The temperature change of a building depends on two factors: direct heating via radiation from the sun and indirect heating by temperature exchange with the surrounding air. The former is in normal conditions much stronger and the primary driver. It's why blinds are so efficient. This factor is directly proportional to the solar radiation, naturally. Indirect heating is delayed since ground-level air indeed accumulates heat until 15:00 to 16:00. At the same time, it's much harder to transfer this heat, especially for a well built house. Sorry, cardboard derivatives don't count. So yes, there is cooling need beyond the solar peak production at noon, but it is much lower once the sun starts to set.

The main reason for the duck curve has little to do with solar radiation, but that it is still very common to let the houses heat up over the day and then suddenly cool down when people come home in the evening.

This has nothing to do with base load or how the electricity is generated. The core point is: to actually do something about climate change, we have to electrify all those things INDEPENDENT of the whole question where the electricity is coming from. Not acknowledging that is just distracting and derailing the discussion.

Now things like Air Conditioning doesn't actually have to increase grid load at all, because for well built houses, cooling demand strongly follows availability of sun light, so PV can perfectly well cover that. Heating on the other hand is a natural buffer system, so any work on electrifying heating on a large scale can help in shifting demands to better match production. That is useful for grid operation in general, again, independent of how electricity is created.

Or maybe we should stop comparing "how we generate electricity" with "all the primary energy use we have". Going nuclear is not magically going to fix the latter.

A good Wifi 6 access point can actually easily saturate a 2.5G ports. The theoretical limit is just under 10G per PHY and you will often see multiple PHYs in the better models.

There might be no need for FTTR, but that doesn't mean it isnt desirable. Power consumption alone is a good argument. 10G Ethernet over CAT5/6 typically requires 2-5W for the PHY, compared to less than 1W for a SFP+ Fibre module. Here, that would easily sum up to 10-20EUR per year for a permanently runningdevice.

With Wifi 6 in every room, 10G connections would actually be desirable for the access points, with Wifi 7 even more so.

Are you aware that the primary cooling methods of lakes in summer is evaporation? Standing waters are surprisingly good at keeping their temperature...

Have you tried thinking that plan through? There are two efficient methods for cooling (locally): flowing water or water evaporation. In summer during a heat wave, both fail. That's not even an issue specific to nuclear power stations, all thermoelectrical power station share that problem. Just letting water sit in a ditch doesn't help all that much.

God, this is such a BS, it's not even funny. Just to put that number into perspective: The cost for on-shore wind energy is currently around 1.2m EUR/MW. 500 billion EUR is enough money to build a capacity of ~400GW or five times the current peak load. That would in fact be enough power to replace all primary energy needs 1:1. Also just for fun: Germany is currently paying something like 60 billion a year for importing fossil fuels. Suddenly, even that BS number doesn't sound all that much, given that it is *poof* gone. Seriously, if you want to make a case for nuclear power, at least use real numbers.

The most important part is missing. TOTP authenticators are not phishing-resistant. They don't ensure that the website asking for the code is the one the code is intended for. That's also why remarks like "TLS already can do mutual authentication" are completely missing the mark. If I connect to s1ashdot.org and get a valid HTTPS certificate, the user doesn't necessarily notice the difference. With FIDO tokens, the public key authentication is also bound to the domain.

The broken function has been around for many years, the latest installation of pushing it down everyone's throat is second quarter of this year. But I guess the typical American nowadays can't tell if a video title sounds like it is written by the current US president, so nothing lost.

If Google's management at Youtube at any kind of diversity, they would know how most multilingual people around the world react to this. AI translations can be a great help if you interact with someone you couldn't interact before, but getting it forced down your throat for a language you know better than the AI?

The only reason why the jobs are obsolete is because the managers don't care about their own products, but just what business school taught them. Copywriter or editors for a magazine with actual content are essential for creating good content. AI absolutely cannot replace that, it can't even do a very good job for grammar correction for many languages. The common theme is replace work with a soulless automation because no one cares about the qualify of the actual result.

Let me give you a very personal exampe of why it sucks. Youtube recently decided to force down translations for video titles down the user's throat. It has exactly one control: the UI language. Most Americans might not understand the problem, but if you are ever fluent in another language, the quality of the automatic translations are infuriatingly hilariously bad. It says a lot about the culture in Google's management that they pushed this change through and no one thought that spending a bit more time on it might be a good idea. But I guess I shouldn't be surprised given that Google cancelled their DEI policies a while ago...

What gives you the right to be a threat to every other car on the road? That's what the "I don't need a break every 3-4 hours" argument boils down to. With modern high-speed chargers, going from 20% to 80% can certainly be done in 15-20min and if you don't skimp on the battery, it will last you 3 hours. So the difference between an EV car and an ICE is less than 10% if you ignore road safety. Otherwise it is the amount of break time sane drivers will take anyway. Ironically, one of the most vocal ICE lovers I know used to take trips with an older car that needed a forced break for overheating issues...

The core of the issue are two fold: perceived exceptionalism on the side of the ICE driver and unwillingness to change. Long distance trips with a EV are certainly possible, they only require a bit more planning and enforced breaks. But ICE drivers are exceptional human beings that can concentrate on a task for half a day without needing a break unlike professional drivers or traffic controllers.

EVs are vastly simpler than hybrid cars, which are kind of the worst of both worlds.