

You can use the open drivers instead of the closed ones, but they’re not as developed.
Admin of lemmy.blahaj.zone
I can also be found on the microblog fediverse at @ada@blahaj.zone or on matrix at @ada:chat.blahaj.zone


You can use the open drivers instead of the closed ones, but they’re not as developed.


Nvidia dropped support for some older cards in their linux driver recently, so that could be it!
That’s more Kaity than me. Can I be three activists in a trench coat?
I feel like I’ve fallen through the cracks on this one!
Kaity fixed it. Sorry, forgot to let you know!
Close! Basically, once the remote instance has verified a signature, they cache it, so they don’t have to double check with the host instance every time. But when the keys change, and the remote instance is hosting a cached copy that doesn’t match anymore, you run in to problems, until the cached copies expire, and the remote instance refreshes their keys.


I had no idea Olympus/OM gear was hard to find in Australia, I wonder why?
I mean, that was decades ago. Digital wasn’t the only camera market, and Olympus was a small part of the larger camera market, so it was a niche within a niche. We could get lenses, but they took a long time to arrive, were over priced, and sometimes simply weren’t stocked.
These days, they’re harder to find simply because Olympus/OM is a small player in the digital camera market. There’s still plenty to go around, but sometimes, only in small numbers, because of the limited market for them. But online ordering is easier now, so it’s pretty much a non issue overall.
I don’t know if you can. You can do it through the default UI though at lem.lemmy.blahaj.zone
There will absolutely be federation issues for a bit due to having to reset the activitypub key pairs for every lbz user and community. Cotent is federating out, as you can see non lbz users interacting with some of the recent 196 posts. The issue is most likely that the keys are still being cached by remote instances, and until those keys fall out of their respective caches, things are going to be inconsistent.
We don’t verify emails. You can put whatever you want in that field.
That was kinda my feel as to what happened as well. It feels like if it was someone targeting us specifically with a multi layered attack, they’d have done something more overtly hostile. Deleted our data, defaced our site, or something. But the fact they don’t seem to have done much of anything after getting in there, and the fact that there wasn’t much in there of much use in the first place felt at odds with the sophistication of the attack. Which is why I am leaning towards it being driven by an LLM
We just haven’t had the time to get it fixed yet. We’re moving house and maintaining paying work. It’s Saturday today though, so it should be back up today


No, we don’t use Cloudflare. I would pull the site down before touching those evil fucks
Your passion project is community
That’s a really good way of putting it!
Same idea as relay.fedi.buzz
It’s useful for smaller servers. Basically, the only remote content that a fediverse instance sees is content produced by a remote account that a local account is subscribed too. So the more users subscribing to remote content, the more content that everyone sees. But in small servers, without a huge amount of incoming content, it can be hard to get a broader view of the fediverse.
This is a way for people on smaller fediverse instances to pull in broader content, because anything that tags.pub sees with the hashtag, gets pushed to your local server.
Thank you, that’s a good catch. We probably need to reset them
Interesting that he frames LLMs as both a tool that he needs to use to compete, whilst also framing their use as a scarlet letter that will tank sales if he uses them.