Interesting. I hadn't considered pronoun reference(since, for humans, it seems to come naturally); but now that you mention it I can see how it would be a fairly brutal mess to try to codify.
(Purely as an aside; I applaud your choice of a rather fascinating, and magnet-obsessed, Jesuit polymath as a pseudonym. His theories may not have aged well; but he is a very, very, interesting guy.)
Indeed, I suspect that (as with low end routers) there are substantially fewer distinct designs than there are brand names and rebadges, which would make 3rd party firmware easier. On the minus side, in areas where rebadging is the rule it can be a real pain to ensure that you get the same hardware reliably: if your vendor is slapping their badge on one ODM's cheapo board today, they could(and not infrequently do) switch to slapping the same brand and model name on an entirely different board with approximately similar capabilities tomorrow.
This is hardly unique to IP cameras and DVRs, the OpenWRT hardware support wiki is loaded with examples of routers that sell under the same model name and number but are totally different internally(as well as ones that are sold by completely different companies, and internally identical) and USB peripherals, the nastier PCI/PCIe cards; and even computers that aren't associated with 'business' brands that promise image stability will sometimes swap chips without notice.
I'm not sure if it's a specific business decision, or some sort of culture/language thing; but these sorts of situations always struck me as an opportunity for some entrepreneurial type in China to simultaneously distinguish their product(albeit for a limited market) and get some software development and localization done more or less for free: Western FOSS tinkerers love cheap hardware to play with; and while some established vendors play fairly nice, the combination of 'IP' enthusiasm and a desire to tie hardware to various cloud services and app stores often limits how cooperative establshed western brands are with what the FOSS people want(eg. Intel recognizes the value of having non-awful, in-kernel, drivers for their NICs and chipsets and stuff, since Linux is serious business in the server market; but takes a "your motherboard comes with cryptographically signed UEFI, and you'll like it." attitude). If you have the necessary contacts and business relationships with hardware manufacturers, access to datasheets, etc. you could position yourself above the other rebadge outfits by assuring that your product has a known, stable, chipset and hardware design inside; and by being as helpful as possible to OpenWRT or an analogous effort; and both reap extra hardware sales from tinkerers who want to be sure that they are getting hardware with good 3rd party firmware support; and have the option of basing your official firmware on the 3rd party work; rather than the in-house atrocities that so often ruin otherwise decent hardware.
I don't doubt that it is harder than it looks; and my Mandarin isn't remotely good enough to try; but if I had hardware that offers excellent value, ruined by firmware that is utter crap, it seems like this could be a win-win.
I strongly suspect that such a product wouldn't go out through the same channels as the product that Samsung is actually proud of; but I have encountered various 'de-branded' hardware items where the item didn't pass muster with its actual vendor for whatever reason(maybe a refurb, maybe a product line that got killed, I don't know what goes on behind the scenes, though I'd be quite interested to) and it ends up having some anonymous packaging and an inferior warranty slapped on it, with a correspondingly lower price tag.
At least in the case of HP, the practice is common enough that the de-brand has its own 'brand': the circle/globe symbol you see on the bottom left of the front panel of this computer(chose the first example I could find, not an affiliate link, no specific endorsement implied).
I wouldn't expect the result of this reworking to be branded as a 'Galaxy Note' anything; quite possibly have all mention of Samsung scrubbed, bootsplash replaced, etc. but the smartphone market is big and price sensitive(especially 'emerging markets') and Samsung is going to be sitting on hundreds of thousands of perfectly good, high end, logic boards and screens. I wouldn't be at all surprised if the deal requires that whoever ends up reselling them make absolutely no mention of Samsung; but a high end touchscreen and top of the line Qualcomm SoC, assembled and ready to go; are worth so much more as the basis for a rework than as scrap that they'd have to be really, really, touchy to just send them to the grinder.
Or just rework them into slightly less sleek phones. Unless Samsung somehow managed to build phones that curse and doom any Li-ion cell connected to them; they'll be fine if they accept a somewhat thicker phone that allows them to use a slightly less power dense, but better tested, battery and offer it a bit more protection. It'd be uglier, and they'd have to design a new backplate and pay for the labor to rework them; but the battery isn't deeply buried; and the hardware is otherwise top of the line, so even an uglified version would probably beat up on competing midrange models; while still selling for enough money to be worth more than scrapping the thing.
I'd be curious to know what will actually happen in this case.
Clearly Samsung can't sell these as-is anymore; but by all reports the problem is with the battery(or possibly with the case not protecting the battery properly); not the logic board or the screen. Given that, it seems crazy to be talking about recycling them(even if we had nearly perfect methods), when the most expensive components are still fully functional.
I imagine that market-cannibalization/brand dilution/etc. concerns might interfere; but Samsung(or a 3rd party, if Samsung wants the result debranded and not associated with them); could pretty much just rip the back cover off; swap in a somewhat bigger and uglier, but non-explosive, battery; put a suitably enlarged cover back on and have the thing ready to go. If they didn't mess with the RF section, they might even be able to reuse FFC and similar certification.
Yes, the resulting product would be less valuable than the Note 7(if it could still be sold); but it would be worth a great deal more than even a perfect recovery of the constituent elements; and by the look of the teardown you could rework to remove the offending battery without damaging the PCBs or the screen with relatively little labor.
Even if perfect recycling existed, why would you grind up something like this? There's a Snapdragon 820, 4GB of RAM, 64GB of flash, and a 2560x1440 display in there; all perfectly fine. Surely CAD-ing up a new backplate and swapping the battery to produce a saleable phone is markedly more profitable than just breaking it down?
Nothing specific; though some IP cameras are incidentally supported because they are built on the same SoCs as routers.
Just by way of example, since one is on my desk, the D-Link DCS-930L is essentially a Ralink RT5350F with a lousy webcam attached to its USB host port; all integrated into a single PCB. Since the RT5350 shows up in all kinds of little routers, it has OpenWRT support; and since it is primarily a router SoC, the camera is a USB device rather than some MIPI CSI atrocity.
More generally, it just varies. A lot of the higher end DVRs are just x86s, since that's a cheap and easy way to get a punchy CPU, as much storage as you deem necessary; and optionally a bunch of PCI/PCIe capture cards to handle legacy analog devices; so putting your own OS on them isn't a terribly heroic endeavor(though support for the capture cards might be, what little support their is is typically aimed either at consumer entertainment devices or scientific/industrial framegrabbers, since the former has the biggest userbase and the latter has the deeper pockets). The cheap seats tend to be some ARM or MIPS SoC running a truly shoddy linux port(and have fun getting GPL compliance out of the vendor, not that you'd want see their kernel 2.4 hackjob anyway...); and so could be supported; but are likely to be a somewhat heroic undertaking unless enough interested people have the same hardware to work on it together.
