3 ms·

But I already see many "subreddits" (here called communities) on the main website: https://dev.lemmy.ml/communities https://dev.lemmy.ml/communities Also you h

by simias 6y ago

But I already see many "subreddits" (here called communities) on the main website: https://dev.lemmy.ml/communities https://dev.lemmy.ml/communities Also you have a big "create community" button at the top. Surely that doesn't spin a completely new instance of the application every time? And if not, how can we tell which instance a community belongs to? Honestly I don't really understand the need for something like that to be federated. In the olden days you had a bunch of forums/BBS/IRC network/Whatever that served various niches but didn't communicate with each other. For instance, what would we gain if we decided to turn HN into a Lemmy federated service? If anything it seems like in the long run it would be a disservice, as very large communities with lower standards would end up spilling and wrecking niches where the community is more tightly knit and post higher quality content. This fediverse thing makes some sense for IM and similar applications where you want to be able to connect easily with anybody. For forums however, it feels rather pointless to me.

rglullis 6y ago
Well, currently Lemmy does not even have a working federated implementation. Maybe it's too early to guess? > For instance, what would we gain if we decided to turn HN into a Lemmy federated service? It doesn't have to be a "Lemmy" Service. But let's pretend that dang decides to implement ActivityPub for HN. What would we get? Some guesses: - Less people trying to game/break HN. There is great value in gaming HN now because of its centralization. If HN is just one in a place with a bigger number of actors, I would guess the incentive to game it would be reduced. - Easier to have cross-pollination of ideas. There is an overlap between some subreddits, HN, lobste.rs, etc. Now, we can accept these services are big enough that there is always some cross-posting. What about the other topics that are HN-worthy (gratifies one's intellectual curiosity) but are under-represented elsewhere? - More room for dissenting/non-status quo views. With ActivityPub, your client could easily allow you to subscribe to an account. So let's say that I want to see whatever more controversial people post - e.g, idlewords. With HN, I need to either stalk him or hope that the echo chamber has decided on his favor on a given day. Granted, I think is highly unlikely that dang or YC would have any interest in doing something like that. They would be giving away control of the conversation and the risks are unknown for very little benefit. But is it really our job to be concerned about this? I'd rather have more people and more actors sharing this control than having to trust entities that become too big to fail.

by simias 6y ago

rglullis 6y ago

Well, currently Lemmy does not even have a working federated implementation. Maybe it's too early to guess? > For instance, what would we gain if we decided to turn HN into a Lemmy federated service? It doesn't have to be a "Lemmy" Service. But let's pretend that dang decides to implement ActivityPub for HN. What would we get? Some guesses: - Less people trying to game/break HN. There is great value in gaming HN now because of its centralization. If HN is just one in a place with a bigger number of actors, I would guess the incentive to game it would be reduced. - Easier to have cross-pollination of ideas. There is an overlap between some subreddits, HN, lobste.rs, etc. Now, we can accept these services are big enough that there is always some cross-posting. What about the other topics that are HN-worthy (gratifies one's intellectual curiosity) but are under-represented elsewhere? - More room for dissenting/non-status quo views. With ActivityPub, your client could easily allow you to subscribe to an account. So let's say that I want to see whatever more controversial people post - e.g, idlewords. With HN, I need to either stalk him or hope that the echo chamber has decided on his favor on a given day. Granted, I think is highly unlikely that dang or YC would have any interest in doing something like that. They would be giving away control of the conversation and the risks are unknown for very little benefit. But is it really our job to be concerned about this? I'd rather have more people and more actors sharing this control than having to trust entities that become too big to fail.