jo909

While spinning, the blades store a miniscule amount of kinetic energy.

After removing power even that small amount ends up as heat through friction ( both in the bearing but mostly in the air turbulence). And the blades end up in the same zero energy state: sitting still.

So it is correct that a 100% "end up" as heat

Overhead high voltage conductors are not insulated with a coating, probably for many reasons but certainly for cost and heat dissipation.

That means the path through the air to some conducting materials needs a certain distance, and that even when wet or iced over or whatever can happen up there.

This view is just very extreme, it is much less zig zag. It is just mounted to the wall at the high points and slack in between. Certainly there is also a reason for the exact amount of slack like thermal expansion.

https://cdn.ca.emap.com/wp-content/uploads/sites/9/2018/04/l...

The money you buy stock with l goes to the former/selling shareholder, which is most often not the company. It is possible the company is holding its own stock and selling for cash, or emitting new shares for cash, but that is much much rarer.

Remember that you are supposed to replace the entire thing because the other components like the sensor or simply capacitors also age. It is a very cheap safety device and simply not worth taking any risks by stretching it to say 15 years instead. The proper way would be to replace them while they were all still fine by making a note in the calendar.

There are two cases:

Your products are faulty and at least one has not made their intended 10 year lifespan. I'd change them all for better ones.

Or

They have reached their lifespan and you only noticed because the first one failed. I'd replace them all.

Couldn't LLM provider just fine-tune their model for these tasks specifically - since they are static - to get ad value?

They could. They would easily be found out as they loose in real world usage or improved new unique benchmarks.

If you were in charge of a large and well funded model, would you rather pay people to find and "cheat" on LLM benchmarks by training on them, or would you pay people to identify benchmarks and make reasonably sure they specifically get excluded from training data?

I would exclude them as well as possible so I get feedback on how "real" any model improvement is. I need to develop real world improvements in the end, and any short term gain in usage by cheating in benchmarks seems very foolish.

And assumption like that are probably a great source of accidents, our mind needs to take some shortcuts like that and isn't always right. Grandmas car got sold last week and daddy is alone in the car and late for work and will be racing to get in front of you.

I'd rather have a computer keep track of everybody just the same but with millisecond reaction to all changes. Something that I can't do lacking eyes all around and processing power.

I'd guess that they used to have very good discounted contracts but are not able to renew them anymore.

The cost they charge also includes the cooling, so you could see it as for example 30c/kWh for power, 15c/kWh for cooling to compare it to a consumer contract. Maybe they have published their ratio of that split "on average" before.

I do not understand why one would fetch stacks of paper from the bank if one believes toilet paper is the more valuable type of paper you should hoard and maybe also trade with.

Seriously, why would the end of fiat cause a bank run? I would expect it would cause a switch back to trading physical goods directly, or the emergence of another currency-like thing or good (like cigarettes might have been in the past).

I have a Pi 4 home server, and the biggest issue right now is that my home upload is a bit weak for remote video streaming.

"To ensure that every Pi at our decentralized locations always has enough network throughput, the uplink and downlink is fixed at 10 Mbps."

Not sure about your use case, but for me that is way less bandwidth than I have at home.

I'm sure there is and will be demand for good static IPv4 tunnel brokers. I'm personally fine with dealing with dynamic DNS and port forwards for my home setup for now, but once I no longer have a public IPv4 assigned I would be a potential customer.

How do you deal with the global scarcity of IPv4-addresses that you would need to scale your service? I think this can only work long term if you own the address space yourself and are not dependent on some specific provider or cloud.

Also very important is a local endpoint to get a reasonable end to end latency.

No, scaleway is owned by the same parent as online.net, which is a competitor to OVH. Both are french though.

There is a very large industry that specializes in exploiting the human mind: advertising/marketing.

Of course the behavior/opinion changes they are after are somewhat limited and not too bad, and the effectiveness debatable, but they are trying hard. And machine learning is already in use.

This is what makes this no better than any other VPN service. I still need to completely trust that the provider is not watching me. Even with full hardware access, there could be an invisible sandbox or hardware DMA to some chip I can't see. There is no easy scalable solution to avoid this in a technical level. But you can be one anonymous user in a sea of many others and hope for the best, which is why I'd advise to go to a trusted mainstream VPN provider.

You would debug an unikernel just like you can debug the OS kernel. Either it brings its own internal "remote debugging API", or the hypervisor helps you with that and provides interactive tracing/debugging features and dumps with CPU and memory state that you load into a local debugger.

The kernel debugging technology exists for a long time now, and I don't think its far fetched to see public clouds exposing those abilities in a secure manner, and unikernels including their own tools for remote debugging. Any kernel can be debugged, and unikernels are no exception.

Since the Eula comes with the driver, hetzner (and other hardware rentals) shouldn't be covered directly. They don't download and install the driver for you, so aren't a party of that contract in any case.

As defined in the article: "when a computer is reset without following proper procedures (what’s known as a cold/hard reboot)"

Even if you disagree, "cold boot attack" is the established name for the actual attack, the new aspect presented here is how to circumvent a certain firmware protection that would overwrite the memory on a cold boot to prevent that attack.

