And I don't think there was anything quite like an OS revolving around a web browser so thoroughly before ChromeOS came out.
Actually, there were tons of almost "browser-only" OS for ages by that point. - Lots of them using heavily customize and cut down versions of the "Embed" editions of Windows, mostly in the corporate world. - And lots relying on customized Linux installation in more hacking-friendly environment. BUT ALL OF THEM being exclusively used in "Kiosk" mode : as single use terminal to browse a captive site or to display a slideshow and other information on a public screen.
Before ChromeOS, all web-browser-based OS where single-use, because there wasn't much you could do on the web anyway.
The "tour de force" of Google isn't as much of being yet another company capable of design yet another variation of the web-kiosk paradigm. The novelty of Google is being the first company to actually PROVIDE lots of application making a web-based-os actually useful for everyday work, thank to the rest of Google's portfolio, in particular all their web-applications (GMail, Google Docs, etc.) instead of being just "yet another screen showing the weather forecast in the hotel's lobby".
I think game streaming makes sense on some level. I know they have data centers for miles and miles. And it wouldn't surprise me if some amount of them sat idle for some percentage of time.
This is exactly what I immediately though of (calling it the "equivalent of Amazon's EC2 for Google's AI servers"). They probably have tons of GPU power laying around to be used for their deep-neural-net AI. (They reported that they still rely on GPGPU for the training part, and they tensor-accelerator cards are best at applying a already-trained model). Might probably use them for video compression of newer-ish algorithm that don't have dedicated hardware compressors neither (VP-9, experiments with upcoming AV-1, etc.)
Between peaks of usage, these GPUs would otherwise stay unused. Streaming games seem a good way to monetize them during down-times.
How the latency/compression thing would be resolved I have no idea.
Also Google has data centers sprinkled all-over the world - i.e.: most of them within a reasonable ping-time of any potential customer - meaning that the latency problems are halfway solved for them, unlike for game streaming startups which need to build infrastructure from the ground up.
(And most modern GPUs are capable to use their onboard video hardware to compress their own output for streaming purpose - think Twitch, etc.)
I'm not saying that Google won't have difficulty. I'm just saying that when you are a massive corporation with datacenters all within 5 ms worth of hops of any internet enabled Chromcast HDMI stick, already stacked with giant piles of GPUs, setting up a game streaming service is a tiny bit easier than when you're a small start-up dreaming big.
And in the US (which is the jurisdiction which applies where REddit is based), these rights specifically control *commercial exploitation* (i.e.: a company deciding to sell products and earn money by using your face on them without your consent).
It absolutely doesn't cover stuff like horny teenagers using pencils to make their fantasy happen on hand-drawn comics. That sweaty 14 y.o. drooling why drawing a sex act involving Kate Perry ? Perfectly legal. (Well, until he decides to show it to his friends, then in the US it becomes a matter of under-age access to porn).
DeepFake is basically just a better pen enabling horny boys to draw their fantasies - THAT NEVER happened in the real world.
Criticizing to output of deep fake isn't anymore illegitimate than criticizing a caricature, etc. It's artists rendition of something imaginary that happens to look like a living person.
It's not actually stealing the collection of nude selfie of someone. It's drawing your own imaginary representation of what it would like (but using a better computer rather than a pencil or old-school Photoshop).
As long as it's not mis-represented (i.e.: the doctored movie being clearly stated as such, with credits where it is due to the original material that was used in the remix. Not pretending that Scarlett Johanson actually took part in hardcore BDSM session with you - that probably goes into on slander/libel territory), I fail to see what's the problem.
It should be illegal to manufacture, or offer for sale any device which has a privilege level technically feasible yet unattainable.
On the other hand, due to how things are licensed, it would be illegal for a device to allow someone to emit on frequencies for which that individual doesn't hold a license. You, as a end-user don't hold a license to operate on licensed 3G/4G frequencies, so you can't hack these. The manufacturer of your phone and the service provider you use are the one hold the license permitting them to emit on these frequencies so they get to decide what you phone does, because they have to comply to some regulations.
For some phone (looking at Qualcom, champions of "let's make the modem act as the SoC's northbridge") it's quite difficult to achieve without locking out the user. (Other more open-source friendly phone tend to lock away the closer portion into separate and segregated islets, that only talk using serial+network protocols. See Purism's upcoming Librem 5 and the Pyra handheld console as an recent examples)
The GPS has the same kind of problems : import/export law in the US forbid the commerce of device that can have a high precision at both a high altitude and a high speed (to avoid off-the-shelf parts being used to build missile targeting computers). You can't enforce that on an unlocked phone (unless the GPS is a separate chip that only talks over a serial line).
etc.
Each time, putting such a separate chip has an impact on the cost of the whole device (more component) and an impact on the battery life (a modem acting as a northbridge, means it can directly send audio to the codec or to the bluetooth chip, without any required work by the main CPU. It's possible to have a conversation while the OS is in suspended mode). So few constructors will go through the extra step and requirement to build such phones.
Or you could go with a signing infrastructure where the firmware of the modem and the GPS could be user replaceable, but you tivoize them so only legal-abiding firmware can be used. Which require extra efforts to make the signing infrastructure for specific pieces of firmware, instead of having the whole phone lockable as a giant monolithic bloc).
etc.
Abiding to these law require efforts, and locking the whole damn thing is the cheap lazy solution and most clueless consumer won't know any better anyway. So why should manufacturer care?
There are quite a lot of legal reasons to keep phone locked even before entering into the "protecting dumb users from themselves" territory (making sure only "secure" curated software is ever executable on the phone).
I'm not saying that it make mandatory to lock 100% of all phones (There are still manufacturer who make unlock-able phone that are legal to sell). I'm just saying that it might make it more difficult for a manufacturer and "lock the damn shit" is simpler and lazier.
(And then there are all the potential profits of controlling and selling software for phones. Stupid excuses on ground of "piracy" and MPAA/RIAA, etc)
You have no clue what youâ(TM)re talking about. Windows ME was a disaster. 2000 was the first mostly stable, mostly plug and play OS Microsoft released.
