4 ms·

I don’t think we can go back to some things like ircd or mud talkers because they are too “chatty” to users. People like simplified centralized services with on

I don’t think we can go back to some things like ircd or mud talkers because they are too “chatty” to users. People like simplified centralized services with on screen discovery in the form of popups. The small internet will have to stay small

jchw 2y ago
That'd be more than fine with me, except the small internet competes for attention with the rest of the internet and gets slaughtered by their attention-sucking applications with shiny animations, spammy push notifications, gamification and manipulative FOMO-inducing tricks. This means that the "small internet" for any given niche is very, very small, even compared to what it would've been a long time ago on a vastly smaller internet. User retention aside... Nobody can even find the small internet. It's out there and there are search engines, but even if Google magically wasn't utterly ruined by SEO SPAM, people just don't Google their special interests as much directly anymore. (I can tell from search analytics!) So aside from a struggle to keep users engaged in small communities, there's also not very many users entering smaller communities either, certainly not enough to counteract the bleed.
- jjulius 2y ago
  >This means that the "small internet" for any given niche is very, very small, even compared to what it would've been a long time ago on a vastly smaller internet. This has been my lived experience with a few places the past couple of years, and I love it. It's a completely different experience from the "pop web" that most people use and it's amazing. >Nobody can even find the small internet. It's out there and there are search engines, but even if Google magically wasn't utterly ruined by SEO SPAM, people just don't Google their special interests as much directly anymore. I know that my example can't speak for most/many other places, but the regional hiking forums I frequent (same places I alluded to above) come up a lot on search engines. Whether you're looking for "[region] hiking", or looking up "[name of] trail", or anything related to it, the pages pop up towards the top quite frequently. It's how I found them, and there does seem to be a steady number of new users joining.
  - jchw 2y ago
    Maybe it actually can be alright for a niche as relatively large as hiking, but I think it has done some real damage to smaller niches, which seem to struggle to maintain active forums.
    - jjulius 2y ago
      That's a fair point. WATMM, for instance, is finally calling it quits. https://forum.watmm.com/ https://forum.watmm.com/
dylan604 2y ago
Sounds like someone just hasn't come up with the right app to act as an abstraction layer over the protocol.
- bathtub365 2y ago
  The other big problem with IRC is that if you have a connection interruption you miss messages.
  - jjulius 2y ago
    We don't always need to know everything that happened all the time, whether it's online or meatspace happenings. If my IRC connection dropped back in the day, and there was something that happened in that timeframe that was truly worth hearing about, I'd find out eventually. There's something to be said, at least in my opinion, about keeping a healthy dose of ephemerality in our lives.
    - immibis 2y ago
      Phones interrupt the connection every time you close the app, and if there's even a way to avoid this (yes on Android, no on iPhone) the user sees a notification that something is running in the background (fine) and their battery life is 80% less (not fine). The way IRC works is just inherently incompatible with the way mobile devices work, since IRC assumes stable endpoints. And because it's a protocol not a product, this can't be fixed. Even if a new protocol was created which fixed this, the necessary design change would bring so much baggage that it would become Matrix. To solve the unstable endpoint problem, servers need to store messages until all endpoints retrieve them (which is never, for channels of non-trivial size, since at least one client isn't coming back) or time out (how long do you set that? a week? If you're holding all messages permanently, you might as well never time out clients). The obvious storage design will hold each channel's messages once, not once per client connection buffer. Which means a lot of things: you might as well send it to new clients when they join; each message will have an ID so you might as well support replies and emoji reactions; you have to moderate it for illegal content; since messages have IDs, you might as well retract moderated messages on clients. At the end of the design process, what you have is nothing like IRC any more.

jchw 2y ago

That'd be more than fine with me, except the small internet competes for attention with the rest of the internet and gets slaughtered by their attention-sucking applications with shiny animations, spammy push notifications, gamification and manipulative FOMO-inducing tricks. This means that the "small internet" for any given niche is very, very small, even compared to what it would've been a long time ago on a vastly smaller internet. User retention aside... Nobody can even find the small internet. It's out there and there are search engines, but even if Google magically wasn't utterly ruined by SEO SPAM, people just don't Google their special interests as much directly anymore. (I can tell from search analytics!) So aside from a struggle to keep users engaged in small communities, there's also not very many users entering smaller communities either, certainly not enough to counteract the bleed.

jjulius 2y ago
>This means that the "small internet" for any given niche is very, very small, even compared to what it would've been a long time ago on a vastly smaller internet. This has been my lived experience with a few places the past couple of years, and I love it. It's a completely different experience from the "pop web" that most people use and it's amazing. >Nobody can even find the small internet. It's out there and there are search engines, but even if Google magically wasn't utterly ruined by SEO SPAM, people just don't Google their special interests as much directly anymore. I know that my example can't speak for most/many other places, but the regional hiking forums I frequent (same places I alluded to above) come up a lot on search engines. Whether you're looking for "[region] hiking", or looking up "[name of] trail", or anything related to it, the pages pop up towards the top quite frequently. It's how I found them, and there does seem to be a steady number of new users joining.
- jchw 2y ago
  Maybe it actually can be alright for a niche as relatively large as hiking, but I think it has done some real damage to smaller niches, which seem to struggle to maintain active forums.
  - jjulius 2y ago
    That's a fair point. WATMM, for instance, is finally calling it quits. https://forum.watmm.com/ https://forum.watmm.com/

dylan604 2y ago

Sounds like someone just hasn't come up with the right app to act as an abstraction layer over the protocol.

bathtub365 2y ago
The other big problem with IRC is that if you have a connection interruption you miss messages.
- jjulius 2y ago
  We don't always need to know everything that happened all the time, whether it's online or meatspace happenings. If my IRC connection dropped back in the day, and there was something that happened in that timeframe that was truly worth hearing about, I'd find out eventually. There's something to be said, at least in my opinion, about keeping a healthy dose of ephemerality in our lives.
  - immibis 2y ago
    Phones interrupt the connection every time you close the app, and if there's even a way to avoid this (yes on Android, no on iPhone) the user sees a notification that something is running in the background (fine) and their battery life is 80% less (not fine). The way IRC works is just inherently incompatible with the way mobile devices work, since IRC assumes stable endpoints. And because it's a protocol not a product, this can't be fixed. Even if a new protocol was created which fixed this, the necessary design change would bring so much baggage that it would become Matrix. To solve the unstable endpoint problem, servers need to store messages until all endpoints retrieve them (which is never, for channels of non-trivial size, since at least one client isn't coming back) or time out (how long do you set that? a week? If you're holding all messages permanently, you might as well never time out clients). The obvious storage design will hold each channel's messages once, not once per client connection buffer. Which means a lot of things: you might as well send it to new clients when they join; each message will have an ID so you might as well support replies and emoji reactions; you have to moderate it for illegal content; since messages have IDs, you might as well retract moderated messages on clients. At the end of the design process, what you have is nothing like IRC any more.