much of the Fediverse are.

There are also anti-cohort communities, largely defined by opposition to a specific existing community. Typically, "we've been kicked out of X so we're starting / joining Y". In almost all cases these Go Very Poorly: Gab, Parlour, Truth, SomethingAwful, 4chan, 8chan, 8kun, ... Imzy was this to some extent.

The problems these groups face are on both Problem Set 1 & Problem Set 2:

  • The initial founding cohort is not high affinity to outsiders. Problem Set 1.
  • The mentality driving technical infrastructure is very often opposed to good / effective / competent Problem Set 2 issues.

Basically, the networks have terrible hygiene, sanitation, and trust & safety systems. Where they exist at all they're typically aimed at persecuting designated scapegoat populations. This is good for building a highly-tribalistic group, but poor at either scaling or promoting accurate rather than ideologically-conformant information and messaging.

@humanetech @galaxis @tindall #DistributedC #SocialNetwork #PlexodusWiki #SocialMedia #UserGeneratedContent

Show threadDoc Edward Morbius ⭕​Nov 18, 2021

@ricardojmendez Having said all that...

... it's really hard to see how a platform or network or protocol will be used without seeing extant use cases. And the first issue I've got with DistributedC is that I don't see any content.

G+ faced a similar issue, the "Ghost Town" impression: people would join, register an account ... and then face a blank home stream. Attempts to populate this through identifying friends, suggested users, and/or an exceedingly bland set of topical suggestions ... were at best of limited success.

Onboarding is a challenge. And my view is that onboarding individuals is a broken model, rather, onboarding groups should be a core goal.

An example of this is Discord which originated as an out-of-band discussion channel amongst gamers. Wikia adopted a similar strategy around various fandoms. Rather than onboard individuals, these bring entire extant communities together at once (see earlier). They've remained somewhat niche, but are successful within those niches.

@humanetech @galaxis @tindall #UserGeneratedContent #SocialMedia #PlexodusWiki #SocialNetwork #DistributedC

Show threadDoc Edward Morbius ⭕​Nov 19, 2021

@ricardojmendez The other issue I have is that I don't see how DistributedC is actually handling identity. Best I can tell, it permits multiple identities, but still centralises their management.

That's ... an issue.

Mind: centrallised tracking of identities can be critically important in trust-and-safety, and that there's an inherent challenge between providing a decentralised and privacy-conscious service and ensuring various forms of abuse, ranging from name-calling to global crime syndicates or worse, aren't readily facilitated. That's a tall order.

Not sure how well this corresponds with what you have in mind, but I hope that's at least marginally useful.

/end/

@humanetech @galaxis @tindall #DistributedC #SocialNetwork #PlexodusWiki #SocialMedia #UserGeneratedContent

Show threadDoc Edward Morbius ⭕​Nov 19, 2021

#tootstorm (for my own reference)

#UserGeneratedContent #SocialMedia #PlexodusWiki #SocialNetwork #DistributedC

Show threadDoc Edward Morbius ⭕​Nov 19, 2021

@ricardojmendez Addendum 1:

What problem is DistributedC trying to solve? What is/are its goals and methods?

See: https://old.reddit.com/r/dredmorbius/comments/2fsr0g/hierarchy_of_failures_in_problem_resolution/

(That's phrased as a failure chain, but you can invert logic to get a necessary success chain: https://news.ycombinator.com/item?id=28972349)

What community/ies does it address?

How will it sustain itself?

@humanetech @galaxis @tindall

Hierarchy of Failures in Problem Resolution

Places in the process of attempting to resolve or address a problem (or "predicament") at which the process may end in failure. * **Failures of...

reddit
Show threadDoc Edward Morbius ⭕​Nov 19, 2021

@ricardojmendez Addendum 2:

I'd suggest finding some extant group (or groups) with unserved communications needs, and serving those.

All the better if there's a clear growth path from that initial group to a wider population.

This doesn't necessarily have to operate in concert with developing the platform, though I strongly suspect it's best if the two are reasonably well coupled.

@humanetech @galaxis @tindall

Show threadDoc Edward Morbius ⭕​Nov 19, 2021

@ricardojmendez Addendum 3:

The difference between Problem Set 1 and Problem Set 2 is pretty stark. The stuff that helps you in 1 does little for you in 2 and vice versa.

The transition between the two can be quite sudden. As in overnight you go from "how do we survive/grow?" to "how do we deal with our own growth and consequences?"

And you've got to solve the problems in order. First 1, then 2. The other way 'round really doesn't work.

Failure to solve both (in correct sequence) also doesn't work. It's no good if you nail 1 and foul up 2, or vice versa.

Problem Set 1 does, as noted, have a large element of sheer dumb luck. Standard portfolio theory (diversify your risks / spread your bets) applies.

It's usually best to take on an incumbent in a domain in which it does not, will not, or can not compete, rather than to assault it head on.

@humanetech @galaxis @tindall