HDDs are still the most cost effective solution for large storage arrays that don't need particularly fast random data access, although putting an SSD in front of the drive array to act as a cache can make even some of those workloads viable. I think the issue has been more that the size of the array where that becomes a significant enough cost difference to offset the "screw it, let's just go all-in on SSD" has been increasing rapidly.

For instance, it used to be that media creatives would have a SSD for their go-to / work drive and a high-TB HDD or RAID to store the bulk media data, but - at least until AI blew the market apart - unless you were either seriously budget-limited or producing a vast amount of raw content, then a lower-spec high capacity multi-TB SSD or two was a potentially affordable option. In high-end server land, it was similar; you were spending so much on things like per-core software subscription licenses and however many chassis full of CPUs/RAM, that the storage uplift from HDD to SSD on the drive arrays (excluding the stuff that really needs to be SSD, like VM image storage) is largely a rounding error for PO approval until you get up into the 100s of TB or even PB range. But again, then along came AI...

I suspect a lot of people with upcoming hardware refreshes and large SSD drive arrays are going to be taking a good hard look at how much of that data *really* needs to be on SSDs until the AI bubble pops. It might be a bit of a last hurrah for the tech, but the next few years could be very good for distributors and other bulk suppliers of HDDs if those reviews go the way I expect.

Yeah, that's what I was getting at. Subtract one number from the other, and you get the profit on the movies from theatres - after all the usual Hollywood math has been applied. That's a few points in the black, which is fine by some metrics, but is still only for the movies made since the original trilogy, so my main point was that it doesn't seem to include the profits coming from other areas, which are going to make the RoI for Disney look a lot better overall.

Point taken on the level of success on the movies alone though. Ignoring all the other income streams from the franchise, that's still a pitiful amount of RoI for a franchise with a reputation and loyal fanbase that should have ensured every single one of the movies landed in the top end of the highest grossing movies of all time list if they'd even been remotely decent. I doubt Disney are going to be unhappy with the results, but equally they are probably not as happy as they probably hoped to be when they bought LucasArts.

It's not the best phrasing, but that's just on the box office (after "Hollywood Math" has been applied), and it's a profit which is as far as many people will need to read. I'm pretty sure once you add in the TV shows for overseas broadcasts and people who subscribed to Disney+ specifically for them (less their production costs), all the license costs for all the toys/collectibles/video games, etc. produced by third parties, and all the other ways Disney rakes in money from the franchise it's even more financially rosy for them.

Because it becomes more cost effective to use less environmentally-friendly alternatives that produce more CO2, like coal, instead of natgas.

Profits > somewhere healthy to live. One of the many "Fsck you, I've got mine!" mantras of unfettered capitalism - it's almost like they're trying to come up with their own version of the Ferengi Rules of Acquisition...

So you have the truck equivalents of Marine Pilots, combined with an airport "drop and go" style parking system. The autonomous truck arrives at the lot, parks up in the designated assigned at the gate or whatever waiting bay, or queue, and when there's a loading dock ready, the "pilot" manually drives the truck the last few 100m.

Besides, while it's certainly possible that all the loading docks might be in use, if you have a "complete jumble of trucks", then that's a fairly major yard design failure and major safety issue since anyone on foot will need to content with trucks coming any which way. Surely any one with the slightest clue would come up with a layout with a suitable combination of one-way systems, a suitably sized overflow lot laid out like a holding park at a ferry terminal, and a whole bunch of clearly marked safe walking routes. Arguably, the pick a dock and arrival time issue shouldn't be a problem either - modern cars are some of the most connected devices of all time; what makes you think any autonomous trucks will be any different? It wouldn't be that hard to come up with a system that tells in an inbound truck that there won't be a bay free for 30min or whatever, and to slow down a bit on the highway to delay arrival accordingly, or just go around a suitable block a few times.

So, ~4,400 satellites that are now going to be a little over 10% closer to Earth. That means they are also going to be ~10% closer to each other and moving across the sky faster as well, and therefore the light trails they leave across the sky in long exposures will be even more densely packed.

I felt a great disturbance in the Force, as if millions of ground-based telescope using astronomers suddenly cried out in frustration and were suddenly silenced.

It might just be profiteering, but beside a little RAM they do contain other ICs which, like RAM chips, are all made on wafers. Any fab capable of turning blank wafers into ICs that is interested in turning a profit (e.g. ALL of them) are going to be prioritising those lines with a higher profit margin (RAM/GPUs/NPUs) and ramping up costs to match demand. That is eventually going to create a knock-on supply/demand problem for other chips that are not in high demand to fuel AI startups as well, and as supplies of those chips run low, that will drive thier prices up. Combined with tariff/export tit-for-tats, these kinds of supply issues were starting to impact the automotive industry some months ago, so it shouldn't be surprising it's now expanding into other sectors.

Asus is a big company, so probably wasn't entirely based on a "just in time" supply chain. If they are running low on parts inventory and that lack of supply of general components is now starting to bite, then we're going to see similar price rises from other vendors fairly soon too. If so, then I'd be preparing for hardware pricing to spike across the board before the AI bubble pops and making "buy now or wait" decisions accordingly.

The analogy you want is "handbags". Leica makes some very capable cameras but, like Hermes with handbags, they realised a long time ago that they have a very devoted following that is willing to pay an awfully big premium for that little red badge, and thus started the seemingly endless chain of Leica's "limited edition, collector's models" and then, inevitably, along came the copycats. Kudos to them for finding a new business model in a shrinking market, I suppose. Anyway. Like the top handbag fashion brands, functionally they are not much better than the equivalent product from any random high street store, and although quality is often better (but not always), you're mostly paying an awfully big premium for the "right" badge.