Why is it that, even after a a couple of decades of experience and examples, a company would just ignore security researchers?
If you are a real asshole; and think you can get away with it, I can see why you might try to threaten them into silence; but if you can't do that, or you aren't scum, they are doing your work for you. What do you gain by not taking advantage of that?
Especially in this case. By the look of their product lineup, these guys appear to have aspirations higher than the '8 lousy cameras bundled with a POS DVR and sold under a mystery brand' caliber of 'security' product(they allege some institutional deployments; and they offer some SAS-attached drive shelves for their bigger DVRs); if you are trying to play with the professionals, isn't free security advice a bonus?
Unless you have the tech for full 'augmented reality'; and some good ideas about how to actually make that a virtue, it is hard to make a terribly compelling case for 'mobile VR'. The fact that a modern smartphone screen is just about the right size to be shoehorned into a low rent VR headset is worth a few tech demos; but nothing battery powered that fits on your face currently has the punch for VR work; and 'wearing giant, ridiculous-looking blinders' is a bad idea in public.
The main win for 'mobile' in general isn't on absolute quality; but on the fact that it is in your pocket right now and other sources of distraction aren't. Especially for cellphone stuff, which doesn't have the advantage of hardware buttons designed with games in mind; but offers a very, very, low friction path to downloading and playing something.
So long as VR gear is moderately ridiculous looking; and largely blocks the surrounding reality, it's not exactly a compelling choice for on-the-go entertainment, which generally demands something reasonably unobtrusive and capable of being used without missing your station/walking into things/etc.
VR more generally has some definite use cases(which, in part, is why deep-pocketed research types have been enduring considerably lousier and vastly more expensive VR setups for a couple of decades now); but are going to have trouble escaping 'novelty' unless the install base is larger.
Whenever you have a feature that is cool; but only some of your players have, you force developers to choose between drastically narrowing their customer base, ignoring the cool feature entirely, or doing something with the cool feature that is sufficiently unimportant that the game can still 'fail gracefully' for people who don't have access to it.
We saw a similar thing, though less dramatic, with 'PhysX': when they first came out with their dedicated PPU card, approximately nobody owned one, so any games that could use the additional physics processing used it for visual tinsel that could be removed or faked without causing any real problems in gameplay. Even after Nvidia ate them developers couldn't necessarily rely on particularly high performance physics acceleration being available(yes on higher end Nvidia setups; but limited on feebler Nvidia GPUs and CPU backed on Intel and AMD setups), so the effects remained mere flavor. Often rather pretty flavor; but nothing gameplay essential; because it still has to work if the physics acceleration isn't available.
If I could find any tapes, it'd take about 45 seconds to make the tape player lurking in my basement tell you that it is sentient, has a soul, and aspires to understand the meaning of life.
That's the trivial bit. Not sounding like a combination of naive keyword searches and cliches aimed at being vaguely suitable to the broadest possible set of situations? Less trivial.
GX270/280-era 'capacitor plague'. I forget the exciting story of industrial espionage and vendor shoddiness; but for some reason a lot of substandard electrolytic capacitors made it into the supply chain. They had a tendency to swell, leak; or just derate far faster than expected. When the capacitors are supposed to be part of the circuit that supplies the CPU with appropriately regulated power, this does wonders for reliability.
It wasn't exclusive to dell, pretty much all desktop motherboards of the period used electrolytics, and the flawed capacitors were widespread; but they had a massive number of affected units and did their best to be total scum about honoring warranties, so they came off looking pretty bad.
So, even if we make the generous assumption that lots and lots of aerial photos are a useful tool, rather than some combination of vendor snake oil and lazy technophiles looking for any excuse to sit in some sort of 'command center' with a comfy chair and some giant monitors instead of having to go outside and do boring police stuff; how is secrecy a good plan?
Solving crimes is nice; but what people really like is when your 'deterrents' cause them to just not happen in the first place. You might be able to justify some concealment of the fine details in order to frustrate attempts to circumvent the measure; but keeping the existence of the entire program secret massively reduces its potential as a deterrent, which is effectively choosing to have more crime in the hopes of closing more cases rather than increasing the perceived risk of engaging in crime.
There are, of course, other reasons for secrecy; but they aren't very flattering.
That is a fair case: the fact that you are buying two CPUs, and Intel's "Because nobody else makes x86s this fast, now do they?" pricing on high end Xeons; plus the slow (compared to desktop and laptop SKUs) adoption of new core designs makes just buying the fast ones up front seriously painful; and buying used actually worth the swap.
As for service life, I've never (knowingly) seen a CPU die, except when my deeply unwise overclocking and overvolting killed it(and I've done enough tech work that my sample size, while haphazard and informal, is at least a few thousand units of ages varying from 'brand new' to 15+ years). It's possible that I have, out of a belief in CPU reliability, misdiagnosed a CPU issue as a motherboard/memory/etc. issue; but they really do seem to be impressively durable.
It's actually somewhat curious: I can understand why solid-state parts would be more reliable than lousy electrolytic capacitors or HDDs with delicate moving parts; but CPUs(despite having some of the most alarming power densities of any logic silicon in the computer) still seem more rugged than GPUs and a lot more reliable than RAM.
They don't necessarily offer the best value; since bundling leaves room to charge for 'convenience'; but you can get arrangements that are essentially what you describe.
Something like this Kingston one gives you a USB enclosure, a copy of some version of Acronis for the data transfer, and the SSD. It's also not uncommon to see packages(especially when aimed at desktop upgrades, since 3.5in HDD USB enclosures are more expensive, being larger and needing a wall wart) where they provide just the imaging software and expect you to plug both drives into internal SATA ports and power.
You should probably consider how much of what we just take for granted counts as 'automating your work'.
Yes, if you can't function without a point-n'-drool EZ Wizard; you are probably of limited use and facing limited future prospects; but do we seriously expect all those lazy, incomptent, losers who rely on a "file system" to provide a convenient abstraction when a real man could just access the block device directly; or the pansies who use 'compilers' because they suck too much to produce machine code themselves to be heading for the bread lines?