You have no clue how the real world works.
You where "sheltered" from Microsoft's disasters, mostly by being a geek and thus having a clue, and likely because you were already working in some IT field (your enterprise's IT department) which was more likely to pay attention to the business line of Windows (WinNT 3.5, Win NT 4, Win 2000), or at least worked in a company whose IT department got business OS installed (either by ordering business line desktops from a manufacturer, or by buying license for a business OS and installing it).
Most "normal people", just "bought a computer" from the electronics shop (or worse, from the supermarket) and are completely oblivious to what a OS is, and just use whatever shit comes preinstalled on the computer (include horrible Microsoft attempts at making home-oriented OSes, bloatware packaged together by money-seeking manufacturer, extra bonus shit installed by the shop, etc.) So most people "just endured" Windows ME because this despicable shit is what came pre-installed on most home computers. Windows 2000 was a business-oriented OS that didn't come preinstalled on these, and most people did have a clue to go a separately buy it.
Windows 2000 was NT version 5.0. XP was NT version 5.1. That is to say that XP was Windows 2000 rebranded and repackaged with a different UI and Internet based Product Activation and marketed toward consumers because the NT code base provided to be better than the bastardized 95/98/ME codebase ever was.
If you followed magazine back then, the initial plan of Microsoft was to make a "Windows 2000 Home" - they actually didn't want to make any 9x Windows after 98, they new really well that the weird infrastructure of bolting a semi-modern OS over an MS-DOS base layer wasn't a brilliant long term strategy.
It just took them more time than expected to develop it successfully and for the market to evolve to the point where they can accept without any problem a NT-based windows with no MS-DOS layer (they only managed it by the time of Win XP), and decided to fill-in the time in between with attempt to stretch the longevity of the 9x serie (98SE was a successfull successor of 98, WinME was a rushed "oh my god, 2000 Home isn't ready, we must quickly make a fill-in" with frankenstein-bolted bits on it in an attempt to keep it modern).
Windows 2000 is one of the best operating systems Microsoft ever produced. Period. End of discussion.
Well if you use "Microsoft ever produced." as a benchmark, you're setting the bar extremely low.
The plugin is a proof-reading tool. It makes all the nice colored wavy line under your mistakes.
It works in an TEXTAREA> <INPUT TYPE="text"> etc.
This particular plug-in doesn't do the proof reading it self, it sends the text-to-be-corrected to some cloud server where the actual proof reading algorithms run.
So for the plugin to work (and colored wavy line to appear), the plugin needs to send everything you type out of your computer.
It's basically a giant keylogger - BY DESIGN.
It's just that some attackers have found a way to tap into the traffic and benefit from the built-in key-loging too.
But it's the whole design of Grammarly which is flawed to begin with.
Based on the adverts I've seen for this service, it looks like it is first-and-foremost a browser-based keylogger anyway, with the copy editing features just being the hook to get people to install (and pay?) for the 'service'.
Yup, I find it personally disturbing that people will let some shady 3rd party unknown server somewhere in Ukraine access (for "proof reading") every single thing they type online.
You're better off using some technology that can be installed locally (or on your own-controlled servers):
e.g.: LanguageTool - it has a webextension - it can be downloaded as a stand-alone version. (- and of course, you can point the extension to the URL of your stand-alone server)
(both of the above are Free/Libre OpenSource Software, so auditable against nefarious code)
It's kind of freaky that it's population is growing so quickly and exponentially.
Actually no surprise. These are not animal that care for their offsprings. So they tend to lay a ton of eggs in hopes that a few of them manage to survive into reproductive adulthood.
The surprise isn't the massive amount of egg. (that would be a surprise for mammals like humans. that's the normal modus operandi for crayfish) The surprise is that without any fertilization happening, the eggs are able to hatch and grow into an adult egg-laying female. (Wikipedia mentions the animal having some chromosomic aberration and being triploid - thus the meiosis going wrong when trying to produce egg cells)
I wonder what kinds of havoc it is wrecking on local ecosystems.
Well, given that its current ecosystem is "aquariums" : its mostly pissing off their respective owner who simply don't know what to do with all the newly hatching crayfishes.
If let out freely in the nature :
- it should have predators in the nature. Not all eggs will hatch and grow to a reproductive adult. There's a reason why the reproductive strategy of all similar animals is "lay as many eggs as possible in the hope that some survive". The population is currently explosing only because it's happening in the sheltered environment of aquarium, with humans taking care of their pets, and not much predation.
- if too many offsprings do hatch, they'll have to eat : they'll be in competion with all other animal in the environment, and they'll be in competition with all the other offsprings. So mostly they'll probably stave very quickly, either by being outcompeted by other animal and/or by depleting the resources in their immediate environment.
So the damage to the environment depends on how fast they'll die.
Was this thing man made?
Technically, given that it currently survives because it's living in aquariums : yes, we human have contributed a bit to it. But no, it definitely doesn't look like something coming out of some lab.
Wasn't there supposed to be some sort of replicative fading where the telomeres or some such got shorter over time until the clones were not viable?
Yes : *body* cell lines have such kind of control to avoid cells replicating too much.
But eggs are formed from *germ* cell lines, which are (by Nature's "design" - as its supposed to be the way a specie reproduce) not having any replication limit.
These are not "dolly the sheep"-style clones (a body cell used to form a new clone by transfering the nucleus, and might inherit some of the replication limits of the body cell).
These are parthenogenesis-style (see christian mythology about virgin mary) clones : some how, an non-fertilized egg-cell managed to grow into a full grown individual (usually, in other species, there's some kind of bug in the "meiosis" - the process is supposed to split the normally "diploid" genome into half "haploid". The wikipedia mentions that they are "triploid" - so having triplets instead of chromosome pairs - so it's quite a big bug in the meiosis procedure).
But it is an egg cell (although a chromosomaly abnormal one) that produces the whole individual, so no telomeres problems nor any other cell division limitations.
