435
Lemmy.world down, probably following the upgrade. A reminder to move to smaller instances for a better experience
(files.catbox.moe)
this post was submitted on 29 Jul 2023
435 points (92.7% liked)
Fediverse
17677 readers
70 users here now
A community dedicated to fediverse news and discussion.
Fediverse is a portmanteau of "federation" and "universe".
Getting started on Fediverse;
- What is the fediverse?
- Fediverse Platforms
- How to run your own community
founded 4 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
A reminder that this constant advice people blindly parrot to install and flock to smaller instance has now created something like 1000 new servers in 50 days that are poorly run and already going offline as quickly as they went online.
Github Issue 2910 is the kind of PostgreSQL problems that the developers ignored for months and people still defend the developer choices to have the code doing real-time counting of every single comment and post for numbers nobody needs to needs done in real-time.
PostgreSQL is voodoo to this project, they do everything they can to avoid going to !postgresql@lemmy.ml community and asking for help, learning 101 about how to fix their SQL TRIGGER logic like Github Issue 2910 spelled out June 4.
I know you are salty about how you are getting treated over at GitHub, but you should look at it objectively, Blaze is clearly advocating that people join top instances that's not lemmy.world or lemmy.ml, not nobody instances that only have 1-2 users. They certainly aren't going offline as quickly as they come online.
I am always advocating for any of the top 25 instances that are not Lemmy.world or Lemmy.ml
For the rest of your post, I don't know what that has to do with people aggreating on LW.
aggregation refers to the lemmy database tables, site_ aggregates, community, person. The SQL TRIGGER logic lemmy_server uses that has been the source of so many crashes the past 60+ days.
Even if the SQL was top notch, it would not be a good thing for 50% of the active users to be on a simple instance. Just makes it easier to take down by any potential attacker.
Shit dude, you still going on about the GitHub issues?
You are really fucking stupid and it shows.
And this will always.. always be the biggest problem in the FOSS community.
"I dont like X, so I'm going to leave and make my own version of X"
So userbases get spread thin, manpower gets spread thin, developers get spread thin, and the user experiences degrades for everyone until it pushes them back to the bullshit websites and products.
This is exactly what federation is meant to solve: power in numbers without the centralization. Is that so hard to understand?
Sometimes I question why people not in favor of the decentralization are commenting on a Fediverse platform. Why not go to Tildes, Squabbles or another centralized alternative? There is plenty of fish in the sea.
This is another big problem in the FOSS
"If you dare offer valid criticism, then why are you even here? get out and go somewhere else!"
Your criticism is nonsensical. It's literally criticizing the purpose of the project.
Your answer didn't justify lemmy.world being treated the same as Lemmy as a whole. It's just a bunch of people who don't understand federation.
The response to you being basically "Not small-small, just smaller," is a great example of why people don't want to leave established sites for this kinda fediverse nonsense.
"It doesn't matter what instance you use!" Turns into, "Well, you chose the wrong instance," when the one you went with went down for any reason. Maybe you chose one that was too big (or got too big after you chose it). Maybe you chose one that was too small. Maybe you chose one where the admin was just not committed to running it and gave up after some time. There's a whole list of reasons why instances will have problems or fail or whatever.
Most people don't want to play this game.
I mean having a bunch of new servers is not a problem. Just choose one that's been up for a while and more stable.
But each additional row in site_aggregates table was causing the instability itself. The SQL code had major flaws. Adding more servers actually made Lemmy crash more.