Outside of some(often well compensated, since they are genuinely difficult) specialty situations that bring you very, very, close to the metal(or make you responsible for designing 'the metal') almost all software development is done on top of a fairly gigantic pile of tools that automate assorted ugly details; it's not something exclusive to ITT tech java monkeys.
While the overclocking and impecunious-PC-performance-chasing of my misspent youth makes me sad that this is the case; the "CPU upgrade = buy new computer" mentality isn't really all that irrational.
With laptops; it is effectively mandatory. Even in laptops with socketed CPUs, unless you went out of your way to buy the absolute worst version of a laptop with fairly high end options, you'll find that the fastest socket and TDP compatible CPU upgrade just isn't all that much faster. Plus, if it's a reasonably new laptop, buying the CPU that is a worthwhile upgrade will be pretty expensive; and if it is an old one it'll be cheap; but leave you with a laptop that is showing its age in both specs and wear and tear.
With desktops; you are also likely to have limited socket-compatible upgrade options, so getting a meaningful CPU boost often means swapping the motherboard as well(unless you started with the lousiest option for a given socket, in which case there might be meaningful improvements to be had); and if you hit the DDR2 to DDR3 or the DDR3 to DDR4 transition you'll need new RAM as well. PSU can probably be reused, unless it is particularly grim; and expansion cards, HDDs, optical drives, and case can be reused; but bumping the CPU speed in any serious way tends to mean ripping out most of the expensive parts(unless your GPU is fancy enough to count as the really expensive part of the system).
An SSD, by contrast, is an easy swap except on laptops that really hate you; and even on ancient systems limited to 1.5Gb/s SATA, the improvement in latency and IOPs over a mechanical drive is pretty dramatic; plus compatibility is almost universal unless your system is so old that you still have PATA; or you want to boot from an NVMe device.
It would be interesting to know if the somewhat narrow 'window' of feasibility helped encourage them to stop fighting.
Actual 'Arduino' products sell at a considerable premium compared to just the relevant Atmel microcontroller with bootloader(for the minimalists) or the assorted mostly adequate clones(especially if you don't need 'shield' compatibility, the standard anonymous Chinese ebay special gets you a USB connector and handles the fiddly soldering for barely more than the price of the components); which is apparently a premium that enough people are willing to pay, either out of a desire to support the project or because the absolute cost of just buying all the official components so that things will Just Work is pretty small; but not really room enough for two confusingly similar and mutually antagonistic vendors.
Aside from merely dividing the available marketshare, having the ongoing feud spoils both the 'support the project' and the 'make life simpler by just buying everything from the official vendor' incentives; at which point you have to be pretty motivated to try to figure out who the 'real' Arduino is; rather than either just buying one of the upmarket clones from Sparkfun or Adafruit or the like(if you are relatively cost-insensitive and don't want to deal with any uncertainties) or grabbing a pack of ten of whatever washes up on ebay(if price is the main factor).
I really have no idea who was in the right on this particular feud; but that is likely part of the problem: if there is a clear winner and a clear loser(as with Xorg vs. xfree86) that is one thing; but if not, the people who are willing to pay more for a helpful and sane vendor have good reason to flee to one of the higher end clones, and anyone who wants to 'support the project' has to wade through an ugly and confusing mess to even figure out which purchase that implies.
Gold is undeniably a compelling leader in the "Hey, do you need an handy abstract representation of value?" market.
It is effectively impossible to counterfeit(all the metals that look kind of golden aren't nearly dense enough; Tungsten and DU have the density about right but are wrong in basically all other respects, nuclear synthesis isn't really counterfeiting but is uneconomic, it's tricky to alloy with something cheaper without being caught by even fairly primitive measurement of volume and weight; etc.), it's pretty scarce, it can be divided/combined/melted down/reshaped easily(unlike precious stones, say, where the value of two halves of a diamond is markedly lower than the value of the larger stone), people find it appealing, and so on.
The problem is just knowing what situations do, or don't, reward possessing a handy abstract representation of value. Too little civilization and you either can't find anyone willing to sell you stuff; or run into somebody who knows that the exchange rate between gold and iron is actually pretty favorable when the iron is of the right shape to stab the guy with the gold. Too much civilization and the fact that it's an inert, unproductive, comparatively cumbersome to transport/store/transact with lump of deadweight makes it a pain compared to whatever currency is being reasonably well managed at the time.
It's only in the intermediate situations, where you are developing a real market; but don't have anyone competent enough to produce worthwhile currency; or have a real market but a previously stable currency is on the rocks; where it really shines. Outside of that, it's just jewelry, anticorrosion coating, or a specific commodities position that might be useful under certain specific conditions as part of a larger portfolio.
Even gold depends on the shared belief that there will be somebody else willing to accept it in exchange for goods of actual use within a survivable period of time after whatever crisis you are expecting passes. Certainly more durable than a few electronic IOUs or fiat currency issued by a nation state that is now on fire/crawling with zombies/etc; but the intrinsic utility is pretty limited. If the apocalypse needs corrosion-resistant connectors, gold has you covered; you could substitute it for lead in ballistic applications; but that's pretty much the list.
With the exception of people expecting to deal with explosions(where bunkers are a natural fit; and fairly commonly used in varying degrees of sophistication); a lot of this disaster-prep stuff falls into an unhelpful category of being both overprepared and underprepared: If you are concerned, it's pretty easy to justify enough supplies to weather a breakdown in our efficient-but-tightly-stretched supply chains; but you don't usually need a bunker to do that. If you have a crisis more serious than not being able to buy groceries for a few months in mind, however, the problem stops being "Do I have enough MREs?" and turns into "Am I set for subsistence farming and/or tribal warfare; and do I really want to bother with that shit anyway?" unpleasantly quickly.
It all seems aimed at a (not impossible; but not necessarily plausible) medium-size disaster; which will somehow be big enough that the 'stash of supplies in the basement' crowd is doomed; but small enough that your bunker isn't going to be plundered by local militias and there will be a society worth living in waiting for you when it's time to open the door again.
While I fully agree that (while we sometimes do our best to pretend otherwise) US constitutional protections of speech are vastly better than the world's in general, even the parts not commonly considered to be despotic hellholes; and I think that letting Team Morality into the TLD business will be a clusterfuck; it isn't clear that ICANN being under US contract ensures that US standards prevail on the internet; or that the aforementioned clusterfuck is avoided.
