Fnoord

Must be -> could be. Which is the point. And it is an easy way to communicate based on merit. It is likely done from a terminal with 'normal internet access', next to one of the intranet ones (or even airgapped) where the issue is occurring. Even if you'd be able to hack AOL, you'd likely end up on a dead end, or a front company. There's an added bonus: NSA might even know who it was/is internally, so someone who isn't in the 'in group' would get noticed.

You can notice something similar on Ghidra or Cyberchef repos. People in government agencies and military follow this M.O. all the time; hiding in plain sight.

Nowadays, probably Gmail fullfils the very same role, with a caveat: clearly available world-wide. For Germany, this used to be GMX or T-Mobile (Deutsche Telekom).

What about retaining or defending your license/copyright?

Stealth Model, is this a CTF?

LLM needs to become more transparent, not less. Hence, this idea (and trend, possibly) is disgusting.

How can we even possibly verify 'Prompts and completions are retained by the provider and are not used for training...'? What if the training is done, but used internally?

5-7 years? With daily use device 100% drain your battery will be about 85-90% health after 1,5 years. Once it gets to 80%, Apple recommends replacement. I'd say replacement should occur at approc 2-3 years of daily usage.

IIRC costs and/or size. I agree, silly change. Going from the blue to green theme is another silly change, as is renaming scheme. I like many things this company does, but some decisions I question.

I also bet a smaller, less powerful smartphone with 3.5 mm would sell like hot cakes.

I believe your story, the thing is... once my Babboe is driving, I can reach 20 km/hour. The problem is, it takes some time, and I'm fine with the 15 km/hour to 20 km/hour taking time or effort. I'm cool with staying at 15 km/hour as well. The problem is only after I am completely at 0 km/hour at a traffic light or stop sign, getting velocity. I wouldn't need better brakes there. Since I would never go higher than 20 km/hour electric, I wouldn't need better brakes. If they say I shouldn't go faster than 15 km/hour, I'll stick to that, no problem.

But for people who'd reach the max 26 km/hour, if they'd need better brakes, then if they reach such speed manually (going downwards, for example) would suggest my current brakes aren't adequate either. But I'm not interested in going from to 20 km/hour ASAP. I am interested in getting a little of help at 0 km/hour going forward. After that, the electric part can go off for all I care.

Here's my cargo bike: (non-e variant): [1]. The company got bankrupt a couple of days ago.

It has three wheels. Two smaller ones at front, one back. As such, I'd prefer to install the motor on the back. We do have that speed limit of 26 km/hour (my current e-bike has it as well), but since more force is required, the kit would need to know this as well.

I got it second hand, same as my e-bike. Otherwise, these aren't affordable for me. 6k EUR for a cargo e-bike is just insane IMO.

[1] https://www.babboe.nl/media/akeneo_connector/media_files/b/a...

Looks like you can't have a passenger anymore either. I like to have my kid on the backseat of my e-bike.

I'm not in the market for this, but I am in the market for making my cargo bike electric. Having to slow down at a traffic light and then start from zero absolutely sucks. It goes very slow, it is rough, and you won't have time. I'd use it just for that specific use-case.

Your Soviet comparisons are a dishonest take. They're also moot, since an issue would be authoritarianism. Regardless, yes, authorities need to exercise their authority to ensure the integrity of elections. It is difficult to do this in a fair way, and regardless it will stir controversy (another vector to throw oil on).

The external influences are documented. We know the Russians meddled with elections in The Netherlands (the 2016 Ukraine referendum), Romania (recently), Brexit, Trump 2016 (via the Clinton hack), 2020, and 2024, and likely also Hungary. In regards to Romania, yes the elections were repeated. That thing tends to happen when elections are fraudulent.

> Fortunately in this case of Brazilian elections, there is nothing secret about the list of accounts which have been limited. The censors are very open about who has their account limited and why (all political candidates during elections).

Because it is about a local law regarding airtime. A law Musk doesn't respect. But because now it requires an active action of the mark to get into a propaganda tarpit, instead of the timeline, it is a reasonable deterrent.

> That's deep into 1984 or Soviet / Chinese censorship territory.

Not if you consider the context of external influence, e.g. [1], and not wanting a Jan 6 in Brazil. Not letting such know their efforts are ineffective, is a tool to combat their effort. It is akin to shadowbanning.

Amplification of radical right by external entities is nothing new, btw. Russia did this in 2016. Musk does it now. But Musk won't show those parts of his machine. He still does shadowbanning, too.

[1] https://medium.com/dfrlab/election-fraud-narratives-surge-on...

> But if I were Brazillian, I'd be immediately following all ~700 accounts (and probably unfollowing all 700 after the election).

I would certainly not, as it is most likely propagandists; these aren't interesting.

But either way, the purpose of this is to protect democracy, and by mentioning the very accounts, this could have a Streisand effect. That by itself is influencing the election, and Musk knows this. That is the purpose of his action. Musk does the same with regards to other democracies, such as AfD in Germany.