If you would give your definition we could see if it is right, too.

It underscores how little you should trust your own numbers (or numbers from strangers on the internet). Forget one small factor and the whole picture shifts a lot.

"And I have one particular server that we have loved for a long time."

Yes, for over 10 years now! (I looked up the manufacturing date from the serial number, week 29 of 2007)

I feel for your acute pain, but I think you got your moneys worth and then some out of that system, there should be plenty of saved budget to replace that with a new storage server. One that has support from the vendor.

And the preemtible price is only $0.22 per GPU. Not that the GPUs used here seem to be preemtible, but this might give some hint to the baseline cost.

Well I've seen many environments, and some only ran puppet manually whenever they would before had used ssh and made a change directly. Another allowed one group of admins to just continue to log in and make small config changes, then got alerts and diffs of those changes from running puppet in noop and incorporated those changes into their git after the fact.

It's a tool to be used however you see fit and what you describe sounds reasonable, but the most common (because default) setup I've seen is the 30 minute interval.

Which is why I would argue "time of last puppet run" is mostly fairly recent and consistent for all machines, but terraform apply is most often not run automatically, not even at daily intervals.

Puppet rules are typically applied constantly in short intervals. The installed version should always be the same everywhere, the latest available from the repository. The risk is more that you can get updates installed at inopportune times. Of course what's right is that the version change is not reflected in the configuration management, but this isn't normally a problem for minor version changes.

We can and we have with mobile networks and wifi positioning.

That's just not accurate enough for many applications, and resolution varies a lot.

To be honest monthly pricing and full year commitments to get special deals no longer do it for me.

I'm sure you'll find some customers at those prices, but it does not compare with what Hetzner is offering here just because of that.

https://wiki.hetzner.de/index.php/Rechenzentren_und_Anbindun...

"How are the Data Centers connected to each other?

The two Data Center Parks are both connected to Frankfurt (FFM) and each other with dark fiber. Thus, a redundant loop is formed, which ensures the availability of a Data Center should one of the connections fail. The n10 Gbit/s connections provide ample bandwidth between the Data Centers.

The bandwidth of the connections between Nuremberg-Frankfurt, Nuremberg-Falkenstein and Falkenstein-Frankfurt are at least 120 Gbit/s. Through the Frankfurt location data is transported to the peering partners at DE-CIX and also to the uplinks Noris, GLBX, Aixit, AMS-IX, Init7 and Level3. At the Nuremberg location there are connections to Noris, KPN, Init7, Level3 and N-IX.

In each Data Center several Juniper EX Core switches, each with 64x 10 Gbit/s ports, are operated and bundle the streams of the Data Center to the n10 Gbit/s backbone and then over the various uplinks. "

And here is a list of their peerings: https://www.hetzner.com/unternehmen/rechenzentrum/

EDIT: I seem to be wrong. This is a hybrid of a EU regulation that needs to be implemented into local law, and "EU law" that has direct application everywhere. Also there are specific rules on how much stronger the member states can make their local rules. I'm leaving my error in the text below.

That is not technically correct, and the difference matters.

The EU regulation forces it's members to implement their own local laws in alignment to the EU regulation. The EU regulation can get a member state into trouble if not implemented in time or correctly. They describe a set of "baseline" data protection laws for the whole EU, but the members can still have stronger local rules. Only the local law is what can get a resident company in trouble.

Germany has (had for a long time) very strong data protection laws that are still going to be stronger than the EU regulation.

In the context of renting one of their VPS this is fair criticism. The information handed out by Hetzner is not adequate regarding their exposure and timeframe for patching. It's been several weeks since the public disclosure, and if they are still vulnerable for meltdown this puts all customers systems at massive risk.

"The host systems will be updated to fix the vulnerabilities as soon as possible. The necessary reboots will be announced on Hetzner Status. You may subscribe to be notified. "

The VPS's were never billed hourly, so you wouldn't have used them in a cloudy, disposable, dynamically scalable way.

As far as I understand this is a historic safety mechanism.

Trains were very popular and important in the UK, transporting nearly the same amount of passenger pre World War 1 as they do today.

At this time no central locking existed and many train compartments, especially for the rich, had their own door that opens to the outside.

With handles on the inside it is much more likely to open by accident. The door contained the only window in the aisle between the seats of the compartment, where you would stand to look outside, so people naturally lean against it.

At the time central locking came around it wasn't necessarily reliable and people were already used to the situation, so there is little reason to suddenly introduce internal door handles again.

His conflict started with the Ontario Human Rights Commission policy which clearly includes gender pronouns. That is often getting mixed up with when he also spoke out against C 16 later, with different arguments.

http://www.ohrc.on.ca/en/policy-preventing-discrimination-be...

I think you are right that just in C 16 there is nothing about compelled speech. My understanding of his position is that it's part of a slippery slope, and the interpretation of discrimination or hate speech could easily extend to using the wrong gender pronouns.