Since ICANN has no actual teeth, aside from the inconveniences of being incompatible, it is already routine for entities of various sizes; from home networks with cheapie routers whose embedded DNS servers give themselves a name friendlier than 192.168.0.1; to companies trying to keep employees off facebook at work; to the Great Firewall of China and the 'Halal Internet' of Iran; to implement a 'mostly what ICANN says; except when we disagree' assignment of internet names and numbers; and when you have legal authority(and sometimes direct ownership or control) of the ISPs, that counts for a lot.
On the Team Morality and the TLDs side; the Department of Commerce couldn't, or didn't feel like, keeping ICANN from treading the road of total insanity of allowing gTLDs to proliferate like mad. That terrible plan already has the trademark lawyers, the morality police, every last idiot with sensitive feelings about anything, and so on up in arms and likely to keep squabbling until the heat death of the universe.
If it were actually the case that ICANN is a good vehicle for advancing America's better nature worldwide; and that maintaining American operation of ICANN was instrumental to keeping it that way; I'd be 100% against letting anyone else touch it. I'm just not convinced that either of those are particularly true. It is easy, and people already do wherever they have the requisite power, to operate "ICANN, minus what I dislike" and have it stick against all but moderately determined users. Some countries do it much harder than others; with the US most enthusiastic when copyright is involved; but ICANN has zero effect on how easy it is.
I know that "That depends" is the second most frustrating answer(after "yes and no"); but it is true here.
Across what geographic area, set of topics, etc. are these minutely articles distributed?
If you consider a global scale, and a fairly wide variety of interests(not necessarily serious niche stuff; but all the sections that a major Sunday print newspaper traditionally had); one article a minute is downright patchy coverage.
If you are talking a local news outlet; or a "just the foreign events large enough to be relevant" offering; it strongly suggests that they are really, really playing hard with the 'minimum publishable unit' concept.
If the once-a-minute number is across a whole stable of publications catering to different interests; then it might be the case that once you remove the celebrity gossip they actually only publish every ten minutes; Given how few genuinely just-a-local-paper operations exist these days, the quoted publication rate is probably across a media empire that isn't expected to appeal to any single individual: it'll probably have local news for more places than any one person could live/work; cultural tidbits across more fields any one person cares about; politics from around the world, and so on.
I'd argue that there are really two better questions: Ignore the stated total output; and ask "How much are they publishing that I find worth reading?" and "Is their focus on speed killing their ability to focus?" The first question is obvious: you don't enjoy news by the pound or by the word; you enjoy news by how much you actually feel like reading. The second is slightly trickier: Mere 'data' are pretty easy to come by. The sorts of news reports that you get when you give an experienced reporter plenty of time and room to dig into a matter he is experienced with are much less so. If an outfit's metrics-driven chase after viral listicles has caused them to cancel all reporting that can't be reworded from AP feeds by interns within 20 minutes; they've hollowed themselves out and it barely matters how fast they churn out "content" because none of it will add up to anything. If they just generate a lot of material because they have a lot of people reporting; that's a different matter.
Isn't that a pretty easy one? Unless you adhere to a reading of the constitution that allows for virtually no federal government activity at all(in which case ARPA probably shouldn't have ever had the cash to spend on the project; and the Department of Commerce either shouldn't exist or should be a tiny fraction of the size and scope); the US government clearly has the authority to spend allocate DoD funds to an R&D project deemed to be of military interest; to hire somebody to handle the technical work bundled under the Internet Assigned Numbers Authority; and to transfer the contract for the same functions over to the Department of Commerce once it became clear that civilian and commercial applications of the technology were where the action is.
That doesn't mean that the US has any right to get other people to care what its DNS servers say; what media types it defines, etc; but it takes a pretty narrow reading of their powers to suggest that they don't have the authority to set up a body to publish that sort of thing in the hopes that others will adopt it because being compatible is more valuable than getting to DIY every aspect of the system.
So far as I know, nobody has ever claimed US authority over 'DNS'(indeed; back in the heady days of the.com bubble, companies trying to get users to point to their nameservers so that they could sell shitty vanity domains were a dime a dozen; and nobody even argued that US nationals had any duty to abide by ICANN-defined names and numbers; it's just that the market value of DNS servers that live in a strange world of their own turned out to be pretty limited). ICANN's authority, to the degree it has any, is founded in the fact that it's a pain in the ass to administer and maintain systems that have drifted out of compatibility with what the majority is using.
Even today, and for years now, DNS servers and other infrastructure routinely flout ICANN in situations where the benefits are greater than the costs(oddball hostnames on LANs; lazy content blocking by providing bogus IPs for sites you don't want users getting to, just choosing your own damn port because you feel like running your protocol on it, etc.) They pay more attention in places where incompatibility would hurt more: competing claims on various TLDs would get to be quite a mess; your life would really suck if your pet flavor of IP starts to differ enough that you need custom routing hardware, that sort of thing.
Nobody needs ICANN's blessing to just ignore them; but it's pretty easy to justify the Department of Commerce paying some people to be DNS jockeys.
Even if one considers the ICANN handoff to be a terrible plan; that still leaves the "and a state would have standing to block this why exactly?" problem unsolved.
I'm having trouble thinking of a reading of the constitution where one of the several states gets to stop the feds from making a change in management to a Department of Commerce contractor(handling a job previously done by a DoD contractor) overseeing the outgrowth of a federal military research project.
It'd be like Vermont going to court because they think that selling F-15s to the Saudis is a terrible plan. They may well be entirely correct; but even pretty minimalist readings of the constitution tend to give the feds most of the foreign policy power.
I'm deeply unclear on what the world thinks they'll get from ICANN that they haven't under US administration; and also unclear on what we have to gain from changing the situation; but I'm still baffled as to what possible standing state governments have on the issue.
Privacy policy. The UK, among certain others, substantially limits the degree of anonymity a donor is allowed to preserve(with respect to the recipients and the children produced; the sperm bank always wants to know a fair bit about the donor); and the trend has been toward even further emphasis on the 'right' of the child to know about their parentage; which makes even the people who are OK with the current level of information jumpy about what might happen in the future and be retroactively imposed on them.
It's not terribly hard to pay people enough to masturbate into a cup, even after you impose the various restrictions that the state of the tech(sperm counts, motility) and the state of the demand(height, educational attainment, absence of unpleasant genetic conditions) create. It is...less simple... to pay people enough to deal with the possibility that at least one mystery child might end up knocking on the door and calling them daddy in a decade or two.
Slashdot Mirror
Interesting. I hadn't considered pronoun reference(since, for humans, it seems to come naturally); but now that you mention it I can see how it would be a fairly brutal mess to try to codify.