Yup, indeed, wrong link (about Telegram being pulled out of the appstore due to Child pornography).
Speaking of which :
- How did Apple check the existence of Child pornography ? Wasn't Telegram supposed to be exclusively using end-to-end encryption ? Or did they ask investigating police officers to start chat with CP-distributors ?... (read the Wikipedia article...) Ah okay. end-to-end encryption isn't default and users need to initiate "secret chat" to enable it.
- Telegram is only a communication service enabling end user to exchange message. It shouldn't be liable for what the end-users are exchanging. (Just as the ISP isn't responsible for the internet use of their client).
- Also given that the suspected criminals are stupid enough to NOT use the end-to-end encryption and be easily checked by Apple, it should be easy to prosecute *them* directly.
so the manufacturers & carriers were responsible for the updates (if any) and in their business model, it's not worth the time since people just toss their phones away about every 12-24 months,
It's more precisely the business model that they *wish* to have.
Reality is a bit different. Some old phone remain for use much more longer, usually changing hand several time (second hand market) and eventually ending up in third-world countries.
To everybody can actually afford to re-buy a brand new shining phone every other year. And not a lot of manufacturer can cater to poorer markets with modestly speced phone at bargain bin prices.
But meanwhile, the money that the manufacturer could have made has already been made eons ago (at purchase time), and the manufacturer have no financial incentive to keep spending engineer's salaries to support extremely old phone that won't bring in any penny more. Better direct those engineer to making the next "shiny gadget phone stuff" that you can persuade people to fork money on because the previous one is "so outdated" (It's so last year! Buy the one! Now 0.1 mm even thiner ! you can finally cut cheese with it ! - DISCLAIMER: and break it when sitting with it in your pant's back pocket. )
So planed obsolescence more or less invite itself on the scene due to how the market is organized.
Hope fully some modularity will help against it:
Perhaps going forward, past Oreo, if more manufacturers adopt the project Treble, which splits the core OS from the UI/bloat, Google can get the updates to the core of the system faster, through the play store
(Cue in the "I have a 32GiB smartphone, but only 18GiB of actually free space, because a massive amount of space is lost to a gazillion of different partitions" typical android-phone complaint...)
Yup, indeed, I hope that it help getting fixes easier in the future.
(Google can easily update the core. And with their insisting on kernels to be at a minimum (LTS) version, there's a chance that the community might also be able to churn out update to the non-blob part - see partition above - of the Linux kernel)
As a side note : That's also an advantage of some full blown GNU/Linux smartphone OS - like Sailfish OS. The "mer" core of the OS is shared among all phones. So whenever Sailfish OS gets an update (e.g.: Sailfish X recently made available for Sony Xperia X in partnet ship with Sony's Opendevice program), all the older devices all the way back to 2013's Jolla 1 smartphone can benefit from it, even if the kernel is stuck to some antique 3.2 version due to Qualcom not releasing any blob update.
I know that joking that "if Skynet becomes self-aware with human-level emotions now during the internet era it will be m ore likely to become over-obsessed with kittens rather than with winning a war against humans" has been a staple of internet humor.
But it was supposed to be a joke. Not a prediction / design plan for you first smartphone embed neural net chip.
And given the design of the previous record holder he is only slightly switching the design around replacing a pair of rockets strapped on a car with a car strapped on a triplet of rockets.
It sounds like AMD should come out of this quite well.
At least much better than the giant pile of mess that is Intel. That's why some experts are pissed at Intel trying to muddy things and pretend all CPUs are equal. (Nope. All CPU are equal in *Spectre v1* only. Intel's peculiar way to optimize at the cost of everything else including safety and sanity stands out a lot in Spectre v2 and Meltdown).
I can't decide whether variant 1 sounds "possibly dangerous" or not. I suppose it depends on how applications segment their data.
Yup. There's a reason why web browsers have moved (Chrome) or are moving (the whole reason to switch Firefox from XUL to WebExtensions is to enable Electrolysis by default) to multi-proc models. Eventually none of the critical data (e.g.: Password Manager extensions) and externally provided arbitrary code (Javascript on websites) will be living in the same process. And there's a reason why the JITting of eBPF isn't enabled in the Linux kernel by default.
But I'm really skeptical about speculative execution in hyperthreads in any case.
Hyperthreading and speculative execution are completely orthogonal to each other.
They are two completely different strategies in answering the problem of how to keep the pipeline fed, each time it stalls (e.g.: while waiting for something to be fetched from the memory, or some long computation to finish).
You don't know what you should do next : - Speculative execution :...so you make your best guess, and try to do it anyway. If your guess turns out right, you're gaining some execution speed. - Hyperthread:...but there are N other task currently wait for which you DO know what to do next. Do them instead.
Speculative execution comes at lots of complexity (in order to be able to invalidate wrong turns) with devil lurking in the implementations details (side-channels, security checks done too late). Hyperthreading is much closer to normal execution and simply require doubling some already existing facilities in order to enable the CPU tracking N tasks.
OTOH, given the way applications are currently written, I can see why they did it, I just think it's a poor local optimum.
Hyperthreading is the *better* solution... except that, for it to work, it requires to have N other tasks in the wait. Hyperthreading works better in heavily multi-tasking use cases. So you'll find it on servers (typical server-only CPUs like the UltraSparc Niagaras had 8 threads per CPU core) and on GPUs (as anyone used to CUDA knows, the basic strategy is keeping as fucking many threads in flight as possible).
It works on task which are heavily parallel (tons of servers and daemons, or tons of pixels, etc.)
But it doesn't help on task that as mostly single threaded so it doesn't look nice on lots of benchmarks (e.g.: older games, lots of compression tasks).
In other words, Multithreading won't make your Windows go faster (Even more so as Windows has traditionally lagged behind the starte of the art process schedulers). "Local optimum" as you say.
For instance, just because you store Game of Thrones in Postgres doesn't mean you are not infringing copyright by sharing that database.
