Show HN: Mu – The Micro Network | Hacker News Reader

Show HN: Mu – The Micro Network

github.com

65 pointsasim41 comments

I've been catching myself thinking about this idea for the last two years. Maybe it's my old obsession with PKI and "personal digital infrastructures" that were both promised to us in the early days of the commercial Internet, but never turned concrete for various reasons.

IMHO, the best we could have today in terms of digital infrastructure is a personal/family level custom Mastodon node with basic Internet services like email, posts, tasks, chat, IM etc. but implemented in a way that all data would be portable to other services (open standards) and its storage would be "bottomless", meaning that users wouldn't have to worry about storage limitation for photos/videos for instance, as they would be sharing resources with other nodes worldwide. There would have to be some monetary incentive(s), of course, but they would be secondary to the bigger cause of keeping a true cyber interconnected community outside big tech.

I agree with a lot of that. I think the hard part is, who runs the nodes. If you hand a piece of hardware that people run in their house that's one thing. But if you expect them to run it themselves in the cloud it never goes well unless you have an engineer in the family. Maybe automation can allow the ability to spin up these nodes but ultimately it might be easier to support multitenancy and let a group of people run it e.g like every other saas service. But I guess the difference is the values on which it's founded. Every commercial or VC funded product goes the same way. Whereas stuff like Let's Encrypt has gone in a different direction. I'm not saying I have all the answers but some of these things we always seem to struggle to overcome. One thing I will say, the people who run it matter. Their motivations and their morale flexibility affects direction e.g ChatGPT has led OpenAI in a very different direction than first intended. Why is that?

Hey! I think I actually posted this some days ago and it went nowhere. I keep looking for feedback but you know one of those karmic things, we don't always get it when we want it but maybe when we need it.

I guess I just look at the way tech is now and feel disappointed. Its very addictive, "social" yet lonely. I don't find the alternative networks like Reddit, Mastodon, etc much better. I have held off doing anything, cut back most of my social usage but still found problems and years ago, yes I tried to make this idea work and shut it down. But I've felt like I really needed a solution, and so I use this everyday as an alternative to other forms of media consumption. And the hope is longer term it continues to develop as a useful utility. I'll keep investing time in, slowly iterating, figuring out what works and what doesn't. I'm not looking to gamify it or seeking fame or notoriety. Tbh, if it stays small great, if I can figure out how to make "conversational networks" [1] work even better. Anyway, happy to answer questions.

[1] https://arxiv.org/pdf/2503.11714

Ambitious. Very hard to achieve. But if we use go, github and discord for building the "anti big tech" stack - then we have failed from the start.

Why go?

I've been writing software with Go for over a decade now so it's just down to ease of use and what I know. It's performant, straightforward, compiled. It's a no nonsense language and does what it says. I'm not the type to get into language wars. I have a tool, I use it, that's it. Thanks for the question.

Go is a programming language. It’s not exactly pushing google’s ad agenda.

Your criticism of using github and discord are somewhat valid, but asking people to re-invent the wheel while they re-invent the bicycle seems like arbitrarily making up rules so everyone fails. Is there some influence you expect to leak into their platform thru github or discord?

go does push silly things. besides being overhyped, they block anonimizing tech when accessing package repos for example.

you dont want anything to do with google or anything they make or host. they will only do it to extract shit from you. they dont give nything dont be naive.

> It’s not exactly pushing google’s ad agenda.

Maybe not their ad agenda, but certainly one of their agendas. Specifically, the agenda they have to get young, inexperienced developers prepared for specifically the software development practices they employ internally.

It was directly stated that "Go is not for clever developers" and that their target is recent graduates with limited experience. It punishes you for trying to think about what you're building and to design sophisticated software, relying more on brute force. It doesn't encourage you to reach higher.

I don’t know where you are on your career journey, but having worked with countless clients, business domains, and projects as a freelancer I value readability over everything else.

If you’re working on a greenfield project in a team of one then I suppose it’s great to get in an expressive mood and emit your code poetry from those fingertips.

It’s very different to inherit a quirky puzzle and reverse engineer a mental model from there.

> Go is not for clever developers

What “clever” code is required to write a BBS-over-IP? 99.99% of code isn’t clever and shouldn’t be clever.

> It punishes you for trying to think about what you're building and to design sophisticated software, relying more on brute force.

Can you give an example of this? I have not written much go, so i am unable to think of a case where golang encourages brute force over sophistication?

FOSS projects have their constraints on time and usually not talent, like for-profit teams often are, so they should want the code they write to go farther.

They've explicitly stated that they wanted to discourage building abstractions, since "abstractions are hard to learn".

Concretely, this is evident in how channels and goroutines both poorly compose together, in part a result of the unsophisticated type system. It's difficult to build very generic libraries that can be leveraged as force multipliers, like the tokio-tower ecosystem does. You can do it, but it comes at performance costs or involves relying on codegen.

Google's Bazel build systems are designed around reasoning about and checking-in generated code, but the standard go tooling doesn't do this well and git workflows also don't really grapple with it well. This aspect of the design is very clearly an example of internal Google processes leaking out.

They are standing on the shoulders of giants. There's nothing wrong with the ideals and the motivation, but it begs the question: Could Mu exist without go? could go exist without google? Could Mu exist without google?

And all of that culminates with: Could the level of technology and the internet reach the state it is today without big tech? And if not, was the price we paid to get here worth it?

> but it begs the question: Could Mu exist without go? could go exist without google? Could Mu exist without google?

No, it doesn't.