(Purely as an aside; I applaud your choice of a rather fascinating, and magnet-obsessed, Jesuit polymath as a pseudonym. His theories may not have aged well; but he is a very, very, interesting guy.)
Indeed, I suspect that (as with low end routers) there are substantially fewer distinct designs than there are brand names and rebadges, which would make 3rd party firmware easier. On the minus side, in areas where rebadging is the rule it can be a real pain to ensure that you get the same hardware reliably: if your vendor is slapping their badge on one ODM's cheapo board today, they could(and not infrequently do) switch to slapping the same brand and model name on an entirely different board with approximately similar capabilities tomorrow.
This is hardly unique to IP cameras and DVRs, the OpenWRT hardware support wiki is loaded with examples of routers that sell under the same model name and number but are totally different internally(as well as ones that are sold by completely different companies, and internally identical) and USB peripherals, the nastier PCI/PCIe cards; and even computers that aren't associated with 'business' brands that promise image stability will sometimes swap chips without notice.
I'm not sure if it's a specific business decision, or some sort of culture/language thing; but these sorts of situations always struck me as an opportunity for some entrepreneurial type in China to simultaneously distinguish their product(albeit for a limited market) and get some software development and localization done more or less for free: Western FOSS tinkerers love cheap hardware to play with; and while some established vendors play fairly nice, the combination of 'IP' enthusiasm and a desire to tie hardware to various cloud services and app stores often limits how cooperative establshed western brands are with what the FOSS people want(eg. Intel recognizes the value of having non-awful, in-kernel, drivers for their NICs and chipsets and stuff, since Linux is serious business in the server market; but takes a "your motherboard comes with cryptographically signed UEFI, and you'll like it." attitude). If you have the necessary contacts and business relationships with hardware manufacturers, access to datasheets, etc. you could position yourself above the other rebadge outfits by assuring that your product has a known, stable, chipset and hardware design inside; and by being as helpful as possible to OpenWRT or an analogous effort; and both reap extra hardware sales from tinkerers who want to be sure that they are getting hardware with good 3rd party firmware support; and have the option of basing your official firmware on the 3rd party work; rather than the in-house atrocities that so often ruin otherwise decent hardware.
I don't doubt that it is harder than it looks; and my Mandarin isn't remotely good enough to try; but if I had hardware that offers excellent value, ruined by firmware that is utter crap, it seems like this could be a win-win.
I can only hope(though doubt) that the person responsible for this 'innovation' is appropriately ashamed of who they are and what they have done.
There are plenty of people who are useless; but this winner is actively making the world just a little bit worse. I hope that weighs on them.
I strongly suspect that such a product wouldn't go out through the same channels as the product that Samsung is actually proud of; but I have encountered various 'de-branded' hardware items where the item didn't pass muster with its actual vendor for whatever reason(maybe a refurb, maybe a product line that got killed, I don't know what goes on behind the scenes, though I'd be quite interested to) and it ends up having some anonymous packaging and an inferior warranty slapped on it, with a correspondingly lower price tag.
At least in the case of HP, the practice is common enough that the de-brand has its own 'brand': the circle/globe symbol you see on the bottom left of the front panel of this computer(chose the first example I could find, not an affiliate link, no specific endorsement implied).
I wouldn't expect the result of this reworking to be branded as a 'Galaxy Note' anything; quite possibly have all mention of Samsung scrubbed, bootsplash replaced, etc. but the smartphone market is big and price sensitive(especially 'emerging markets') and Samsung is going to be sitting on hundreds of thousands of perfectly good, high end, logic boards and screens. I wouldn't be at all surprised if the deal requires that whoever ends up reselling them make absolutely no mention of Samsung; but a high end touchscreen and top of the line Qualcomm SoC, assembled and ready to go; are worth so much more as the basis for a rework than as scrap that they'd have to be really, really, touchy to just send them to the grinder.
Or just rework them into slightly less sleek phones. Unless Samsung somehow managed to build phones that curse and doom any Li-ion cell connected to them; they'll be fine if they accept a somewhat thicker phone that allows them to use a slightly less power dense, but better tested, battery and offer it a bit more protection. It'd be uglier, and they'd have to design a new backplate and pay for the labor to rework them; but the battery isn't deeply buried; and the hardware is otherwise top of the line, so even an uglified version would probably beat up on competing midrange models; while still selling for enough money to be worth more than scrapping the thing.
I'd be curious to know what will actually happen in this case.
Clearly Samsung can't sell these as-is anymore; but by all reports the problem is with the battery(or possibly with the case not protecting the battery properly); not the logic board or the screen. Given that, it seems crazy to be talking about recycling them(even if we had nearly perfect methods), when the most expensive components are still fully functional.
I imagine that market-cannibalization/brand dilution/etc. concerns might interfere; but Samsung(or a 3rd party, if Samsung wants the result debranded and not associated with them); could pretty much just rip the back cover off; swap in a somewhat bigger and uglier, but non-explosive, battery; put a suitably enlarged cover back on and have the thing ready to go. If they didn't mess with the RF section, they might even be able to reuse FFC and similar certification.
Yes, the resulting product would be less valuable than the Note 7(if it could still be sold); but it would be worth a great deal more than even a perfect recovery of the constituent elements; and by the look of the teardown you could rework to remove the offending battery without damaging the PCBs or the screen with relatively little labor.
Even if perfect recycling existed, why would you grind up something like this? There's a Snapdragon 820, 4GB of RAM, 64GB of flash, and a 2560x1440 display in there; all perfectly fine. Surely CAD-ing up a new backplate and swapping the battery to produce a saleable phone is markedly more profitable than just breaking it down?
Nothing specific; though some IP cameras are incidentally supported because they are built on the same SoCs as routers.
Just by way of example, since one is on my desk, the D-Link DCS-930L is essentially a Ralink RT5350F with a lousy webcam attached to its USB host port; all integrated into a single PCB. Since the RT5350 shows up in all kinds of little routers, it has OpenWRT support; and since it is primarily a router SoC, the camera is a USB device rather than some MIPI CSI atrocity.
More generally, it just varies. A lot of the higher end DVRs are just x86s, since that's a cheap and easy way to get a punchy CPU, as much storage as you deem necessary; and optionally a bunch of PCI/PCIe capture cards to handle legacy analog devices; so putting your own OS on them isn't a terribly heroic endeavor(though support for the capture cards might be, what little support their is is typically aimed either at consumer entertainment devices or scientific/industrial framegrabbers, since the former has the biggest userbase and the latter has the deeper pockets). The cheap seats tend to be some ARM or MIPS SoC running a truly shoddy linux port(and have fun getting GPL compliance out of the vendor, not that you'd want see their kernel 2.4 hackjob anyway...); and so could be supported; but are likely to be a somewhat heroic undertaking unless enough interested people have the same hardware to work on it together.