In your example, the art assets (video content) is what is copyrighted (the Postgres is irrelevant. It's just a container, functionally equivalent to storing the stream in Matroska).
The DCMA takedown notice mentionned in the article doesn't list any copyrightable asset (no artworks, etc.)
Duplicating a collection of factoids isn't copyrightable.
And all the arguments of the notice are similar to "Waaah! Their table of spells looks just like ours back then!"
NOTE: Not the artwork of the spells (which is on the game disk legally bought by the players, not on the servers). Just the numbers giving the strength of the spells and maybe a few of their serial numbers.
Basically, what is reported to be duplicated looks like a "phone book of spells". And you definitely can't copyright a phone book in the US.
Also similarly, SCO Unix' arguments against Linux about content of *system header files* comes to mind.
If indeed there were no Blizzard copyrighted assets (think giant maps, artwork) but only general look-alike-ness of a couple of structures and a few numbers - as seems to be indicated by the cited DMCA takedown notice, this classic server emulator show have a solid case, even in the US.
- you need to replicate functionality in order to emulate in a compatible way. - you can't copyright numbers (- and you can't copy right a collection such as a database *by itself* (i.e.: without any further creative work) , e.g.: a phone book's content isn't copyrightable, though a nice illustration on its cover could be. the SQL designated in the DMCA take-down notice border on being a "phone book (-like) of objects".)
If you look in the URL linked in the summary (Yeah I know,/. and actual article reading) : it might seem debatable.
The complains hinge around 2-3 sql file using names and having a few data that looks like the data used in wow (spells have the same characteristic as old versions of wow, same old trademarked names are used, etc.).
Fantasy Names – “Script” files and folders are named after and reference WoW fantasy names.
They're not complaining about game assets being lifted of blizzard's own software (e.g.: bitmaps, etc.) They are complaining about the code using official Blizzard trademarked name to designate Blizzard's said trademarked characters. (Note: e.g. it's not a trademark violation when you use microsoft's trademarked "Microsoft Windows" name to speak about Microsoft Windows itself).
They're complaining that the datamodel is very close to how it used to be in old servers:
The LightsHope spell table has identical layout and typically identical field names as the table from early WoW. We use database tables to represent game data, like spells, in WoW. In our code, we use.sql files to represent the data layout of each table (i.e. the fields of each specific table, like a spell name or the magnitude of its effect). MaNGOS, the platform off of which Light’s Hope appears to be built, uses a similar structure. The LightsHope spell_template table matches almost exactly the layout and field names of early WoW client database tables.
(Looks like the devs made their "Classic" recreation by using old dumps / backups as a referrence).
Matching Record IDs – There are “scripts” that reference database records directly by ID; there are cases where these IDs directly match the ID from WoW’s content.
"Hey, their serial numbers looks suspiciously close to our serial numbers !" Oh, come one. (Numbers aren't copyrightable in the US. That's basically Intel complaining that competitor's 386-compatible chip also use numbers like 386)
None of the complain is anything that looks like : "these huges chunks of code are actually a un-licensed copy of the network code of our server".
Overall : Some of the complain could almost fell under the "but these old numbers are necessary to get interoperation" exemption that exists to copyright in some jurisdiction (other/.ers have mentioned Canada as an example).
AMD considers their CPUs potentially vulnerable to Spectre Variant 2 - "Branch Target Injection". (The one were one attacker application is able to do it bidding into a completely different and innocent target application)
Some more recent AMD processors do indirect branch prediction. But the way they do this indirect branch prediction is completely different. Currently the Google demo code against Intel Xeon doesn't work (well, obviously). Nobody has managed to write a successful exploit of that variant. AMD engineer believe that it's a terribly difficult task that might not be doable. So they might indeed end up being more or less Intel-specific-ish.
Sèectre variant 1 - "boundary check bypass" is the one affecting every CPU that those speculative execution and is basically "speculative execution working as it should/as documented, but now somebody has found a way to use it as a side-channel attack to have a software see its own data".
All the cells in all the multicellular organisms are cancerous.
For asexually reproducing organisms, there is nothing called cancer. The cells keep mutating and dividing. So what is cancer for them?
Nope. Not at all. Cancer is not simply cell division. (Though there are some cell that do indeed not divide (e.g.: neuron, for obvious practical reasons) and whose population is repleted by progenitor cells (in that case, that would be neuroblast ; mostly happening in the amydalia region of the brain), there are other cell population were dividing cells are pretty much the norm (e.g.: cardiomyocytes in the heart do divide to replenish the population).)
Cancer is about complete uncontrolled cell divsion. Not only *unchecked* but utterly beyond any control or even coherent organisation.
Obligatory/. car metaphore: - if cells dividing are cars driving forward - then cancerous cells are cars witl the pedal's mechanism completely stuck, while in full throttle position. With all consequence that entail with that (including colliding everywhere)
A fully cancerous cell (once it has moved beyond hyperplasia toward full blown cancer ), will divide completely chaotically, even before it's actually ready to divide. It won't simply divide even when unneeded/unrequired by the body (that's "hyperplasia" and that's what you're thinking with your "revolution against the oppressor" logic) It will divide even when it doesn't make any sense : - Before having checked that the duplicated genetic material is correct (cancerous cells accumulate mutation at an alarming rate. On a global scale they are terribly inefficient : lots of them die just because they've completely destroyed their DNA. It's just that, on scale of division speed happening in a tumor, there are still a few that miraculously manage to be still semi functionnal enough to keep reproducing. It's survival of the fittest, but with the production of unfits turned up to eleven). - Before even having correctly duplicated its genetic material (chromosomic aberration are abundant in a tumor, leading to lots of dysfunctional cell) - Before even having accumulated enough resources to be functional. That's why you don't see cancer in single-cell organism : a cancerous cell isn't even able to function anymore, and require an organism on which to parasytically to rely in order to sustain. (A cancerous amoeba would be unable to eat and will die after a couple of divisions - cancerous cell are defined by their loss of function). (Just like the metaphorical car would run out of fuel pretty fast. From that point of view a cancer is closer to a trolley-bus : able to tap into the city's resources (electric grid) to still drive. And just like the cancer cell, a lot of the trolley bus would derail their electrical feed and die of)
A cancerous cell isn't simply "reverting to a pre-multicellular state". A cancerous cell is going batshit insane about division in a way that could not be survivable outside a multicellular organism.
