Live data from Hacker News

The “smolnet”, build for friends and friends of friends

communitywiki.org

51–60 of 104 posts

Re: The “smolnet”, build for friends and friends of friends

#51
post #50
post #37

Earlier quoted context omitted.

Chat is a very bad flow for this, because chat is a stream, not a garden. It's made for quick feedback. What you want is closer to a wiki, but I understand you need to share it with other people and discuss about it. I don't know of any software that does that, but the good thing is that maybe you don't need a new platform. Creating new platforms just splits the communities and is against the Unix philosophy we hold…

Yeah, I've considered slapping something together in Django or what have you—libraries exist to make a lot of this not that hard—but... kids. And the new Zelda, for that matter :-)

The new Zelda has utterly destroyed my progress on side projects and also playing guitar.

That's ok though, this happens to me every now and then - a game will come out that is so good / compelling that I just need to drop everything else and jam all my free time into it. It's almost like something in my psychology is saying "play the shit out of this so that we can get over the obsession and get back to other pursuits quickly."

I don't game all that much these days, so when something really good comes along, I just let myself go nuts on it. I've realized I seem to work in cycles, so after my Zelda obsession dies off, I'm sure I'll get back to other stuff.

Re: The “smolnet”, build for friends and friends of friends

#52
post #5

I like this. Speaks to my own goal of creating "micro" communities. https://micro.mu . We limit groups to 20 people. I think that's an ideal number for real connections. Need more, it's likely for a different type of communication.

20 seems a bit low. what happens if i want to invite 21 people?

many groups have a few active people and many less active people. one group i am in has about 80 members, 10-20 or which are active and keep a steady stream of communication. if the group were to grow much beyond that, it would be to much.

so you are right that 20 active people is a good group size but if there is a technical limit of 20 that means i have to police all the inactive people in order to have room for active people. seems to create more overhead than i'd be comfortable with.

wechat groups have a limit of 500 people. it's big enough to allow some flexibility for groups that have lots of members but little communication, but don't allow the groups to get so large as to get out of hand. the ideal technical limit for me is somewhere between 100 and 500. less than that and i'd run into to many corner cases because some group somewhere hit the limit. most of the groups i am active in have less than 100 members, but many have more than 20, but i am also in some groups that have more than 200 members, most of which are just reading and a few contributing.

Re: The “smolnet”, build for friends and friends of friends

#53

What's with the sudden influx of "smol" things? Is this related to "smol" ML / "smol" AI models, and/or the HN user that (IIRC) is associated with them and likes to use the word "smol" in their comments?

It's a cat/dog/animal thing. At least that's where I see "smol" used the most, in relation to kittens and puppies and other newborn animals. And it's been brewing there for a few years now, and the word's escaping into more contexts.

> It's a cat/dog/animal thing

And a pipi thing too.

Re: The “smolnet”, build for friends and friends of friends

#54
post #52
post #5

I like this. Speaks to my own goal of creating "micro" communities. https://micro.mu . We limit groups to 20 people. I think that's an ideal number for real connections. Need more, it's likely for a different type of communication.

20 seems a bit low. what happens if i want to invite 21 people? many groups have a few active people and many less active people. one group i am in has about 80 members, 10-20 or which are active and keep a steady stream of communication. if the group were to grow much beyond that, it would be to much. so you are right that 20 active people is a good group size but if there is a technical limit of 20 that means i hav…

what would be interesting is a feature where the moderators can limit the group size. start with 20, and if that limit is almost reached, have an option to raise it. by that time a moderator will have an idea if the group can handle more growth or not.

Re: The “smolnet”, build for friends and friends of friends

#55
post #42

Earlier quoted context omitted.

Good eye. Yes we're thinking about that already. I think when we think about this tool from a community and even family or friends perspective, keeping track of everything you're doing or need to do becomes difficult even if you're living by the calendar. It's because that calendar isn't shared. So I think for us "events" is something we want to find a way to integrate at the core of the experience. That at to-dos, b…

Sweet! Yeah my partner and I have made a shift to keep our lives in our personal calendars which has been a huge QoL upgrade for us, but it's still so hard to keep track of who in the friend group is available when, as well as keep everyone remembering what's coming up in the next week or so. I feel like this was the biggest value+add of Facebook, but that particular product is now all but unusable :(. Adding shared,…

I think NextCloud pretty much fits the bill here. I think it's a fantastic free tool that's easy to host on your own for exactly this kind of thing.