Why is it that, even after a a couple of decades of experience and examples, a company would just ignore security researchers?
If you are a real asshole; and think you can get away with it, I can see why you might try to threaten them into silence; but if you can't do that, or you aren't scum, they are doing your work for you. What do you gain by not taking advantage of that?
Especially in this case. By the look of their product lineup, these guys appear to have aspirations higher than the '8 lousy cameras bundled with a POS DVR and sold under a mystery brand' caliber of 'security' product(they allege some institutional deployments; and they offer some SAS-attached drive shelves for their bigger DVRs); if you are trying to play with the professionals, isn't free security advice a bonus?
"They are said to place work central in their lives, to avoid wasting time and to be ethical in their dealings with others."
Because if there is one word that comes to mind when one examines the period of history the baby boomers wrought it would definitely be 'ethical'...
Unless you have the tech for full 'augmented reality'; and some good ideas about how to actually make that a virtue, it is hard to make a terribly compelling case for 'mobile VR'. The fact that a modern smartphone screen is just about the right size to be shoehorned into a low rent VR headset is worth a few tech demos; but nothing battery powered that fits on your face currently has the punch for VR work; and 'wearing giant, ridiculous-looking blinders' is a bad idea in public.
The main win for 'mobile' in general isn't on absolute quality; but on the fact that it is in your pocket right now and other sources of distraction aren't. Especially for cellphone stuff, which doesn't have the advantage of hardware buttons designed with games in mind; but offers a very, very, low friction path to downloading and playing something.
So long as VR gear is moderately ridiculous looking; and largely blocks the surrounding reality, it's not exactly a compelling choice for on-the-go entertainment, which generally demands something reasonably unobtrusive and capable of being used without missing your station/walking into things/etc.
VR more generally has some definite use cases(which, in part, is why deep-pocketed research types have been enduring considerably lousier and vastly more expensive VR setups for a couple of decades now); but are going to have trouble escaping 'novelty' unless the install base is larger.
Whenever you have a feature that is cool; but only some of your players have, you force developers to choose between drastically narrowing their customer base, ignoring the cool feature entirely, or doing something with the cool feature that is sufficiently unimportant that the game can still 'fail gracefully' for people who don't have access to it.
We saw a similar thing, though less dramatic, with 'PhysX': when they first came out with their dedicated PPU card, approximately nobody owned one, so any games that could use the additional physics processing used it for visual tinsel that could be removed or faked without causing any real problems in gameplay. Even after Nvidia ate them developers couldn't necessarily rely on particularly high performance physics acceleration being available(yes on higher end Nvidia setups; but limited on feebler Nvidia GPUs and CPU backed on Intel and AMD setups), so the effects remained mere flavor. Often rather pretty flavor; but nothing gameplay essential; because it still has to work if the physics acceleration isn't available.
If I could find any tapes, it'd take about 45 seconds to make the tape player lurking in my basement tell you that it is sentient, has a soul, and aspires to understand the meaning of life.
That's the trivial bit. Not sounding like a combination of naive keyword searches and cliches aimed at being vaguely suitable to the broadest possible set of situations? Less trivial.
GX270/280-era 'capacitor plague'. I forget the exciting story of industrial espionage and vendor shoddiness; but for some reason a lot of substandard electrolytic capacitors made it into the supply chain. They had a tendency to swell, leak; or just derate far faster than expected. When the capacitors are supposed to be part of the circuit that supplies the CPU with appropriately regulated power, this does wonders for reliability.
It wasn't exclusive to dell, pretty much all desktop motherboards of the period used electrolytics, and the flawed capacitors were widespread; but they had a massive number of affected units and did their best to be total scum about honoring warranties, so they came off looking pretty bad.
So, even if we make the generous assumption that lots and lots of aerial photos are a useful tool, rather than some combination of vendor snake oil and lazy technophiles looking for any excuse to sit in some sort of 'command center' with a comfy chair and some giant monitors instead of having to go outside and do boring police stuff; how is secrecy a good plan?
Solving crimes is nice; but what people really like is when your 'deterrents' cause them to just not happen in the first place. You might be able to justify some concealment of the fine details in order to frustrate attempts to circumvent the measure; but keeping the existence of the entire program secret massively reduces its potential as a deterrent, which is effectively choosing to have more crime in the hopes of closing more cases rather than increasing the perceived risk of engaging in crime.
There are, of course, other reasons for secrecy; but they aren't very flattering.
That is a fair case: the fact that you are buying two CPUs, and Intel's "Because nobody else makes x86s this fast, now do they?" pricing on high end Xeons; plus the slow (compared to desktop and laptop SKUs) adoption of new core designs makes just buying the fast ones up front seriously painful; and buying used actually worth the swap.
As for service life, I've never (knowingly) seen a CPU die, except when my deeply unwise overclocking and overvolting killed it(and I've done enough tech work that my sample size, while haphazard and informal, is at least a few thousand units of ages varying from 'brand new' to 15+ years). It's possible that I have, out of a belief in CPU reliability, misdiagnosed a CPU issue as a motherboard/memory/etc. issue; but they really do seem to be impressively durable.
It's actually somewhat curious: I can understand why solid-state parts would be more reliable than lousy electrolytic capacitors or HDDs with delicate moving parts; but CPUs(despite having some of the most alarming power densities of any logic silicon in the computer) still seem more rugged than GPUs and a lot more reliable than RAM.
They don't necessarily offer the best value; since bundling leaves room to charge for 'convenience'; but you can get arrangements that are essentially what you describe.
Something like this Kingston one gives you a USB enclosure, a copy of some version of Acronis for the data transfer, and the SSD. It's also not uncommon to see packages(especially when aimed at desktop upgrades, since 3.5in HDD USB enclosures are more expensive, being larger and needing a wall wart) where they provide just the imaging software and expect you to plug both drives into internal SATA ports and power.
You should probably consider how much of what we just take for granted counts as 'automating your work'.
Yes, if you can't function without a point-n'-drool EZ Wizard; you are probably of limited use and facing limited future prospects; but do we seriously expect all those lazy, incomptent, losers who rely on a "file system" to provide a convenient abstraction when a real man could just access the block device directly; or the pansies who use 'compilers' because they suck too much to produce machine code themselves to be heading for the bread lines?