Are you also saying we as a society shouldn't try and shut down sketchy con-artist retailers because you're not stupid enough to fall for what they're selling and should be able to waste your money if you want to?
Or maybe, we should reconsider the education system a bit : Maybe instead of blocking targets that are semi arbitrary labelled as "con artists" on the grounds of trying to shelter stupid people, we should instead educate stupid people and teach them how to use their brain and do their due information search to not fall for cons/snake oil/conspiracy theory/etc. ?
Maybe if people just didn't forget how to use their brains, we'll have a lot less problems with people falling to stupid tricks.
On the other end, none of my media players (Xbox One, Nvidia Shield, Chromecast, AppleTV, etc) support it or my 4k projector (All support MPEG2/h.264/HEVC).
The idea of AV-1 creators, is to use GPGPU approaches for the hardware acceleration for the first devices supporting it.
So your Xbox, Shield and Chromecast will likely get the support added in a future OS upgrade, probably managing to get fluid play (but maybe at a higher power consumption than MPEG-AVC/H264). After all, AV-1 is also going to be IETF's NetVC, and all these device boast web browsers.
Apple, being one of the HEVC licensor is probably going to drag their feet much longer.
HEVC is already entrenched in the UHD standard, so that isn't going away any time soon.
From being used in TV ? Indeed, it's going to stay because of various UHD DVB-xxx standards, UHD BlueRay, etc.
From being present in gadgets ? Well, you'll have to check the bargain bin at your local electronics store : not all dead-cheap no-name asian tablet feature actual hardware MPEG-HEVC/H265 acceleration.
That's actually a complain of the chipset manufacturer : licenses fee for H265 are so high, that they become a significant fraction of the price of the crappy cheap SoC that are used in such tablets. (Hint: the SoC isn't the most expensive component in a crappy tablet. The screen and the battery are. The SoC are actually quite cheap. When a no-name table manufacturer shops for parts, they'll try to get the cheapest possible part, and H265-less SoCs are actually significantly cheaper at the volume that interests them).
I'm not making this up, it's actually been cited as a rationnal behing starting the whole Daala, AV-1, NetVC story.
So unless they can get the content creators (TV Shows and movies) to switch, then AV-1 will be a "web-only" technology unfortunately or we will have to wait for a UHD 2.0 standard.
Yes, for the close future, AV-1 / NetVC is going to be an internet-only technology (mostly web and apps), just like OPUS is currently (Digital Radio Mondial unofficial experiment non withstanding). (Though I wouldn't be surprise if there are similar experiments in using AV-1 in some long range digital media showing eventually, as part of small scale local , experiments or in less wealthy region of the world).
But given amount of media consumed online, that's actually the lion's share of the audiovisual world. By volume, *that* is what is going to be relevant.
There was a/. entry not so long ago that even in the US, the tipping point is being reach where more media is consumed over the net than over TV.
And again, like Vorbis before it, because it's free to implement, lots of no-name gadgets are going to implement it eventually, just to add another bullet point on their feature list, even if it's not going to be advertised on the stickers on the box.
Eventually, if TV standards are still relevant by then, it might show up in some UHD2.0 standard (DVB-T3... ?) but the whole question if TV is going to remain relevant by then remains. It might even show up in some media descendant of BlueRay disc if such a thing still exists by then. Streaming is probably going to be the most widespread media consumption mean, be it on some service provider's box (cable box, except relying more on IPv6 than on DVB-C), or with website / apps (think Youtube, Hulu, Netflix, etc.)
The above AC would have noticed what difference in Sun vs. Moon eclipse means, and would have told you that you actually can't go blind from a lunar eclipse,
but he cannot read anymore because hE WENT BLIND! by attempting to use eyes during the last solar eclipse.
Slashdot Mirror
And I don't think there was anything quite like an OS revolving around a web browser so thoroughly before ChromeOS came out.
Actually, there were tons of almost "browser-only" OS for ages by that point.
- Lots of them using heavily customize and cut down versions of the "Embed" editions of Windows, mostly in the corporate world.
- And lots relying on customized Linux installation in more hacking-friendly environment.
BUT ALL OF THEM being exclusively used in "Kiosk" mode : as single use terminal to browse a captive site or to display a slideshow and other information on a public screen.
Before ChromeOS, all web-browser-based OS where single-use, because there wasn't much you could do on the web anyway.
The "tour de force" of Google isn't as much of being yet another company capable of design yet another variation of the web-kiosk paradigm.
The novelty of Google is being the first company to actually PROVIDE lots of application making a web-based-os actually useful for everyday work, thank to the rest of Google's portfolio, in particular all their web-applications (GMail, Google Docs, etc.) instead of being just "yet another screen showing the weather forecast in the hotel's lobby".
I think game streaming makes sense on some level. I know they have data centers for miles and miles. And it wouldn't surprise me if some amount of them sat idle for some percentage of time.
This is exactly what I immediately though of (calling it the "equivalent of Amazon's EC2 for Google's AI servers").
They probably have tons of GPU power laying around to be used for their deep-neural-net AI. (They reported that they still rely on GPGPU for the training part, and they tensor-accelerator cards are best at applying a already-trained model).
Might probably use them for video compression of newer-ish algorithm that don't have dedicated hardware compressors neither (VP-9, experiments with upcoming AV-1, etc.)
Between peaks of usage, these GPUs would otherwise stay unused.
Streaming games seem a good way to monetize them during down-times.
How the latency/compression thing would be resolved I have no idea.
