Mbin is a decentralized content aggregator, voting, discussion and microblogging platform running on the fediverse network. It can communicate with many other ActivityPub services, including Kbin, Mastodon, Lemmy, Pleroma, Peertube. It is an open source alternative to other link aggregator services like Reddit. The initiative aims to promote a free and open internet.
Mbin is focused on what the community wants, pull requests can be merged by any repo owner (with merge rights in GitHub). Discussions take place on Matrix then consensus has to be reached by the community. If approved by the community, only one approval on the PR is required by one of the Mbin maintainers. It’s built entirely on trust.
It seems it’s claim to fame is being more open and accepting of community changes and improvements. It can install as either bare metal/VM or as a Docker container.
Although anyone can install it and self-host it, their project page also contains a link to various instances that already exist and which anyone can register on.
See https://github.com/MbinOrg/mbin
#technology #opensource #Fediverse #linkaggregator #decentralised
Honest question: what’s the reason to make a fork instead of contributing do kbin itself?
Saw somewhere it was said the kbin side was going too slowly and not accepting some commits that their community gave. Some wanted to move quicker with newer features and enhancements.
My understanding of the situation is that Ernest, the main developer behind Kbin, thinks of the current Kbin as a proof of concept, and he is doing profound rewriting of the codebase to better fit his vision of how it should be working.
Meanwhile, other people wanted to contribute to Kevin directly, developing a better product on top of what Ernest considers to be too shaky foundations. So he’s not all that interested in pursuing that part of the development before he is happy with the core.
This also leads to a dynamic where he still has his own vision for the project and it goes through him, whereas other contributors want to make it their own more and develop something different.
It’s hard to see how to make everyone happy here without forking. Hopefully both projects can still gain from each other in the future: Mbin can benefit from the rewritten codebase of Kbin, and Kbin can implement features from Mbin after seeing that they are good and work well. In either case, the continued development as separate projects is probably not all that bad.
That makes sense!
For my basic knowledge working side by side with a lot of devs, I totally agree with the way Ernest thinks. It’s essential for a product that’s growing up to have a solid core that will not need to be rewrote in the near future.
But also, as an user, I keep wishing we could see more features, like the API for mobile apps.
I mean, besides what Ernest and Melroy thinks that’s the right way, there’s also what the users need, what the users want and what the project needs to escalate (new features vs core rewriting). And probably there’s not a right answer for that.
Yeah, I think it became a bit of an impossible solution the second Ernest’s proof of concept suddenly attracted a whole bunch of users and attention after the Reddit exodus. Kbin was clearly not ready, and I admire him for staying on course with the development after that despite pressures.
That said, users are not wrong to want easy to implement features asap. So I personally think the fork makes a lot of sense, though everyone could do without the occasional bad faith from some of the people involved.
deleted by creator
Melroy, lead dev, is the kind of person that if you don’t take his “advice”, he goes “fine, my way or the highway!”. He used to be a part of Lemmy dev supposedly. Then he was part of kbin dev. Then he tried to just make his own instance. Then when that didn’t work he started Mbin(which, if you haven’t realized, is named after him - Melroy Bin). Dude is pathologically egotistical.
Hi there, mbin dev member here. I do not know melroy personally, but I have never gotten the vibe that he is egotistical or wants to make the project his own. Never heard that he contributed to Lemmy…
Btw. we do not have a lead dev. He is the repo owner though
The core problem I had with kbin was that Ernest is just the kind of person who likes to work alone and in his own ways. That is just not a good fit for a project that gets contributions from the community (which I think he is not interested in). For example: I implemented a subscription panel in June/July 2023 and it got no reply from ernest for months. Then he replied once with the things he wanted changed and I did, then no reply anymore. I think it is still not implemented, but I lost intered. After I opened a PR about adding the same code to mbin I got some replies, answered them, changed the stuff that was complained about and voilá it got approved… It is just not encouraging to contribute to a project where your changes get accepted after a year if you’re lucky. Mbin is just more open to people contributing.
But yeah the start was rocky as can be seen here: https://kbin.social/m/fediverse/t/610675
some notes:
- exposes reputation/karma points in profiles ( i know some lemmy users are looking for this)
- does not federate downvotes (maybe someday!)
- i would recommend >=8GB of ram (4 is just too close to minimum)
- processing sensitive to external instances… lemmy.ml knocked me over for a minute when it came back up due to my low processing resources.
I don’t think karma should ever be visible for any reason and downvotes should be disabled. But I know that is not a very popular opinion. “Karma” should only be in the background, and for the purposes of moderation/admin needs (such as thresholds). Downvotes invariably become disagree buttons, and net karma doesn’t really mean anything other than someone has been on a long time.
yes! perfect example of how there is no one-size-fits-all.
many instances will and many will not. you do you.
Yeah I think that’s definitely something beehaw gets right. It noticeably alters the tone of conversations when people can’t just downvote everything they don’t like. Kbin’s making votes public I think helps somewhat though.
The big question…
Do any of the apps work with it? I’m using Voyager right now, and this is a dealbreaker for me if it doesn’t. (I’m assuming the answer is no, but I thought I would ask.)
Both kbin and mbin function very well as progressive web apps, and shouldn’t /require/ a discreet app like, imo, lemmy does.
There Matrix discussion forum may be best place to ask - https://matrix.to/#/#mbin:melroy.org
deleted by creator
I just tried signing up at fedia.io, and I got the response “429 Too Many Requests”. It doesn’t inspire confidence that I can’t sign up for the largest and most popular (and represented in the screenshot above) instance.
That was the instance I signed up at, about 10 mins before I posted this link. Lemmy also went down in the last day, so nothing is bulletproof. But the site is working as I’m browsing and commenting right now.
Okay, I’ll give it a few and try again.
And you shouldn’t choose an instance just because it is “the largest” ;-) A big problem from which the fediverse suffers at the moment
Yeah, I definitely understand, which is why I moved off of lemmy.world. But the other mbin instances are really small in terms of members, so that makes me nervous that they aren’t going to be around long term. I get it, there’s no guarantee with any instance, but one that has a miniscule number of users makes me gun shy.
Well I can promise that my 2 instances will be around for a while :D That would be https://thebrainbin.org (English, stable) and https://gehirneimer.de (German, uses with some of the latest commits for testing)
What happened to Lbin?
Feel like the *bin and the *key have a lot of cool features, but way too many forks to have a understandable product. It’s pitty
It’s one of my main concerns with the fediverse, that tech people’s (and people generally) inclination to not work together and instead do their own thing will be detrimental on the whole.
Which is EXACTLY this. Melroy(Mbin dev), was a Lemmy dev that was “ignored”, then he was a kbin dev who was… “ignored”. But then he ran his own instance, which was again, ignored. So then he started Mbin, which of course he named after himself(Melroy Bin). In between each of these attempts, he’d go to numerous communities and post long rants trying to call out the other devs about how terrible everyone else was and how no one is listening to his brilliance and he’s the One True Savior of the fediverse.
Dude is pathologically egotistical and is exactly the kind of person I don’t want admining anything.
One great instance of Mbin I’ve been on is kbin.run. So far, very stable and no complaints
What are the key differences? How is it more open?
kbin had not been accepting some commits and apparently were moving quite slowly with newer features. So, this is more like a dev version type implementation. It is more “open” to changes and commits apparently. Not more “open” as in open-source.
I don’t get the pull request part. does that mean each instance does not control who they federate with???
No, the pull requests are to do with submissions of source code to the core project. The project owner has to review and accept those changes for them to happen (or not).
