If you know what it actually treats, you also understand that an underlying inflamation by unknown worms is resolved, hence one thing less for the body to worry about. I was actually surprised about the other uses of hydroxychloroquine in regular medicine, read: not for malaria, that it had emergency approval made sense.

The fact that your country allows to buy veterinary medicine and ethanol (read: not denatured alcohol) off the shelf might be more troublesome to begin with. That is not the same within the European Union. In fact, since this year it is even illegal to use a non-registered source as a veterinary medicine. Hence if you were using oxalic acid in the past, you may now only use a registered counterpart of that same substance as medicine. And I think that kind of legislation is bat ass crazy, it is the same material at prices of at least ten folds higher, but you can see the contrast.

I wonder how this would have played out if Redis Corp would have relicensed towards AGPL. If Linux Foundation would have then forked towards the earlier, more liberal, BSD license that may have had some serious political impact, as a "non-profit" serving their donating members, assisting on removing a business model for the original copyright holder. In our reality Linux Foundation is still on the moral highground.

The big difference between DuckDB and PostgreSQL is that for DuckDB you don't run a separate database process, it is an in process database. This is directly the major drawback of DuckDB: you can't do concurrent read/write operators. In my perspective DuckDB replaces remote crunching solutions, because it made column storage accessible in the formfactor of SQLite.

Dismissed

With a few farmer protests all the pesticide reductions were gone. New elections are coming, lets see what remains of these ambitions.

What about: no backup? Wouldn't the primary thing for this kind of research not to spread the risk over multiple sites?

Please read the amount of heat they start with: Czech: 160 degrees. In Iceland: 130 degrees. That is high calorific heat. There is no dispute on that being valuable (and maybe even a bit too much). But the argument that has been made for quite some years now: 25-33 degrees maximum is barely usable for this purpose. Low temperature water heat distribution is in the ballpark of 30-60 degrees already. So in order to get it there a significant electric investment should be added, and there the question rises: why are we trying to do this for this poor heat resource in the first place?

There you have it: locally. Now transport the water that is at 25 at most 33 degrees over long distances (since data centers are typically not nearby urban housing, not even in Amsterdam) and you would end up with... nothing. But nothing is a perfect way to lose the heat... for free...

The low calorific energy that comes from cooling would maybe allow a heat exchanger to have some degrees offset, but is not as useful as industry scale residual heat. To me it sounds like the lobby of the past years by the industry: getting free cooling (water).

So no.

China announced the launch of a liquid-salt thorium reactor yesterday.

When you state "it has the wrong bootloader" do you mean it does not work like one could expect from true pxe implementations? "Most PCs" here is only the PC's that have UEFI enabled, a common use case like booting Virtrual Box is hardly possible with the EFI option enabled. Principally what they implemented has cut some corners, personally I would like to see that fixed. The Option97 that does the "RPI" identification might be a stable way to identify the boards, but just feels like a Microsoft approach to standardisation. So the real benefit would be we don't have "Internet Explorer" approaches in network booting, that it may have migration costs is something that is obvious, but nothing prevents the end user to use a hybrid set up to catch both the categories.

Why isn't aarch64 correct in your opinion, but "0" (x86) is?

It is obviously nice that such functionality appears. But the current PXE identification of the RPi4 is an x86 PC, instead of aarch64. There are other attributes that prevent an ordinary identification of the board outside of the MAC address realm. Fix those things first.