Also Google has data centers sprinkled all-over the world - i.e.: most of them within a reasonable ping-time of any potential customer - meaning that the latency problems are halfway solved for them, unlike for game streaming startups which need to build infrastructure from the ground up.
(And most modern GPUs are capable to use their onboard video hardware to compress their own output for streaming purpose - think Twitch, etc.)
I'm not saying that Google won't have difficulty. I'm just saying that when you are a massive corporation with datacenters all within 5 ms worth of hops of any internet enabled Chromcast HDMI stick, already stacked with giant piles of GPUs, setting up a game streaming service is a tiny bit easier than when you're a small start-up dreaming big.
In most of the world, you have the right to control your own image.
And in the US (which is the jurisdiction which applies where REddit is based), these rights specifically control *commercial exploitation*
(i.e.: a company deciding to sell products and earn money by using your face on them without your consent).
It absolutely doesn't cover stuff like horny teenagers using pencils to make their fantasy happen on hand-drawn comics.
That sweaty 14 y.o. drooling why drawing a sex act involving Kate Perry ? Perfectly legal.
(Well, until he decides to show it to his friends, then in the US it becomes a matter of under-age access to porn).
DeepFake is basically just a better pen enabling horny boys to draw their fantasies - THAT NEVER happened in the real world.
Criticizing to output of deep fake isn't anymore illegitimate than criticizing a caricature, etc.
It's artists rendition of something imaginary that happens to look like a living person.
It's not actually stealing the collection of nude selfie of someone.
It's drawing your own imaginary representation of what it would like (but using a better computer rather than a pencil or old-school Photoshop).
As long as it's not mis-represented (i.e.: the doctored movie being clearly stated as such, with credits where it is due to the original material that was used in the remix. Not pretending that Scarlett Johanson actually took part in hardcore BDSM session with you - that probably goes into on slander/libel territory), I fail to see what's the problem.
It should be illegal to manufacture, or offer for sale any device which has a privilege level technically feasible yet unattainable.
On the other hand, due to how things are licensed, it would be illegal for a device to allow someone to emit on frequencies for which that individual doesn't hold a license.
You, as a end-user don't hold a license to operate on licensed 3G/4G frequencies, so you can't hack these.
The manufacturer of your phone and the service provider you use are the one hold the license permitting them to emit on these frequencies so they get to decide what you phone does, because they have to comply to some regulations.
For some phone (looking at Qualcom, champions of "let's make the modem act as the SoC's northbridge") it's quite difficult to achieve without locking out the user.
(Other more open-source friendly phone tend to lock away the closer portion into separate and segregated islets, that only talk using serial+network protocols. See Purism's upcoming Librem 5 and the Pyra handheld console as an recent examples)
The GPS has the same kind of problems : import/export law in the US forbid the commerce of device that can have a high precision at both a high altitude and a high speed (to avoid off-the-shelf parts being used to build missile targeting computers).
You can't enforce that on an unlocked phone
(unless the GPS is a separate chip that only talks over a serial line).
etc.
Each time, putting such a separate chip has an impact on the cost of the whole device (more component) and an impact on the battery life (a modem acting as a northbridge, means it can directly send audio to the codec or to the bluetooth chip, without any required work by the main CPU. It's possible to have a conversation while the OS is in suspended mode).
So few constructors will go through the extra step and requirement to build such phones.
Or you could go with a signing infrastructure where the firmware of the modem and the GPS could be user replaceable, but you tivoize them so only legal-abiding firmware can be used.
Which require extra efforts to make the signing infrastructure for specific pieces of firmware, instead of having the whole phone lockable as a giant monolithic bloc).
etc.
Abiding to these law require efforts, and locking the whole damn thing is the cheap lazy solution and most clueless consumer won't know any better anyway.
So why should manufacturer care?
There are quite a lot of legal reasons to keep phone locked even before entering into the "protecting dumb users from themselves" territory (making sure only "secure" curated software is ever executable on the phone).
I'm not saying that it make mandatory to lock 100% of all phones (There are still manufacturer who make unlock-able phone that are legal to sell).
I'm just saying that it might make it more difficult for a manufacturer and "lock the damn shit" is simpler and lazier.
(And then there are all the potential profits of controlling and selling software for phones.
Stupid excuses on ground of "piracy" and MPAA/RIAA, etc)
You have no clue what youâ(TM)re talking about. Windows ME was a disaster. 2000 was the first mostly stable, mostly plug and play OS Microsoft released.
You have no clue how the real world works.
You where "sheltered" from Microsoft's disasters, mostly by being a geek and thus having a clue, and likely because you were already working in some IT field (your enterprise's IT department) which was more likely to pay attention to the business line of Windows (WinNT 3.5, Win NT 4, Win 2000), or at least worked in a company whose IT department got business OS installed (either by ordering business line desktops from a manufacturer, or by buying license for a business OS and installing it).
Most "normal people", just "bought a computer" from the electronics shop (or worse, from the supermarket) and are completely oblivious to what a OS is, and just use whatever shit comes preinstalled on the computer (include horrible Microsoft attempts at making home-oriented OSes, bloatware packaged together by money-seeking manufacturer, extra bonus shit installed by the shop, etc.)
So most people "just endured" Windows ME because this despicable shit is what came pre-installed on most home computers. Windows 2000 was a business-oriented OS that didn't come preinstalled on these, and most people did have a clue to go a separately buy it.
Windows 2000 was NT version 5.0. XP was NT version 5.1.
That is to say that XP was Windows 2000 rebranded and repackaged with a different UI and Internet based Product Activation and marketed toward consumers because the NT code base provided to be better than the bastardized 95/98/ME codebase ever was.
If you followed magazine back then, the initial plan of Microsoft was to make a "Windows 2000 Home" - they actually didn't want to make any 9x Windows after 98, they new really well that the weird infrastructure of bolting a semi-modern OS over an MS-DOS base layer wasn't a brilliant long term strategy.
