It is surprising that this topic hasn't been studied in depth by the scientific and economic community, right? Oh wait, it was... This article is the same FUD as the recent "wind turbines are creating toxic trash" crap.

More like resets. An AED will typically stop your heart. That's still far better than the arrhythmic pulsing it did before, since those actually damage the heart muscle without pumping any blood. But you still need the CPR to actually keep the patient alive.

Please don't spread misinformation. Breaking rips is not a sign of doing it right. It is a sign of doing something though and a broken rip is far easier to mend than any kind of brain damage from lack of oxygen.

Funny that you claim that C++ is not stable when talking about Rust. The number of backwards incompatible changes in C++ is much lower than Rust per generation and Rust has a heck more generations than C++ per three years.

I'm not sure if that is supposed to be ironic or not. Encryption for one is a heavy SIMD work load and one of the primary use cases for AVX-512 even. Most video codecs are heavy SIMD too. Database systems can be a mixed bag, but there are quite a few parts can benefit from SIMD too. That leaves essentially the "web traffic" item of your list. Given that node.js fundamentally depends on IEEE 754 arithmetic, I'm not sure whether you know what you are writing about or whether this is trolling...

And what a brilliant PR stunt that was by Israel. There used to be a huge difference between being anti-Semitic and being anti-Zionist, but thanks to relentless efforts of Israel over the last 30 years, the second has gone now.

Reference counting by itself is not even a garbage collection technique. You still need cycle detection.

I love how everyone goes on and on about data races. Preventing data races is for safe concurrency what standardized tests are for quality education: they seem superficially related, but fall apart when looking at the details. ACID doesn't ensure safe concurrency by itself, but transactions at least provide a useful building blocks for creating systems that can provide safe concurrency. Data race freedom is a red herring at best.

IPv6 is not IPv4 with larger addresses and never was. That's BS that has been repeated over the year mostly by people that refuse to work on updating their networking knowledge and often also still talk about network classes because they haven't heard of CIDR. The reason why IPv6 has been fighting since the beginning against stateful (!) NAT is that it breaks one of the core goals of a good network, end to end addressing. I find it amusing to see you ramble about STUN etc. as it is a horrible hack around the problems created by NAT. The reality of the world is that there are a lot of reasons for incoming connections to devices in a home network. Games, VoIP etc. all need hacks to work around the addressability issue.
There are very good reasons why every good network admin should have been thinking about IPv6 support for a long time. It can help massively improve the architecture of the local network by removing all the scarcity reasons that created the mess a lot of larger installations are. Old internal devices are often used as scapegoat, but often are just excuses as dual stack support just continues to work. For some reasons, a lot of naysayers just pretend that it is an all or nothing decision.
We are past the point where major international ISPs have moved towards a native IPv6 core network exactly because it allows them to cut a lot of network layers. CGNAT is expensive for them too and doing it on the edge is more so. At the moment, it is mobile ISPs pushing IPv6 because it provides significant benefits for them.
If you as sysadmin believe that systems that are not directly addressable are safer, you might want to look for a new job. The number of attacks using various devices like phones or printers or end user systems as proxy is so long, that it isn't even funny. If it can connect to the internet, it should be assumed that the internet can connect back.

I have no idea what you mean. Prefix Delegation has been working exactly as intended for many users in different networks for years. Please don't confuse Prefix Expiration with Prefix Delegation. They are not the same thing. The former happens if your ISP likes to cut your link for some reason and then isn't willing to give you back the same prefix. So the router is between the hammer and the anvil, it can't retroactively reduce the announced life time of the prefix and always sending a much shorter life time creates a different set of issues. But that's not really a problem of Prefix Delegation.

"Hiding" the network behind the router would pretty much defeat the purpose of dropping NAT in first place. It is also the wrong objective. The IPv6 Private Extension in combination with an inbound connection filter pretty much provides all that NAT ever did without many of its problems. PE essentially means that a client will have two or more IPv6 addresses at all time. A stable address for incoming connections, e.g. based on the MAC address, and one or more temporary random addresses for outgoing connections. Depending on the OS and the settings, a new one will be created every so often, so you can't reliably enumerate the hosts in a network.

South Korea also has shown that it takes only one moron to effectively break containment. In the case of SK, a single cultist moving from one congregation to the next is responsible for thousands of infections and many direct deaths from that.

You can always save the PDF to disk or open it with a separate application.

Frankly, I do blame America. I can understand mistaking normal hornets for the Giant variation if you are nowhere careful. But something is deeply wrong with your education if you can't tell bees, wasps and hornets apart and at the very least you shouldn't be hunting insects in that case. But completely mindlessly rushing ahead seems to be much stronger in certain parts of America than most of the first world.

I remember not to long ago when people said this is no worse than the flu. The numbers of death in the USA have now passed the estimates for the entire 2017/18 flu season. That one was one of the worst in recent time and had case numbers 45 times as high as the current numbers for the corona virus. Factor in the lack of natural immunities and even with a factor 10 of yet-unknown asymptomatic cases, a million deaths are still entirely realistic if containment measurements are dropped on a US wide level. As far as the icebergs are concerned, you must be an utter moron or leaving in a very narrow bubble to not recognize that the loss of arctic ice is likely beyond any saving with the way the world acts...