The purpose is the same, too. It's not just, or at all, about the photography, or having something stylish to carry your stuff in; it's about being seen to be on trend (usually because some random influencer/celebs was recently photographed holding one) and having a lot of disposable cash within the right set. You know what else all these cameras have in common? They are compact enough to fit easily into an expensive handbag ready to be brought out on a moments notice whenever the paparazzi show up.

Who cares, as long as the company that doesn't need that group of workers in 6 more years pays their 1% and the re-training cycle repeats? Well, at least until everyone has been downskilled into menial work at minimal wages; typical peasant labour doesn't generally need all that much training.

As always, when people propose things like this, the first thing to do is follow the money and figure out where is that "1% for retraining" is intended to go, and it doesn't take a genius to realise a lot of it would end up going to companies like the Khan Academy. Not exactly the most subtle bit of shilling for self-interest's sake, but still better than some of the far more blatant "do this (and give me money)" sales pitches we've had of late.

What's to complain about? They're no secret, nor is their purpose. A government subsides a developing technology to offset the initial investment required for companies to undertake development, drive market adoption once they have a product, and establish a high market share and, ideally, market dominance for your preferred - e.g. domestic - manufacturers. Once that happens, a government will generally try to recoup those subsidies through taxation of sales revenue, and - in some cases - on domestic users of the product (e.g. the UK's plans for a per-mile tax on EVs). The size of the subsidy generally reflects their confidence in the size/importance of the potential market, and therefore their ability to recoup their investment. EVs are not the first market this game has been played with, and it surely won't be the last.

A government pulling those subsidies, while their competitors maintain theirs, is simply them saying they don't feel this market is going to yield a return on their investment because reasons, or that they feel the money is better invested in other markets with a larger potential for return. The governments that maintain their subsidies are simply placing a contrary bet. No, it's not a "free market" move. There never has been a "free market", so stop kidding yourself about it - capitalism and free markets have always been about protectionalism of corparate and national interests first and foremost, and always will be.

The real question here is which technology you feel will be the long term winner, ICE, EV, or maybe even something else entirely? Given that, which goverments are playing their hand correctly should be QED.

That's what I'm puzzled over. Surely you can already do all this with additional accounts and forwarding rules, and do so without any of the stated restrictions? Google even lets you link multiple accounts and have them appear side by side in the same inbox (in both the web and app UIs) if check the settings, if that's what you want to do. Or you can mix and match (I have three main GMail accounts; two linked for regular email and mailing lists, and one completely separate, plus $deity knows how many throwaways over the years since it was still in Beta).

The only possible usage case I can see for this is for people too lazy/stupid to set up the required forwarding rules and manage the account transition process themselves. I suspect that is going to be a depressingly large number of people though...

Hopefully not, at least in the commercial space. If so, that presents a massive opportunity for Linux and other FOSS tools to gain an advantage where it really matters if they do take this approach; on the bottom line.

At some point arguments like: $xxx for RAM (per server, per month) + $deity-knows-what in commerical software licenses (again, per month since we're renting in the cloud) vs. let's say a reasonably achievable target of 1/3 of that on RAM (per otherwise identical "hardware") + no software costs to do exactly the same thing, and probably do it faster, is really going to start to register in the C-suite.

And, with a few adjustments, that argument even more so than cloud if you are doing all this on-prem.

Not quite. There's an implicit dichotomy, yes, but the way to cut through it is more about being tolerant of other viewpoints and accepting of the fact that some people may have views or beliefs different to your own. Free speech entitles you to say what you like without censorship, and specifically government censorship in the US, it does not require others to listen or accept what you say. The flipside is that others get to say what they like about you or things you believe in, but that doesn't mean you have to listen or accept what they say either.

It's perfectly possible to be a devout follower of one faith (or none at all) and still respect the rights of others to worship whatever gods, aspects of nature, or whatever else floats their boat. It's when they decide to act on things like chats of "death to the infidels" or whatever that it gets messy, but at that point it's also more about actions than beliefs, and those tend to fall under much clearer legal frameworks.

Yeah, that's the tricky thing about "free speech". It cuts both ways, and the only way to respect it is to allow people to say things that you might not agree with. You can disagree and try to rebut their arguments as much as you like because of those same "rules", but all censoring them does is demonstrate that you have actually no respect for the right to free speech you keep bleating about and don't deserve any respect in return.

It's similar for a lot of the "inclusive" ideologies too for that matter. You can't be truly inclusive unless you also accept the views of those who hate you for what you are or do, no matter how reprehensible they might be.

Cute, but I think reality would like a few words:

Planning permission. Especially so in Europe, maybe not such a problem with Trump, apart from the fact he *hates* anything "green".
Geology. Not all ground is suitable for building a heavy structure like a dam, or retaining the water it would hold back, reducing the number of glaciers this could apply to somewhat, and a proper survey can take a lot of time. You definitely do not want to build a dam on unsuitable ground.
Geography. You'd need to be able to get construction materials to the dam, and multiple glaciers may run off into the same valley, reducing the number of potential locations for hydro even further. Also, in the kinds of places where you find a lot of glaciers (mountainous regions, duh!), the valleys tend to be heavily used for things like habitation and agriculture, which would need relocating first.
The clock is ticking. Once the glaciers are gone, your dams are only going to be dealing with runoff from precipitation. Which isn't what it used to be where the glaciers are or the glaciers wouldn't be shrinking in the first place.

So, even with a viable location, you've got to get planning permission, relocate anything in the way, build the dam & turbines, connect it to the grid, and generate as much electricity as you can before... Oh, wait, wasn't there a glacier up there when we started?