It just took them more time than expected to develop it successfully and for the market to evolve to the point where they can accept without any problem a NT-based windows with no MS-DOS layer (they only managed it by the time of Win XP),
and decided to fill-in the time in between with attempt to stretch the longevity of the 9x serie (98SE was a successfull successor of 98, WinME was a rushed "oh my god, 2000 Home isn't ready, we must quickly make a fill-in" with frankenstein-bolted bits on it in an attempt to keep it modern).
Windows 2000 is one of the best operating systems Microsoft ever produced. Period. End of discussion.
Well if you use "Microsoft ever produced." as a benchmark, you're setting the bar extremely low.
The plugin is a proof-reading tool.
It makes all the nice colored wavy line under your mistakes.
It works in an TEXTAREA> <INPUT TYPE="text"> etc.
This particular plug-in doesn't do the proof reading it self,
it sends the text-to-be-corrected to some cloud server where the actual proof reading algorithms run.
So for the plugin to work (and colored wavy line to appear), the plugin needs to send everything you type out of your computer.
It's basically a giant keylogger - BY DESIGN.
It's just that some attackers have found a way to tap into the traffic and benefit from the built-in key-loging too.
But it's the whole design of Grammarly which is flawed to begin with.
Based on the adverts I've seen for this service, it looks like it is first-and-foremost a browser-based keylogger anyway, with the copy editing features just being the hook to get people to install (and pay?) for the 'service'.
Yup, I find it personally disturbing that people will let some shady 3rd party unknown server somewhere in Ukraine access (for "proof reading") every single thing they type online.
You're better off using some technology that can be installed locally (or on your own-controlled servers):
e.g.: LanguageTool
- it has a webextension
- it can be downloaded as a stand-alone version.
(- and of course, you can point the extension to the URL of your stand-alone server)
(both of the above are Free/Libre OpenSource Software, so auditable against nefarious code)
It's kind of freaky that it's population is growing so quickly and exponentially.
Actually no surprise.
These are not animal that care for their offsprings.
So they tend to lay a ton of eggs in hopes that a few of them manage to survive into reproductive adulthood.
The surprise isn't the massive amount of egg. (that would be a surprise for mammals like humans. that's the normal modus operandi for crayfish)
The surprise is that without any fertilization happening, the eggs are able to hatch and grow into an adult egg-laying female.
(Wikipedia mentions the animal having some chromosomic aberration and being triploid - thus the meiosis going wrong when trying to produce egg cells)
I wonder what kinds of havoc it is wrecking on local ecosystems.
Well, given that its current ecosystem is "aquariums" : its mostly pissing off their respective owner who simply don't know what to do with all the newly hatching crayfishes.
If let out freely in the nature :
- it should have predators in the nature. Not all eggs will hatch and grow to a reproductive adult. There's a reason why the reproductive strategy of all similar animals is "lay as many eggs as possible in the hope that some survive".
The population is currently explosing only because it's happening in the sheltered environment of aquarium, with humans taking care of their pets, and not much predation.
- if too many offsprings do hatch, they'll have to eat : they'll be in competion with all other animal in the environment, and they'll be in competition with all the other offsprings.
So mostly they'll probably stave very quickly, either by being outcompeted by other animal and/or by depleting the resources in their immediate environment.
So the damage to the environment depends on how fast they'll die.
Was this thing man made?
Technically, given that it currently survives because it's living in aquariums : yes, we human have contributed a bit to it.
But no, it definitely doesn't look like something coming out of some lab.
Wasn't there supposed to be some sort of replicative fading where the telomeres or some such got shorter over time until the clones were not viable?
Yes : *body* cell lines have such kind of control to avoid cells replicating too much.
But eggs are formed from *germ* cell lines, which are (by Nature's "design" - as its supposed to be the way a specie reproduce) not having any replication limit.
These are not "dolly the sheep"-style clones (a body cell used to form a new clone by transfering the nucleus, and might inherit some of the replication limits of the body cell).
These are parthenogenesis-style (see christian mythology about virgin mary) clones : some how, an non-fertilized egg-cell managed to grow into a full grown individual (usually, in other species, there's some kind of bug in the "meiosis" - the process is supposed to split the normally "diploid" genome into half "haploid". The wikipedia mentions that they are "triploid" - so having triplets instead of chromosome pairs - so it's quite a big bug in the meiosis procedure).
But it is an egg cell (although a chromosomaly abnormal one) that produces the whole individual, so no telomeres problems nor any other cell division limitations.
Android Oreo took five months to pass the 1 percent adoption mark.
Yup, indeed, wrong link (about Telegram being pulled out of the appstore due to Child pornography).
Speaking of which :
- How did Apple check the existence of Child pornography ? Wasn't Telegram supposed to be exclusively using end-to-end encryption ? Or did they ask investigating police officers to start chat with CP-distributors ?... (read the Wikipedia article...) Ah okay. end-to-end encryption isn't default and users need to initiate "secret chat" to enable it.
- Telegram is only a communication service enabling end user to exchange message. It shouldn't be liable for what the end-users are exchanging. (Just as the ISP isn't responsible for the internet use of their client).
- Also given that the suspected criminals are stupid enough to NOT use the end-to-end encryption and be easily checked by Apple, it should be easy to prosecute *them* directly.
so the manufacturers & carriers were responsible for
the updates (if any) and in their business model, it's not worth the time since people
just toss their phones away about every 12-24 months,
It's more precisely the business model that they *wish* to have.
Reality is a bit different. Some old phone remain for use much more longer, usually changing hand several time (second hand market) and eventually ending up in third-world countries.
To everybody can actually afford to re-buy a brand new shining phone every other year. And not a lot of manufacturer can cater to poorer markets with modestly speced phone at bargain bin prices.
But meanwhile, the money that the manufacturer could have made has already been made eons ago (at purchase time), and the manufacturer have no financial incentive to keep spending engineer's salaries to support extremely old phone that won't bring in any penny more.