Outside of some(often well compensated, since they are genuinely difficult) specialty situations that bring you very, very, close to the metal(or make you responsible for designing 'the metal') almost all software development is done on top of a fairly gigantic pile of tools that automate assorted ugly details; it's not something exclusive to ITT tech java monkeys.
While the overclocking and impecunious-PC-performance-chasing of my misspent youth makes me sad that this is the case; the "CPU upgrade = buy new computer" mentality isn't really all that irrational.
With laptops; it is effectively mandatory. Even in laptops with socketed CPUs, unless you went out of your way to buy the absolute worst version of a laptop with fairly high end options, you'll find that the fastest socket and TDP compatible CPU upgrade just isn't all that much faster. Plus, if it's a reasonably new laptop, buying the CPU that is a worthwhile upgrade will be pretty expensive; and if it is an old one it'll be cheap; but leave you with a laptop that is showing its age in both specs and wear and tear.
With desktops; you are also likely to have limited socket-compatible upgrade options, so getting a meaningful CPU boost often means swapping the motherboard as well(unless you started with the lousiest option for a given socket, in which case there might be meaningful improvements to be had); and if you hit the DDR2 to DDR3 or the DDR3 to DDR4 transition you'll need new RAM as well. PSU can probably be reused, unless it is particularly grim; and expansion cards, HDDs, optical drives, and case can be reused; but bumping the CPU speed in any serious way tends to mean ripping out most of the expensive parts(unless your GPU is fancy enough to count as the really expensive part of the system).
An SSD, by contrast, is an easy swap except on laptops that really hate you; and even on ancient systems limited to 1.5Gb/s SATA, the improvement in latency and IOPs over a mechanical drive is pretty dramatic; plus compatibility is almost universal unless your system is so old that you still have PATA; or you want to boot from an NVMe device.
It would be interesting to know if the somewhat narrow 'window' of feasibility helped encourage them to stop fighting.
Actual 'Arduino' products sell at a considerable premium compared to just the relevant Atmel microcontroller with bootloader(for the minimalists) or the assorted mostly adequate clones(especially if you don't need 'shield' compatibility, the standard anonymous Chinese ebay special gets you a USB connector and handles the fiddly soldering for barely more than the price of the components); which is apparently a premium that enough people are willing to pay, either out of a desire to support the project or because the absolute cost of just buying all the official components so that things will Just Work is pretty small; but not really room enough for two confusingly similar and mutually antagonistic vendors.
Aside from merely dividing the available marketshare, having the ongoing feud spoils both the 'support the project' and the 'make life simpler by just buying everything from the official vendor' incentives; at which point you have to be pretty motivated to try to figure out who the 'real' Arduino is; rather than either just buying one of the upmarket clones from Sparkfun or Adafruit or the like(if you are relatively cost-insensitive and don't want to deal with any uncertainties) or grabbing a pack of ten of whatever washes up on ebay(if price is the main factor).
I really have no idea who was in the right on this particular feud; but that is likely part of the problem: if there is a clear winner and a clear loser(as with Xorg vs. xfree86) that is one thing; but if not, the people who are willing to pay more for a helpful and sane vendor have good reason to flee to one of the higher end clones, and anyone who wants to 'support the project' has to wade through an ugly and confusing mess to even figure out which purchase that implies.
Gold is undeniably a compelling leader in the "Hey, do you need an handy abstract representation of value?" market.
It is effectively impossible to counterfeit(all the metals that look kind of golden aren't nearly dense enough; Tungsten and DU have the density about right but are wrong in basically all other respects, nuclear synthesis isn't really counterfeiting but is uneconomic, it's tricky to alloy with something cheaper without being caught by even fairly primitive measurement of volume and weight; etc.), it's pretty scarce, it can be divided/combined/melted down/reshaped easily(unlike precious stones, say, where the value of two halves of a diamond is markedly lower than the value of the larger stone), people find it appealing, and so on.
The problem is just knowing what situations do, or don't, reward possessing a handy abstract representation of value. Too little civilization and you either can't find anyone willing to sell you stuff; or run into somebody who knows that the exchange rate between gold and iron is actually pretty favorable when the iron is of the right shape to stab the guy with the gold. Too much civilization and the fact that it's an inert, unproductive, comparatively cumbersome to transport/store/transact with lump of deadweight makes it a pain compared to whatever currency is being reasonably well managed at the time.
It's only in the intermediate situations, where you are developing a real market; but don't have anyone competent enough to produce worthwhile currency; or have a real market but a previously stable currency is on the rocks; where it really shines. Outside of that, it's just jewelry, anticorrosion coating, or a specific commodities position that might be useful under certain specific conditions as part of a larger portfolio.
Even gold depends on the shared belief that there will be somebody else willing to accept it in exchange for goods of actual use within a survivable period of time after whatever crisis you are expecting passes. Certainly more durable than a few electronic IOUs or fiat currency issued by a nation state that is now on fire/crawling with zombies/etc; but the intrinsic utility is pretty limited. If the apocalypse needs corrosion-resistant connectors, gold has you covered; you could substitute it for lead in ballistic applications; but that's pretty much the list.
With the exception of people expecting to deal with explosions(where bunkers are a natural fit; and fairly commonly used in varying degrees of sophistication); a lot of this disaster-prep stuff falls into an unhelpful category of being both overprepared and underprepared: If you are concerned, it's pretty easy to justify enough supplies to weather a breakdown in our efficient-but-tightly-stretched supply chains; but you don't usually need a bunker to do that. If you have a crisis more serious than not being able to buy groceries for a few months in mind, however, the problem stops being "Do I have enough MREs?" and turns into "Am I set for subsistence farming and/or tribal warfare; and do I really want to bother with that shit anyway?" unpleasantly quickly.
It all seems aimed at a (not impossible; but not necessarily plausible) medium-size disaster; which will somehow be big enough that the 'stash of supplies in the basement' crowd is doomed; but small enough that your bunker isn't going to be plundered by local militias and there will be a society worth living in waiting for you when it's time to open the door again.
While I fully agree that (while we sometimes do our best to pretend otherwise) US constitutional protections of speech are vastly better than the world's in general, even the parts not commonly considered to be despotic hellholes; and I think that letting Team Morality into the TLD business will be a clusterfuck; it isn't clear that ICANN being under US contract ensures that US standards prevail on the internet; or that the aforementioned clusterfuck is avoided.
