ED: Looks like it's 24(!) hives per beehome, and they charge $2k delivery ($83/hive) plus $400/mo ($400/hive/yr) for maintenance.

Clearly not something of use to amateurs, and I'm not sure whether you can make that economics work out for professionals, either. I guess it depends on how truly independent it is, vs. your local labour costs.

There is little correlation between "presence or absence of pollution" (what a general term to begin with...) and CCD. There is a strong correlation with the presence / absence of varroa. And this system treats varroa.

I've been thinking about getting into beekeeping (I first need to increase the accessibility of my ravine where they'd be), and had been thinking about a sort of high tech solution, with electric blankets, heat-exchanging baffles, a flow hive, and maybe some mass and/or noise sensors for monitoring colony health. But this is WAY more high-tech than I envisioned, and honestly I'm scared to even look up the price ;)

The best correlation between the presence and absence of CCD is varroa. CCD is probably not single cause, but if you had to pick only one, varroa is that one.

And glyphosate is probably WAY down on the list. At least pick an insecticide as your culprit.

Not with far-UV (~222nm) you didn't.

Leaders aren't there out there e.g. building the rockets or doing the vast majority of the engineering. Musk doesn't get credit for that. But they do set the culture and direction for their companies. And in this regard, the "build quickly, launch quickly, fail quickly, learn quickly, and iterate quickly" culture developed for SpaceX happens to be very effective. Musk gets credit for instilling that. Another thing he should get credit for is the broad design strokes such as "focus on designs that are cheap enough that they can be mass produced, gaining you economies of scale and the ability to iterate quickly during testing, but are still capable of being reused" (this differs from the two previous predominant paradigms, either super-expensive low-volume reusables, or cheap high-volume disposables).

I don't like the guy, but absolutely, credit where it's due.

Except that Musk is allowed to show up at work or not whenever he feels like it, so he's a giant hypocrite who forces everyone else to do that which he himself is not willing to do.

I think a lot of people miss the fact that SpaceX engineers know very well that what they're doing might fail spectacularly, and that this is the cost of speed.

A random example: autogenous pressurization.

It's beneficial to have a rocket's engines pressurize the tanks themselves rather than to haul up pressurant tanks and a separate pressurant. But it's surprisingly tricky. For a methalox rocket, you ideally want hot methane injected into the methane tank, and hot oxygen into the oxygen tank. But hot oxygen is very difficult to work with in an engine, as it tends to eat your engine.

If you're still working on reliably producing hot oxygen, there is a hack available to you, but it's not pretty: just inject exhaust into the oxygen tank; after all, it's not combustible. BUT, it is water and carbon dioxide. Both can settle out as frosts or plated ices, and in the liquid, the water ice will float at the top, while the CO2 will form a snow at the bottom. Frosts / ice plating can block e.g. your RCS jets. The CO2 snow will kill your engines. You can put in filters around their intakes, but it'll clog your filters. You might try expanding the filters, and maybe that'll work for a while, but then you rotate the rocket, the snow rushes ti one side, and a bunch of engines die from clogging. You may put some big mesh plates across the whole tank to keep the snow off the bottom, but they can cause their own problems with fluid flow and still sometimes clog or let snow through during maneuvers. Etc.

So then comes the question: put Starship on hold while working on getting the engines to reliably produce hot oxygen, potentially for years, or forge ahead with a hack solution that you know has a reasonable chance of killing your rocket?

To SpaceX, the question is obvious. You cannot afford to give up years of critical flight data just to avoid some booms. The decision is immensely lopsided in favour of "put in the hack solutions and launch, while you work on the proper solutions". Because you learn SO much from every launch that can be used to evolve your design. And you also learn so much from every rocket that you build, whether you launch it or not, so you might as well launch it.

To be clear, you don't want to lose rockets due to doing stupid things. Like, for example, if it turns out that some SpaceX engineer installed the wrong COPV and caused the recent pad explosion**, basically the only thing they would learn from that is "have tighter controls on your COPV processes", which isn't at all worth the cost of the explosion. But in general, if you launch and it clears the pad, you're getting good, important data from it, it's worth it even if it blows up seconds later, and it's on to the next evolved version of the rocket in your production sequence with both production- and flight lessons learned.

** It's clear that the recent explosion was from a COPV failure, but it's unclear why. Some claimed leaks state that a COPV may have been coded to a higher pressure than it actually was during production, so when they scanned it it checked out as being the right tank, but actually was not designed to handle the needed pressures. But I'll wait for official confirmation on this. SpaceX only makes some of their COPVs, usually not the smaller ones - ones that have washed up ashore were made by Luxfer. So this could be a supplier problem, like the strut failure on a 2015 Falcon flight. But again, too early to say.

"What am I missing?"

That the author of this article is an idiot.

Yes, humans went to the moon in the 1960s. It also consumed a huge chunk of the federal budget. Adjusting for inflation by NASA's NNSI inflation index, the entire Lunar program cost $288,1B. If the US were to prioritize a project to the same degree today as then, accounting for GDP growth in inflation-adjusted terms, it would be $702,3B. NASA's annual budget is around $25B.

The cost of access to space today is a tiny fraction of what it used to be, when accounting for inflation. And keeps pushing lower. No, it's not "easy", but it absolutely is being done.

I've read that dropping X11 in favor of Wayland goes beyond GNOME and reaches all of GTK. This means developers of other GTK-based desktop environments (particularly Xfce) and users of distributions built around those environments (I have Xubuntu and Debian Xfce in mind) will have to make hard decisions.

I think what happened with FOSTA is that some of the more conservative states or the federal government redefined "sex trafficking" to include not only sexually-oriented human trafficking but also all other sexually-oriented commerce.

True, higher-priced prepaid mobile phone plans have unlimited text. Lower-priced plans do not. SMS-based 2-factor authentication is not cost effective for a subscriber making do with a lower-priced plan.

Then you must not have followed FOSTA, a 2018 US law weakened the protections of Telecommunications Act section 230 with respect to sexually explicit material. This made it easier for acquiring banks to get away with debanking sellers of erotica.

How do minors (children and teens not yet old enough for a bank account in their own name) pay for things in Sweden?

When did these advantages of "a locked down hardware BOM" go away?

- Easier for the audience to compare a game's system requirements to a particular device model
- Console maker's imprimatur makes it easier to sort through Theodore Sturgeon's 90 percent crap that fills more open online app stores
- Restriction against software modding reduces likelihood of cheating in online ranked play against strangers
- Restriction against software from unidentified publishers eliminates need for intrusive real-time anti-malware
- Console game discs and cartridges can be resold and used even if you are stuck on slow satellite or cellular Internet with its low monthly cap
- A GPU for a desktop PC costs as much as a console, and then you still need to buy all the other parts of the PC if you want to be able to use the console for something else while someone else in the household is using a PC for gaming
- Consoles have historically had more couch multiplayer games than PC

As for tablets and phones, how many people carry a Bluetooth controller with them? A virtual gamepad on a flat sheet of glass offers no tactile feedback as to where the player's thumbs are. Not all games adapt well to that.