The Seven Deadly UX Sins of the Fediverse Web Experience (To Fix)

They're legitimate friction points.

In regards to instance selection paralysis, I think the solution is to invite people to instances specifically rather than an overarching community, software or protocol.

As for the Mastodon timeline, it should be a filter button that defaults to federated, one click to local, two to home.

like fediverse isn't an open source project that someone could actually contribute work to fix instead of write an article for internet points

everyone wants to be captain; no one wants to swab the poop deck

I think starting a discussion about these kinds of issues is important.
It's an area where all of OSS has issues. It comes with projects being developed by individuals their own requirements.
The challenge comes when "we" (the tech community) want to see such projects expand... Into the non-tech community. If that's to happen with any project, the requirements of that community need to be considered.
And they're a fickle bunch that are unwilling to learn how stuff works.

I really don't see good encouraging even more people on a single instance. I can only see that creating more unbalance.

And what is the solution to the timelines, removing them for all users? I've used the local filter to pick instances when I needed it.

One of the reason, why /kbin has achieved the success (so large, that eventually took it down...) has been it solving much of these UX sins:

This was only by chance, but as /kbin was _{not already} never ready for third-party hosting, there was only one non-Polish speaking instance of /kbin (flagship kbin.social). This is another reason, why registration numbers made briefly kbin.social's MAU greater, than that of entire Lemmy software!
But what about decentralisation? As I probably have argued when /kbin yet existed, the Threadiverse could as well consist only of lemmy.ml, piefed.social and a Mbin instance (or the Feediverse might by only mastodon.social, misskey.io, flipboard.com, and Pleroma/Akkoma instance) and still be decentralised. Several carbon-copy instances are the simplest way to make an interoperating network - but not the only one!
/kbin of course is a Threadiverse app and evades this sin by definition. A Lemmy user never has to browse an empty "timeline"
This is a cursed solution for this sin, but with so great scale of flagship instance (and no alternatives) much of the federation work has already been done by more determined people. And if not? Then...
... we can just say:

This is a very early beta version, and a lot of features are currently broken or in active development, such as federation.

and call it a day. :D

Five. This has been solved partially by Lemmy and its community Announces, and partially by scale.

Six. Of course everything on Lemmy and compatible platforms has to be posted to a community. However /kbin's magazines did more than that. They also aggregated (already federated) microblog content based on included hashtags. The magazine mod could specify, what hashtags would be aggregated under the mag.
The sidebar on /kbin's posts also includes sections like Related Magazines, Related Threads and Related Posts (for toots etc.).

Seven. ... and Active People. And "People" is one of the entries on instance's navbar. Yes, different one for every magazine, so you could follow anyone active in the topic.

All of these has been inherited by Mbin, its fork (well, maybe not its size xd)

Do you have any idea why Mbin never got as active as Kbin was?
- IMHO for the same reason why Myspace has over hundreds of millions of MAU and Spacehey has got only over million of users overall.
  IMHO solving UX sins does not bring new users. It rather helps not to deter them.
  The Great Migration from Reddit did not repeat with the same scale, as with Great Twitter Migrations. Even then, most people IMHO return rather to their already existing accounts on already existing instances. Fedia.io has over 5 thousands total users, less than dozen Lemmy servers.
  With the lemm.ee going down, we will witness a MAU drop. A non-zero number of people care no more about federation and will not make a switch.
  There might not be enough stock of users to repeat /kbin's growth right now. Even recent developments at PieFed make its MAU number only slightly larger than Mbin

Just found it on this post: https://mitra.social/post/01978f30-5295-cb5b-38cc-a151f08fd9f1

Haven't had the time to do an in-depth review yet, probably later.

Tbh it reads more like a concerned tech person thinks that everyone except them is confused little kitten, not like a list of a real problems that someone experienced. Some of the problems are real some are completely made up 'problems'.

In fairness, I agreed with his first point to an extent about the difficulty normal people have even understanding the concept of instances. I did once to try to explain it to my mom and was completely unsuccessful. The best metaphor I came up with was that it’s like choosing your local AAA office. You know they’re all AAA but this one might have a theme or perks that you prefer over others. It’s not Location based it’s more theme-based, but same idea. It’s not perfect, but she still didn’t really get it. But then when he refused to demonstrate his solution to the problem, and said we’ll have to wait for his next post to find out, my mind started wandering and I stopped reading.
- You use different phone providers can still call each other
- You haven't ever used email before? That's an obvious and easy explanation that most people understand.