Yes, well, kind of. Opera had Multiple Document Interface (MDI) which was actually (back then) far more flexible than tabs. But yes, Opera was first with MDI, and also I suppose the first web browser for the smartphone.

I ran Wayland over Waypipe over WireGuard (Tailscale, in my example), and it worked very well. That over a 1 gbit fiber... shared with whole building/office. In fact, I ran remote AOSP this way.

Now, VNC might actually also work very well over WireGuard or SDN based on WireGuard. Since VNC was designed for LAN, in a time where LAN was the common use case.

For me, low latency is from the past. I always got either wired, Wi-Fi, LTE, or Starlink.

Goes very well with music by Total Eclipse [1]

[1] https://www.discogs.com/artist/516-Total-Eclipse

Those people are fine just using arrows and backspace, as it isn't worth it to min-max this. I don't want to learn/use Emacs, this doesn't even work on macOS, and Ctrl+A interferes with my tmux keybind. This is why vi keybinds are nice in shells, but... defaults matter.

I haven't decided yet if I want to use NFS to malware my configs, or git. I think git is more flexible, but I don't want my configs in a public git.

Either way, I use atuin a lot, and before that history | fzf. Which is bound to ctrl+r for me. But again... not by default.

Oh, nice.

I alias'ed it to please as well. But because I use fish, I had to do this: alias please 'sudo (history -1)'

I meant that from the scammers viewpoint, and their colleague/competitor).

If I was a scammer, I'd target people who were rich, and the elder. I would use leaked datasets to find these people. And then I wouldn't give up after I failed once, but guards are up if a scam failed, so seems not useful to retry.

But honestly, those scammers you mention seem low hanging fruit.

Yes, you could forward an Android call to Asterisk server and use STT (Voxstral has real-time module), LLM, and TTS to generate a reply. But I think the added latency is going to be problematic.

I wish you all the best, I hate marketing, but I hate liars even more.

Next one who calls you is likely a colleague of the one who just phoned you. Don't inform the colleague you cannot be converted to a sale. Let them waste their time, not reaching their targets.

Were you a former customer of Vodafone-Ziggo? If yes, did you delete your account?

AFAIK it is similar to how Enigma got cracked, due to certain variables always being the same (end of letter greeting, date, etc.)

See also ECB penguin [1]

[1] https://words.filippo.io/the-ecb-penguin/

> In the early years after 2000, I have witnessed a great number of computer crashes caused by power failures.

Don't enable write cache without a power backup, or on an unstable machine.

> Therefore OpenBSD should be used with an UPS (or on a laptop).

NVMe with PLP is suffice.

The same helps me on my NAS as well, with ZFS and Linux. No need for OpenBSD.

Soft updates was very cool when I first used it, like 25 years ago. Linux didn't have Ext3FS yet, or just about got it. ZFS also worked very early on FreeBSD, but nowadays the standard is OpenZFS.

There's a couple of cool things coming from OpenBSD camp, most notably OpenSSH and PF.

A normal, physical doorbell is 24/7 connected to electricity, else it wouldn't work. It typically has no surveillance capabilities, except when used, it makes a sound/signal/buzz. That's typically its purpose.

A 'smart' doorbell [meta: I hate this terminology, and will just call it 'doorbell camera' for the rest of this post] needs electricity in order to do the very same, but it also serves another purpose: record (via camera and microphone, and sometimes other sensors ie. IR camera). This recording of data can be stored, analyzed, etc. For example, you can perform ML on the feed to detect dogs, you can do data analysis on it for example when postal service usually arrives, you can record postal service stows packages away in the premises, etc. In order for it to be analyzed, it needs networking capabilities (wired, or wireless). Doorbells need to be serviceable, so they can typically be dismantled from the outside. Wired and wireless can both be MITM'ed, and DoS'ed. If that happens, you want the feed to be recorded as much as possible, in order to record the attacker. You also want to secure your feed. In order to MITM or DoS a wired doorbell, physical access is required which is why a wired doorbell is superior IMO.

There's no need for your camera to have internet access. Some ports have to be accessible from your NVR; that's it: you need to be able to update the firmware (e.g. to protect against vulnerabilities such as KRACK), and you need to be able to record your stream (e.g. access the RTSP stream). A speaker or screen can be useful to communicate with a visitor. Your NVR can manage all of this, and so can you (usually via a reverse proxy with internet access, but we shouldn't neglect authentication here). The doorbell camera itself doesn't need such. It does, however, need some form of networking capabilities.

So what I do: apart from having it in a separate VLAN, I record my feed, and immediately store it locally (on-prem) and remotely (via adequate internet access; in my case, not in cloud). However, my doorbell camera is connected via Wi-Fi. This is where the story e

This is why my doorbell is on a separate VLAN, without internet access.

I used to take my laptop or Nokia N810 (remote shell over Wi-Fi with physical keyboard) with me to the bathroom. Nowadays there's cyberdecks, e-readers, ...

