This instance is well into a second year of existence now, and while changes have happened its stable.
Thinking about longevity, i think Aussie Zone could benefit from a direction setting and issue defining annual meeting of sorts.
Issues like the large amount of dead communities and why they’re left, decisions that need to be made about inactive moderators, financing (if this hangs around for a long time this might become an issue for Lodion), broader direction setting, also might be a good way to look back on the years events and assess whats happened perhaps after some time has passed.
Its structure being analogous to a companys AGM, hence the name, the thread could be locked after two days or a week, etc, and amount of questions allowed per user limited.
In saying all this, i’m not forgetting that this is Lodion and Nath’s beast. But i think it could be useful to find out the broader instances thoughts or concerns, for instance if a user doesn’t feel an issue warrants its own thread, an AGT could be useful for that.
And, of course, the structure would have to be thoughtfully done. It can’t set up an us against them scenario in any configuration that could possibly detrimentally occur. Can’t also be an agonybaints session, the rules would have to foster only productive and earnest conversation.
I don’t know, I by no means have a fully fleshed out idea here. But what do people think of an AGT or something along that vein?
Has this actually been an issue?
I like to think we’re pretty quick on the moderation stuff, our main problem is time zone related. People making reports or signing up early in the morning don’t usually get seen until after 7am WA time (9am for half the country).
In my opinion, sometimes, yeah. I think you both do exceptionally well when it comes to report handling. I think Lodion has been handling the routine upgrades quite well, too. But when it comes to things like fixing problems, I think there’s been a lack of time investment put into fixing them.
The 2 things that come to mind were when the image proxying stuff was implemented and broke every single Imgur-hosted image on the site, it was a few weeks (might’ve even been over a month now that I think about it) before it was finally disabled. The other is from when the federation delays with LW were still present. Somebody developed a tool to implement activity batching (iirc) before it was implemented in Lemmy. But Lodion responded that they didn’t have the time to implement it.
There was also the time somebody proposed a migration tool to help people migrate from Reddit. To be fair, I do think the time required to vet it for potential security issues, and manage it, would’ve been worth more than the benefits anybody would receive from it. Perhaps that’s true for the batching tool, too, but I personally would place being able to see and interact with like 90% of Lemmy content fairly high up on the list of priorities. Double tbf: I don’t think Lodion actually said they didn’t want this because of the time investment required. I can’t remember exactly what the response was, but I’m pretty sure it was something along the lines of not being interested in managing any additional software.
I don’t mean this to sound like I’m bashing either of you, or AZ in general, I just wanted to share my perspective, since you asked.
I’ll pay the imgur thing. It happened during school holidays and we went away for the (WA) long weekend in there as well.
I did have a shot at looking into it, but didn’t try very hard. I didn’t really prioritise it, because it only seemed to affect thumbnails and even then, it wasn’t consistent. I didn’t see it as that big of an issue.
The lemmy.world lag drama was more nuanced. It’s not a time issue, rather the precedent that one would require. We would have needed to spin up a new server somewhere close to lemmy.world in Finland, batch their feed and send it home. All for an issue with one remote instance that the Lemmy devs said was about to be fixed.
We didn’t want to eat into the finances for a temporary problem. Had we known in May that it would take until November to resolve, we may have made a different call. But the point is that one wasn’t a time problem.
Thanks for sharing your perspective. I do politely disagree with the statement it wasn’t that big of an issue, as from what I could see, it completely prevented every Imgur image post from loading, thumbnail or not. Perhaps you mean the pictrs troubles (which are still intermittently happening, but primarily affect thumbnails and although annoying aren’t dire)?
Perhaps I’m mistaken, but Lodion specifically mentioned it was a time not financial problem when asked 3 months ago. I’m pretty sure there was a similar response when it was asked previously, as well
And I believe it wouldn’t have required that much VPS power. A $5-10/mo low end VPS likely would’ve sufficed. If that was really a concern, I would’ve pitched in to get it operational without killing the kitty
When the LW lag issue was previously discussed, my main constraint was (and still is) time. Longer term it would have been a financial concern.
Either way, the issue has cleared up now due to a number of factors, none of which required time or finances on our part. I call that a win.
Worth noting, LW have still not upgraded their instance to the version that introduced multiple sending threads to a remote instance. If their volume of traffic were to increase substantially again, we could “lag” again.
I’ve been following this from far away, what were the factors that solved the issue if LW didn’t update?
Looking at the graph, it just seems like the issue just got solved over time?
https://grafana.lem.rocks/d/bdid38k9p0t1cf/federation-health-single-instance-overview?orgId=1&var-instance=lemmy.world&var-remote_instance=aussie.zone
The volume of activity from LW reduced, and it also looks like they made some config changes on their side… at a guess more federation worker processes, allowing an almost constant stream of activity to be sent to AZ.
I see, makes sense