Since ICANN has no actual teeth, aside from the inconveniences of being incompatible, it is already routine for entities of various sizes; from home networks with cheapie routers whose embedded DNS servers give themselves a name friendlier than 192.168.0.1; to companies trying to keep employees off facebook at work; to the Great Firewall of China and the 'Halal Internet' of Iran; to implement a 'mostly what ICANN says; except when we disagree' assignment of internet names and numbers; and when you have legal authority(and sometimes direct ownership or control) of the ISPs, that counts for a lot.
On the Team Morality and the TLDs side; the Department of Commerce couldn't, or didn't feel like, keeping ICANN from treading the road of total insanity of allowing gTLDs to proliferate like mad. That terrible plan already has the trademark lawyers, the morality police, every last idiot with sensitive feelings about anything, and so on up in arms and likely to keep squabbling until the heat death of the universe.
If it were actually the case that ICANN is a good vehicle for advancing America's better nature worldwide; and that maintaining American operation of ICANN was instrumental to keeping it that way; I'd be 100% against letting anyone else touch it. I'm just not convinced that either of those are particularly true. It is easy, and people already do wherever they have the requisite power, to operate "ICANN, minus what I dislike" and have it stick against all but moderately determined users. Some countries do it much harder than others; with the US most enthusiastic when copyright is involved; but ICANN has zero effect on how easy it is.
I know that "That depends" is the second most frustrating answer(after "yes and no"); but it is true here.
Across what geographic area, set of topics, etc. are these minutely articles distributed?
If you consider a global scale, and a fairly wide variety of interests(not necessarily serious niche stuff; but all the sections that a major Sunday print newspaper traditionally had); one article a minute is downright patchy coverage.
If you are talking a local news outlet; or a "just the foreign events large enough to be relevant" offering; it strongly suggests that they are really, really playing hard with the 'minimum publishable unit' concept.
If the once-a-minute number is across a whole stable of publications catering to different interests; then it might be the case that once you remove the celebrity gossip they actually only publish every ten minutes; Given how few genuinely just-a-local-paper operations exist these days, the quoted publication rate is probably across a media empire that isn't expected to appeal to any single individual: it'll probably have local news for more places than any one person could live/work; cultural tidbits across more fields any one person cares about; politics from around the world, and so on.
I'd argue that there are really two better questions: Ignore the stated total output; and ask "How much are they publishing that I find worth reading?" and "Is their focus on speed killing their ability to focus?" The first question is obvious: you don't enjoy news by the pound or by the word; you enjoy news by how much you actually feel like reading. The second is slightly trickier: Mere 'data' are pretty easy to come by. The sorts of news reports that you get when you give an experienced reporter plenty of time and room to dig into a matter he is experienced with are much less so. If an outfit's metrics-driven chase after viral listicles has caused them to cancel all reporting that can't be reworded from AP feeds by interns within 20 minutes; they've hollowed themselves out and it barely matters how fast they churn out "content" because none of it will add up to anything. If they just generate a lot of material because they have a lot of people reporting; that's a different matter.
Isn't that a pretty easy one? Unless you adhere to a reading of the constitution that allows for virtually no federal government activity at all(in which case ARPA probably shouldn't have ever had the cash to spend on the project; and the Department of Commerce either shouldn't exist or should be a tiny fraction of the size and scope); the US government clearly has the authority to spend allocate DoD funds to an R&D project deemed to be of military interest; to hire somebody to handle the technical work bundled under the Internet Assigned Numbers Authority; and to transfer the contract for the same functions over to the Department of Commerce once it became clear that civilian and commercial applications of the technology were where the action is.
.com bubble, companies trying to get users to point to their nameservers so that they could sell shitty vanity domains were a dime a dozen; and nobody even argued that US nationals had any duty to abide by ICANN-defined names and numbers; it's just that the market value of DNS servers that live in a strange world of their own turned out to be pretty limited). ICANN's authority, to the degree it has any, is founded in the fact that it's a pain in the ass to administer and maintain systems that have drifted out of compatibility with what the majority is using.
That doesn't mean that the US has any right to get other people to care what its DNS servers say; what media types it defines, etc; but it takes a pretty narrow reading of their powers to suggest that they don't have the authority to set up a body to publish that sort of thing in the hopes that others will adopt it because being compatible is more valuable than getting to DIY every aspect of the system.
So far as I know, nobody has ever claimed US authority over 'DNS'(indeed; back in the heady days of the
Even today, and for years now, DNS servers and other infrastructure routinely flout ICANN in situations where the benefits are greater than the costs(oddball hostnames on LANs; lazy content blocking by providing bogus IPs for sites you don't want users getting to, just choosing your own damn port because you feel like running your protocol on it, etc.) They pay more attention in places where incompatibility would hurt more: competing claims on various TLDs would get to be quite a mess; your life would really suck if your pet flavor of IP starts to differ enough that you need custom routing hardware, that sort of thing.
Nobody needs ICANN's blessing to just ignore them; but it's pretty easy to justify the Department of Commerce paying some people to be DNS jockeys.
Even if one considers the ICANN handoff to be a terrible plan; that still leaves the "and a state would have standing to block this why exactly?" problem unsolved.
I'm having trouble thinking of a reading of the constitution where one of the several states gets to stop the feds from making a change in management to a Department of Commerce contractor(handling a job previously done by a DoD contractor) overseeing the outgrowth of a federal military research project.
It'd be like Vermont going to court because they think that selling F-15s to the Saudis is a terrible plan. They may well be entirely correct; but even pretty minimalist readings of the constitution tend to give the feds most of the foreign policy power.
I'm deeply unclear on what the world thinks they'll get from ICANN that they haven't under US administration; and also unclear on what we have to gain from changing the situation; but I'm still baffled as to what possible standing state governments have on the issue.
Privacy policy. The UK, among certain others, substantially limits the degree of anonymity a donor is allowed to preserve(with respect to the recipients and the children produced; the sperm bank always wants to know a fair bit about the donor); and the trend has been toward even further emphasis on the 'right' of the child to know about their parentage; which makes even the people who are OK with the current level of information jumpy about what might happen in the future and be retroactively imposed on them.
It's not terribly hard to pay people enough to masturbate into a cup, even after you impose the various restrictions that the state of the tech(sperm counts, motility) and the state of the demand(height, educational attainment, absence of unpleasant genetic conditions) create. It is...less simple... to pay people enough to deal with the possibility that at least one mystery child might end up knocking on the door and calling them daddy in a decade or two.