Re: The “smolnet”, build for friends and friends of friends

#56
post #5

I like this. Speaks to my own goal of creating "micro" communities. https://micro.mu . We limit groups to 20 people. I think that's an ideal number for real connections. Need more, it's likely for a different type of communication.

I think what ruins communities is large sustained growth and not size. Back in ye olden forum days, you'd often see cozy communities with thousands of members, most of which weren't active and a lot of them came and went. You'd see like one maybe two new members every few days. What kills most mainstream social media is needing to show growth to investors. This turns them from a village with familiar faces into a tra…

It's assimilation speed and has been seen time and time again in various contexts.

If I take my American family and move to Australia, we will become Australian over time.

If I take my family and the entire city I live in to Australia, there will be an American city in Australia.

Eternal September is just when the influx overwhelmed the "carrying capacity" of Usenet, and anyone who has been on any forum for a length of time can refer to a similar influx (often complaining that everything has gone to crap since then).

Re: The “smolnet”, build for friends and friends of friends

#57
post #52
post #5

I like this. Speaks to my own goal of creating "micro" communities. https://micro.mu . We limit groups to 20 people. I think that's an ideal number for real connections. Need more, it's likely for a different type of communication.

20 seems a bit low. what happens if i want to invite 21 people? many groups have a few active people and many less active people. one group i am in has about 80 members, 10-20 or which are active and keep a steady stream of communication. if the group were to grow much beyond that, it would be to much. so you are right that 20 active people is a good group size but if there is a technical limit of 20 that means i hav…

I think that's spot on.

The other thing I noticed that in let's say "stable" communities (slow growth, many long term members) there is much more of self-policing going. If someone acts outside of the implicit or explicit rules there is bigger chance someone would respond to that in "hey, behave, we don't do that here" comment and 9/10 that's enough, no mod involvement is needed, no bans or other punishment needs to be applied.

When community is a lot of "random people off the street" (no effort to create account or join, just start writing) that happens much more rarely

Re: The “smolnet”, build for friends and friends of friends

#58
post #45

Earlier quoted context omitted.

What did you use to implement end-to-end encryption?

There's no end to end encryption right now but it's interesting the thought immediately goes there. This is not a replacement for signal, WhatsApp, etc. It's more like an advancement on forums or replacement for Facebook groups that takes things private in the sense that it's invite only and no one outside your group can talk to you. The recent community chats product from Facebook has also foregone end to end encryp…

Difficult to have "real connections" in a small, intimate group if you know you're being watched. Putting a 1984 telescreen in the room has a chilling effect. Without E2EE that's what you are building.

Small, encrypted groups with people you know makes sense (like Signal). Large, unencrypted, open groups make sense (like reddit, facebook groups). I don't understand how you're trying to make something with parts of each, it doesnt make sense to me

Re: The “smolnet”, build for friends and friends of friends

#59

What's with the sudden influx of "smol" things? Is this related to "smol" ML / "smol" AI models, and/or the HN user that (IIRC) is associated with them and likes to use the word "smol" in their comments?

Smol webs, smol creatures, smol frens (often also creatures). Can't have too many.

Re: The “smolnet”, build for friends and friends of friends

#60
post #7
post #4

Looks like they’re using “smol” servers; it crashed under a few hundred HN visits ;)

To be honest, that's kind of the point. When you're "webscale" you're inevitably at a level where you have to pay a lot of money just to keep the site up. Smolnet is supposed to be run from someone's basement.

One of my major beefs with IT conversations is the stubborn adherence to the manufacturing notion of 'economies of scale'.

That's not how algorithms work. Most of the things we do are nlogn complexity. There are lots of breakpoints where 10x as many users cost you 11x as much in hardware, and we pretend it's going to cost us 8x. Any time your estimates are off by 40% should trigger your root cause analysis protocol.

In particular we are lousy at statistics. We have to be right in the middle of the consequences of misreading them before we act. When you have 10 servers you might have to deal with a deployment race condition once a month. When you have 30 servers you might have to deal with one every week. That starts to materially affect your estimates when you have five of those sorts of things going on.

Post reply on HN