ProfessorOwl_PhD@hexbear.net
on 01 Jul 11:46
nextcollapse
It’s pretty obvious that when a server goes offline the total number of posts will go down, but it’s not a helpful metric. Comparing the posts per day, rather than total number of posts on different days, would give a better idea of the effect.
The number of posts per day is a useful metric to measure activity, but the total number of posts, or specifically the reduction in the total number of posts, indicates lost posts. Luckily, thanks to federation, many of the posts live on on other servers, but it's still a huge loss of data,
Beauty of fediverse. An instance with thousands of users and posts goes down, fediverse still lives.
suburban_hillbilly@lemmy.ml
on 01 Jul 12:35
nextcollapse
/dataisugly
sanity_is_maddening@piefed.social
on 01 Jul 14:40
nextcollapse
I was registered on Lemm.ee. Now I'm registered on piefed.social. So, technically not Lemmy anymore. I do suspect a lot us moved to piefed. That was the general buzz. I genuinely don't know if the posts and comments of Piefed don't track on Lemmy. I'm sure the ones amongst piefed users in piefed instances don't though. That alone could justify a substantial part of that drop. But I'm still here interacting with Lemmy.
Lemmy uses ActivityPub, so people can use all sorts of software to interact with Lemmy. I happen to be using Hubzilla, for example. In statistics, I would be counted as a Hubzilla user, not a Lemmy user.
irotsoma@lemmy.blahaj.zone
on 01 Jul 16:47
collapse
Seems like poor metrics. Need an average across time. Also, if that’s the case even with that and with how easy it is for ordinary people to get an account on another server, does that mean we dumped a ton of bots?
threaded - newest
It’s pretty obvious that when a server goes offline the total number of posts will go down, but it’s not a helpful metric. Comparing the posts per day, rather than total number of posts on different days, would give a better idea of the effect.
The number of posts per day is a useful metric to measure activity, but the total number of posts, or specifically the reduction in the total number of posts, indicates lost posts. Luckily, thanks to federation, many of the posts live on on other servers, but it's still a huge loss of data,
Beauty of fediverse. An instance with thousands of users and posts goes down, fediverse still lives.
/dataisugly
I was registered on Lemm.ee. Now I'm registered on piefed.social. So, technically not Lemmy anymore. I do suspect a lot us moved to piefed. That was the general buzz. I genuinely don't know if the posts and comments of Piefed don't track on Lemmy. I'm sure the ones amongst piefed users in piefed instances don't though. That alone could justify a substantial part of that drop. But I'm still here interacting with Lemmy.
I think the fact that you’re here means it’s Lemmy. Just because lemmy isn’t in the url doesn’t mean it doesn’t use the protocol
Piefed is not the same as lemmy, they do interact together. But they use different apis and are tracked differently as mentioned in another thread.
Lemmy uses ActivityPub, so people can use all sorts of software to interact with Lemmy. I happen to be using Hubzilla, for example. In statistics, I would be counted as a Hubzilla user, not a Lemmy user.
This feels like bad correlation.
Seems like poor metrics. Need an average across time. Also, if that’s the case even with that and with how easy it is for ordinary people to get an account on another server, does that mean we dumped a ton of bots?