This post is written on https://normals.me and delivered to !fediverse@lemmy.world over #ActivityPub, so if you are reading it the basic plumbing works. Now let me explain what is this.

Normals is a forum with communities, threads, and questions with an accepted answer. It is a single instance, not software you can host, and the core is closed for anti-abuse reasons. The federation module will be published separately with implementation notes (FEP-67ff).

What federates today Every community is a Group actor following the Lemmy model (FEP-1b12). Follow @selfhosted@normals.me or any other community from Lemmy, PieFed, Mbin or Mastodon.

Posts and comments go both ways. Replies from your instance appear in the thread on Normals, replies from Normals go back to the origin.

Remote content is labelled with its origin, links back to it, and is never copied. Media loads from the source.

Signed fetches, so it works with instances that require authorized fetch. NodeInfo 2.1 is up.

Policies, in case they matter to you, it has no ads, and no paid reach, and of course no AI training on posts. AI crawlers blocked by default, public mod log with reasons, and no automated removals. Numbers at https://normals.me/transparency so please take a look.

If something from your instance shows up wrong here, or something from here shows up wrong on yours, reply with the URL and I'll fix it. Andy Piper mentioned NodeBB uses the same group model, if anyone runs a NodeBB instance I'd like to test against it, if possible.

I hope we all have fun. Cheers!

@pynicolas @sposadelvento Jut installed it, but there's an important missing feature for my use-case, which is to have an "export to email" option that pre-populates the To: line (with my email) and the Subject: line (with the title of the scan). Any chance of that coming in? Or have I missed it?

@leighms

Dog racing has never been great for the dogs. Horse racing is a bit better but probably nothing of value will be lost if it continues to fade away.

I've known a few ex-racehorses kept as pets by horse loving people. The big issue is there are not nearly enough such homes.

Now some screenshots

i went to a movie where little children got melted alive by atomic death rays and they gave me a glow in the dark toy

@fesshole despicable being

I bought a ring light so that my partner and I could film some home porn. But we're only using it to photograph and sell our old clothes online.

god Im so old and bald

🎵 #NowPlaying on https://labr.online

🎵 Banging in Brighton 46
🎤 DJ Liam Wabbit

#LABR #radio #Fediradio #communityradio #fediwave #music #dj #mixes #house #hiphop #rnb #rock

Yay! Thank you, I appreciate the comment! It seems to work!

Stick a bunch of tetrahedra together to get a 3-dimensional manifold. Can you tell if it's a sphere?

Baldwin and Sivek have just shown that if it's *not* a sphere, there's a Turing machine that can check a proof of this fact in polynomial time. And their result relies on the newly proved generalized quasi-Riemann hypothesis!

https://arxiv.org/abs/2611.06705

People already knew this result would follow from GRH: the generalized Riemann hypothesis, which is the conjecture that the nontrivial zeros of so-called Dedekind zeta functions all lie on a specific line. But Baldwin and Sivek write:

"While writing our paper [BS19], we made the unfortunate decision to treat the number-theoretic input as a black box, taking “GRH” as something mysterious but mostly harmless. This is fairly common practice (many of us use “ZFC” in the same way) but not exactly a sign of understanding on our part—for example, it would have been good to know if partial progress toward GRH might actually make the above theorems unconditional, as explicitly suggested in the introduction to [Kup14]. The purpose of this note is to atone for our sins by seeing what exactly we might need from GRH to prove these theorems unconditionally."

It turns out that it's enough for all these zeros to lie in some strip (ε,1−ε). This is the 'generalized quasi-Riemann hypothesis', which was proved and formalized in Lean on October 12th using an AI system called AxiomProver:

https://hexagonmath.org/2610.00200v1

It's cool how new number theory gives new results on computability in topology Technically speaking, the new result is that "recognizing the 3-sphere is in coNP".

Thanks to @Ianagol for pointing out this new paper.

@futurebird
I used to go to the greyhound racing. Good fun at the time but now I'm not so certain. Probably some cruelty involved.

I very much doubt the AI use is confined to the frontend. But I don’t see why the frontend is fine to vibe code while other stuff is not?

Using AI as a tool is fine. It can be very useful for programming. But “vibe coding” is using it to an extreme degree where the AI is producing almost the whole thing, with a very small amount of guidance or design from the user. That’s what I would object to. Such a tool has very little value, anyone can vibe code such a thing.

Hello from RetroLemmy!

If you think choosing tailwind forces you into a certain style of website, you clearly don’t know what you’re talking about.

And you clearly haven’t just used AI for tests. I wonder if the source is closed because you know how obviously AI generated it would look if it was open source?

@alisynthesis
I'll refrain from making any Book Depository jokes.

Has AI been used to create this?

Do you really need to ask? 😅

not really, no.

Hi, I got a couple questions that aren’t answered on your transparency page yet.

• Has AI been used to create this?
• Two AI communities are on top of your communities list. Your terms talk about preventing AI scraping, yet your platform seems to welcome “AI founders”. How does this go together?
• Where is the source code?
• Your terms talk about monetization through Normals Plus. Can you elaborate on this?