On the other hand, the quality of the people developing the kernel seems to be higher than those developing systemd, or at least that is the perception I get from reading all the hate on the Internet.
Hint: That is not a reliable indicator.
Not really 40 years that would be about edition 5 and the Bourne shell was not even invented yet. The SysV init, especially the link farm aspect, is quite different to edition 7 init or even BSD 4.3 init. Once
inetd used to be configured by a single file not a collection of files in
I venture to suggest that if you took a unix sysadmin from 40 years ago forward in a time machine they would be pretty much at sea configuring any modern linux and I don't think their learning curve would be any steeper for systemd based init than sysV based init. They would not be familiar with xinetd, they would not even be familiar with inetd. If your mail isn't getting, and they were from 35 years ago they might waste some time work out where the uucp configuration had gone. And this is just the start. A LOT has changed in 40 years. Think of typical sysadmin tasks such as installing software and configuring printers, GUIs, networks, firewalls and httpd. And what is this "grub" and how can I reboot without console toggle switches and how do I change a disk pack? FWIW I no doubt count as a "greybeard", and I personally regard the SysV way of doing things as a wrong turn, but I am still not too old to learn systemd. I found the documentation adequate (or even good by linux standards).
Modern IQ tests consist of multiple sub tests - 15 for the WAIS IV. General intelligence "g" or a combines score is a weighted sum of the abilites in all these sub tests. A competent psychologist will look at the sub test scores for a richer interpretation, but that doesn't mean that the combined score is useless. General intelligence is based on the observation that people who are smart in one area tend to be smart in other areas (unfair though this may seem). Of course, this isn't always the case and sometimes people who are smart in one are are not smart in other areas. It is more accurate to think of modern IQ tests as a combined measure than as a single measure.
Think of it like CPUs. A combined benchmark like specmark or passmark can not fully characterise CPU performance, and sure, if I want a precise comparison, I need to define exactly what my load will be, yet you will be hard pressed to find a work load which a Pentium 3 performs better than an I7. So it IS useful to have aggregate measures of performance - so long as they are no over interpreted.
Some people don't like IQ tests because they see them as discriminating against socially disadvantaged groups, or racial groups. It is true that there is an issue of cultural bias in IQ tests, which people try an eliminate but can never do so completely. However, used properly, IQ tests can actually help people from disadvantaged backgrounds by identifying those with academic ability which may not be manifesting itself as academic performance for other reasons.
old school AC transformer units have miserable efficiency and are both heavy and bulky.
I'm not quite sure what qualifies as "old school" but AC transformers even at 50/60 Hz can be quite efficient though I grant you, they are heavy and bulky.
Apart from size and mass the big advantage of switch mod power supplies is that they can be regulated with no appreciable loss in efficiency. The alternative is pretty much a linear regulator and that combination (of transformer and linear regulator) tends to have miserable efficiency.
So you would say it is OK for a hotel owner to refuse service to blacks?
It's like virtual particles. Particle (cash) and anti particle (debt) pairs can spontaneously pop into existence. So all we need is a black hole to swallow the debt and cash is the equivalent of Hawking radiation apparently emitted by a black hole. I surmise that the worlds current financial problems are due to the black hole(s) finally evaporating.