Why the caveat that it can only be done if it hasnâ(TM)t been launched recently? So if the user accidentally clicked the wrong thing once, the uninstall will fail? And why allow the user to reinstall something IT has said it doesnâ(TM)t want on the PC? None of this makes sense. It can uninstall it, maybe, but it can come back at any time and if the user actually runs even once a month, it canâ(TM)t be uninstalled? Feels like this activity is a response to some legal action or a demand from a very big customer, itâ(TM)s not really providing IT with a useful capability.

That's an option but since I had to replace existing equipment I felt it easier to go with all electric. Also, sunshine is free and propane is not :)

My regular wristwatch runs about 5 years on a single battery.

Get a Solios watch. My "nice" dress watch is a Solios. The dial is a PV panel. The device is designed to last 30 years without requiring a battery replacement :D Very nice watches, designed in Quebec.

You definitely need a new smartwatch. I wear a $50 smartwatch that has all the usual stuff, I get a LOT of notifications on it, and I only need to recharge it once a week or so (maybe six days).

Also, gasoline generators - I got rid of mine and replaced it with batteries because gasoline generators are a horrible, horrible emergency power solution. You can't store gasoline indefinitely. You can't store a primed gasoline generator indefinitely. The fuel goes bad and the fueling system gets contaminated. Emergencies don't happen often and I don't have time to do a carb clean and rebuild when the power goes out and I need my freezer to keep running. Solar-charged battery solutions are about seventeen trillion times less headache than a gasoline generator, especially the 2-stroke bullshit generators that a lot of people use (because cheap).

OK, I consider the Jackery et al "size of a gas can" type devices (which btw are awesome, I am saving up to buy another one, and the solar addon) - to be not a power bank, but a backup power supply. When people say "power bank" and I think specifically in the context of this conversation, I'm thinking of pocketable devices that can charge a phone.

Didn't even think of this but you're absolutely right. And even worse - phones generally have a thermal cutoff for charging ("This iPhone is too hot. Charging has been suspended until it cools down"). But I bet power banks do not - the ones I've disassemble don't even have a thermistor on the battery.

Actually I messed up the math a bit :) but the premise is sound. Yeah I hear ya. I've just seen too many reviews of these that pointed out that the PV cells were barely there (in some cases actually fake).

I have a couple of fold out three-leaf panels also, they are designed to hang from a backpack while you walk around and charge. This is wayyyyyy more surface area than just the side of a power bank, though.

Probably not even that, for most places that people actually live. It's January and I'm in southwest Indiana. 83 W/m2 at the best time of day and best weather conditions.

Remember when VW vehicles used to ship with a solar cell in them to keep the battery charged while in transit and on the dealer lot? Early 2000s I think.

If you have a vehicle at hand, you already have a better charging solution than a power bank. A power bank is a portable, pocketable backup power supply for your phone for when you're walking around out of range of fixed charging or vehicles with power outlets.

There is no way whatsoever that a PV cell on the side of a pocketable power bank is large enough to receive enough energy to be meaningful. These are joke features. The areas of the world with the highest insolation offer about 300W/m2. The size of a solar cell you can fit in your pocket is perhaps 0.01m^2. So you're looking at 3W of available solar energy - if you are in a subtropical desert at the right time of day and year and with the panel aimed optimally. Take out conversion efficiency of at most 20% if you've got a really good panel (which a cheap power bank will not - probably more like 10%). So, somewhere between 300-600mW. Even ignoring losses in the charger circuit, a typical 10,000mWh bank would require somwhere between 17~33 hours of continuous optimal illumination to charge.

Without arguing that point - here's the counter argument. Regardless of what goal is in mind for these AI datacenters, the stuff in them is being rapidly obsoleted. This is kinda like the early days of bitcoin mining where people were using ever more GPUs and later FPGAs until someone designed ASIC miners that were an order of magnitude more power efficient. Even assuming the market for "AI" keeps growing (which it will not; basically every study that's not sponsored by nVidia shows that AI adds nothing to the bottom line) these datacenters are going to be populated with silicon that is on death row even before it gets powered up.

Trust me, here in SW IN people really really really really understand the problem of electric bills skyrocketing. Besides the corruption and also a president ordering a coal-fired electric plant a few miles from here to stay open after its rainbow bridge date, there has been STRONG activism against the incumbent electric provider. They may find themselves having to deal with torches and pitchforks soon. The local datacenter project got something like 100/3 against/support for this exact reason.

Reminds me of the start of The Running Man... "Food riot in progress. Approximately 1500 civilians. Moving in".