Obviously Mu could exist without Go, if Google stopped development on the language, its current state could be forked (Go 2). Lots of programming languages exist without Google's support, there are even programming languages older than Google.

Woah didn’t know there were programming languages other than golang. Perhaps they can switch to nodejs, Google wasn’t involved with that.

C is also a programming language. Unlike go it foes not depend on a big tech company for it to live. No need to reinvent the wheel. Just use wheel that does not come with a big techh bagagge.

And AI-powered chat.

Otherwise it looked very interesting, I like the idea of a flat-fee structure without subscriptions.

They say that it's free from algorithms, but I found several uses of hash maps just from a quick glance through the source.

Haha that's clever! I hadn't even thought about it like that. I think today most non technical people and the news basically calls anything like an algorithmic recommendation systems "algorithms". The "algorithm" is what drives this sort of addictive feed of content that we keep looking for, hence that. But yea good point.

More of a data structure than an algorithm, if we're being pedantic.

I think that this could be marketed a lot better. The website has no information at all and it's not possible to work out what the project is from it. The github readme does contain a lot more information but it's still not presented very well. The "Overview" section needs to clearly explain what the project is rather than just stating the motivation for the project. I was able to determine what the project is only by reading the whole readme and piecing it together.

The membership is unclear also: it says "try for free" which makes me suspect I will only be able to use it for limited time before needing to pay, but you only need to pay to support the project and early access. Seems like a lot of potential users will be lost because they get the impression it's a subscription service.

Thanks for this comment. Very helpful and I'm going to rectify it now.

No worries. Hope it didn't come off too negative. It's an interesting project and I would hate for it to not be as successful as it could be just because of the website and other minor things that can be easily fixed

No I appreciate the direct comments, it's like instant reaction to the issues which is great!

There is more honesty in failing for the sake of ideals than in winning without them. It is a story that shaped many before and will shape many after, and mu may simply be one more instance of that enduring truth.

Why am I unable to test any of it's features without being prompted to install a Chromium "app"?

Is this why nothing happens when I click on any of the services in Firefox? There is no feedback at all.

I don't know that there's much value to, for example, creating another web based YouTube front-end when Invidious exists

Fair point, I'm not an Invidious user but I also feel like Video is part of a suite of services that I personally use, so I wanted to incorporate that way. There's lots of single purpose solutions to many things, I don't think that's the goal here. The idea is to look at the multiple habits that occur across social and put them in one place in a simple consumable format that just gives you what you need rather than promoting idle scrolling/clicking, etc

Sorry trying to fix that. It's an installable PWA but you can still use it as a website like any other. I have it installed on my phone. I guess I catered for my on usecase. Hopefully its partially fixed.

I logged in and was surprised to see Quran on the main page as "daily reminder"

Yes, so I won't apologise for it but I spent a lot of time trying to build something where I was getting a reminder for verses of the Quran. I think the home screen will be customisable soon enough for anyone who wants to see things differently or organise differently but I guess it was important to me. Maybe not to others, who knows.

Edit: I should say there's a configuration json in home/cards.json but I haven't documented the various settings for cards, RSS feeds, YouTube channels yet.

Is this related to the defunct mu.app from a few years back?

Answer: yes it is, I think

Yes, I did try this once before and had to "check" my own motivations at the time. After making an attempt to raise funding again I realised I wasn't honest in my intentions, so I shut it down and tried to cut back my social usage and figure out what else to work on. Guess I still felt there was something here.

11£ a month is a lot. Any way to try it for free ?

Yea it's free. You can sign up and use it. Sorry I guess it's not clear. The membership is a way to support the cause, get input on features, etc. I might have to go the route of Kagi and make that more tied to usage, I'm not sure yet.

Sounds like App.net all over again.

What's old is new again!

If you're still using all the same services though how exactly are you winning over addictive behaviour? eeehhhhh???

> Big tech failed us.

Uses Go, a language written by and maintained by Google [1]. Uses co-pilot written by GitHub for development [4].

Mu is £11 a month and you cannot see any screenshot of what you are getting [2], the same price you could buy a cheap VPS for [3]. The two authors of the project are asim and co-pilot. The commits have meaningless messages [4].

I would run a million miles away from this.

[1] https://en.wikipedia.org/wiki/Go_(programming_language)

[2] https://github.com/asim/mu

[3] https://www.racknerd.com/BlackFriday/

[4] https://github.com/asim/mu/commits/main/

I guess I should be more specific about why "big tech failed us". They essentially control all the dominant social platforms. While they have great developer tools like Go, GitHub (owned by Microsoft), the consumer products have been a point of exploitation. I think it's OK to both admire yet be critical and to try hold systems and people to a higher standard when they have such an impact on everyones lives.

To address some of your other comments on commits/copilot. Commit messages are about as meaningful as email subject titles. At a certain point they really don't offer much value when you have powerful search tools. Essentially the source of truth is the current codebase. Maybe the commit message is going to provide insight into what was happening at that time, but when you're coding with AI toolings it feels almost irrelevant if not dated. If anything it should autocommit with a useful message if its that much of an issue. Second to that, Yes I use copilot, why, because hand coding is 10x slower. I lay the foundations by hand but then started to rely on copilot for a lot of changes beyond that. Again going back to the point, yes big tech failed us, but on social and consumer. The dev tooling and technology is fine, but the addictive and exploitive nature of the consumer tooling is not.

In all honesty, thank you for highlighting the depths of the warts on the project. It's always good for people to see the truth.

Note on £11/month. It's free to use. Membership is just for those who want to support it and help with roadmap, get access to features, etc but point taken about the screenshot.

Reply on news.ycombinator.com