Just imagine, there was a person on #debian reading and taking a dump.

Just to reiterate: book good, e-reader bad. Phone good, smartphone bad. Scanning through a physical magazine good, doomscrolling TikTok bad.

Except it is about focused vs. diffuse mode of brain, and usually people underappreciate the latter. E.g. walking during a lunch break, thereby getting inspiration for work, is working. The same can be true for taking a shower, sporting, staring outside at the landscape or stars, or meditating. It will also empower your focused mode later on. And, the author's point seems to be a toilet break is a decent opportunity for this. I agree. On my previous job, I regularly sat for 10-15 min in late morning, and it inspired me. But you cannot force people to not take a device with them to toilet. I mean, unless you're working for NSA or something. Don't judge such, look at total performance / efficiency per hour instead. Or per USD spend on the employee.

As a former smoker, this is also what I appreciated about having a fag (in a time before smartphone and vape were common). Some missed the socializing. I hated that part. It seemed to be more like FOMO on gossip, little substance.

The people who deny climate change have won, and accepted the [now] inevitable downfall of humanity as we know it. They are enriching themselves to have the financial power to give them the personal power to adapt to climate change. The rest can rot in hell. In essence, they're able to build their own Arc of Noah, and you and 99,999999% of everyone else (nevermind all the animals) aren't on the guest list.

In order to not cause chaos, they lie (by denying climate change, sowing FUD) or ensure the issue isn't mentioned/covered (by owning media, self-censorship). "Keep calm, everything is going extremely well."

I believe Earth isn't badly off with a lot less humans, I just don't like the distribution of survivorship isn't going to be random. Survivorship bias is going to contain heroic stories, but in the end a lot of people without power/money are going to take the shortest straw. And, that wasn't necessary.

The track Extrawelt - Phoebe has an excellent quote on the matter. Paraphrased:

> "We would be the first lifeform that wipe itself out. But what is unique is we did it knowingly. Why didn't we save ourselves from [...]? Is the answer, we weren’t sure if we were worth saving?"

if you pull stuff like that off, and it gets flagged, it is obvious you are trying to game the system. Same with a system like this. It could be used in addition to human moderators, where the human moderator have to meta-moderate the AI's work. This could also be used as training exercise for new moderators. Then, the good and experienced moderators have more time to spend on edge cases, complex cases, fine-tune the AI, etc. In other words, it is able to do the most boring things to you.

And EU regulate, I mean as a counter example: we are not in panic about a nipple. When I was in a large museum in Paris, multiple women were breast feeding their infant. And why not? Kid's gotta eat. I'll refrain from insulting any world leaders, too easy, but you know many examples are available there regarding censorship.

Finally, it can take that BS argument away of 'oh we don't have manpower to moderate'. That is a low blow, too, by large commercial entities who could, you know hire and train? However, even a small company with not much money to burn could -in theory- win here.

I'd give this model a chance, if not only cause I've been impressed by Mistral past years. Yes, Le Chat / Vibe probably lags behind, but something like Voxtral (real-time and transcribe) is neat, and efficient.

Argue about taste. At least it is more original than OpenAI (haha, 'open') who started as non-profit and then pulled an Infantino. The Le Chat logo is also cool retro :)

On general purpose LLMs, and vibe coding, Mistral lags behind. But I find the targeted LMs much more interesting.

Voxtral [1], maybe Robostral? I just saw on their blog they released OCR 4. Their is usually informative [2]

[1] https://mistral.ai/news/voxtral-transcribe-2/

[2] https://mistral.ai/news/

That is not a suspicion, but a (baseless) theory. I mean, your proof is from Windows 98. Might as well been an outlier or urban legend, who knows.

My theory is that people in countries without extradition laws to USA are going to spray the shit out of Americans. With spam, and crap like this worm, troll farms, and the like.

Also baseless, but people who work at SOC can notice an increase since some changes happened in geopolitics this decade.

That one time I was able to beat a 'good' chess player at school :-) I still remember that one. Usually I lost, but... I also came to the chess club rather to play Magic: the Gathering, so I didn't mind.

With that in mind, it has been a while (more than twenty five years) but what you said sums up the one time I went to my first Magic: The Gathering tournament, by having to build a deck with a limited amount of packs. You don't have the advantage of owning any power cards there (unless if you are lucky). What happened? Well, you could see who was doing well based on where they sat. And, I got absolutely wrecked! But, I also had a very bad (useless) rare [Shauku, Endbringer]. Which I shouldn't even have put in my deck, but I still did, cause.. it was a rare! [1] Still, there's much more luck to be had with decent commons and uncommons. In later tournaments (different types), I did better, and my fav experience (apart from some of my cards being stolen on a later tournament at that location) was the one where you'd buy a bunch of packs, open a pack up, and pass the packs on. Some people would pick only meta cards, some people would only pick expensive ones, others would always pick rare ones. If you don't intend to win, that could be a decent strategy. But I went for the strategy where I would pick the best cards for my deck, and that worked out well.