Better direct those engineer to making the next "shiny gadget phone stuff" that you can persuade people to fork money on because the previous one is "so outdated"
(It's so last year! Buy the one! Now 0.1 mm even thiner ! you can finally cut cheese with it ! - DISCLAIMER: and break it when sitting with it in your pant's back pocket. )
So planed obsolescence more or less invite itself on the scene due to how the market is organized.
Hope fully some modularity will help against it :
Perhaps going forward, past Oreo, if more manufacturers adopt the project Treble, which splits the core OS from the UI/bloat, Google can get the updates to the core of the system faster, through the play store
(Cue in the "I have a 32GiB smartphone, but only 18GiB of actually free space, because a massive amount of space is lost to a gazillion of different partitions" typical android-phone complaint...)
Yup, indeed, I hope that it help getting fixes easier in the future.
(Google can easily update the core. And with their insisting on kernels to be at a minimum (LTS) version, there's a chance that the community might also be able to churn out update to the non-blob part - see partition above - of the Linux kernel)
As a side note : That's also an advantage of some full blown GNU/Linux smartphone OS - like Sailfish OS.
The "mer" core of the OS is shared among all phones. So whenever Sailfish OS gets an update (e.g.: Sailfish X recently made available for Sony Xperia X in partnet ship with Sony's Opendevice program), all the older devices all the way back to 2013's Jolla 1 smartphone can benefit from it, even if the kernel is stuck to some antique 3.2 version due to Qualcom not releasing any blob update.
Come on, Google.
I know that joking that "if Skynet becomes self-aware with human-level emotions now during the internet era it will be m ore likely to become over-obsessed with kittens rather than with winning a war against humans" has been a staple of internet humor.
But it was supposed to be a joke. Not a prediction / design plan for you first smartphone embed neural net chip.
And given the design of the previous record holder he is only slightly switching the design around replacing a pair of rockets strapped on a car with a car strapped on a triplet of rockets.
It sounds like AMD should come out of this quite well.
At least much better than the giant pile of mess that is Intel.
That's why some experts are pissed at Intel trying to muddy things and pretend all CPUs are equal.
(Nope. All CPU are equal in *Spectre v1* only.
Intel's peculiar way to optimize at the cost of everything else including safety and sanity stands out a lot in Spectre v2 and Meltdown).
I can't decide whether variant 1 sounds "possibly dangerous" or not. I suppose it depends on how applications segment their data.
Yup.
There's a reason why web browsers have moved (Chrome) or are moving (the whole reason to switch Firefox from XUL to WebExtensions is to enable Electrolysis by default) to multi-proc models. Eventually none of the critical data (e.g.: Password Manager extensions) and externally provided arbitrary code (Javascript on websites) will be living in the same process.
And there's a reason why the JITting of eBPF isn't enabled in the Linux kernel by default.
But I'm really skeptical about speculative execution in hyperthreads in any case.
Hyperthreading and speculative execution are completely orthogonal to each other.
They are two completely different strategies in answering the problem of how to keep the pipeline fed, each time it stalls (e.g.: while waiting for something to be fetched from the memory, or some long computation to finish).
You don't know what you should do next : ...so you make your best guess, and try to do it anyway. If your guess turns out right, you're gaining some execution speed. ...but there are N other task currently wait for which you DO know what to do next. Do them instead.
- Speculative execution :
- Hyperthread:
Speculative execution comes at lots of complexity (in order to be able to invalidate wrong turns) with devil lurking in the implementations details (side-channels, security checks done too late).
Hyperthreading is much closer to normal execution and simply require doubling some already existing facilities in order to enable the CPU tracking N tasks.
OTOH, given the way applications are currently written, I can see why they did it, I just think it's a poor local optimum.
Hyperthreading is the *better* solution... except that, for it to work, it requires to have N other tasks in the wait.
Hyperthreading works better in heavily multi-tasking use cases. So you'll find it on servers (typical server-only CPUs like the UltraSparc Niagaras had 8 threads per CPU core) and on GPUs (as anyone used to CUDA knows, the basic strategy is keeping as fucking many threads in flight as possible).
It works on task which are heavily parallel (tons of servers and daemons, or tons of pixels, etc.)
But it doesn't help on task that as mostly single threaded so it doesn't look nice on lots of benchmarks (e.g.: older games, lots of compression tasks).
In other words, Multithreading won't make your Windows go faster (Even more so as Windows has traditionally lagged behind the starte of the art process schedulers).
"Local optimum" as you say.
For instance, just because you store Game of Thrones in Postgres doesn't mean you are not infringing copyright by sharing that database.
In your example, the art assets (video content) is what is copyrighted (the Postgres is irrelevant. It's just a container, functionally equivalent to storing the stream in Matroska).
The DCMA takedown notice mentionned in the article doesn't list any copyrightable asset (no artworks, etc.)
Duplicating a collection of factoids isn't copyrightable.
And all the arguments of the notice are similar to "Waaah! Their table of spells looks just like ours back then!"
NOTE: Not the artwork of the spells (which is on the game disk legally bought by the players, not on the servers). Just the numbers giving the strength of the spells and maybe a few of their serial numbers.
Basically, what is reported to be duplicated looks like a "phone book of spells". And you definitely can't copyright a phone book in the US.
Feist Publication vs Rural Telephone Service seems a relevant comparison.
Also similarly, SCO Unix' arguments against Linux about content of *system header files* comes to mind.
If indeed there were no Blizzard copyrighted assets (think giant maps, artwork) but only general look-alike-ness of a couple of structures and a few numbers - as seems to be indicated by the cited DMCA takedown notice, this classic server emulator show have a solid case, even in the US.
- you need to replicate functionality in order to emulate in a compatible way.
- you can't copyright numbers
(- and you can't copy right a collection such as a database *by itself* (i.e.: without any further creative work) , e.g.: a phone book's content isn't copyrightable, though a nice illustration on its cover could be. the SQL designated in the DMCA take-down notice border on being a "phone book (-like) of objects".)
If you look in the URL linked in the summary (Yeah I know, /. and actual article reading) :
it might seem debatable.
The complains hinge around 2-3 sql file using names and having a few data that looks like the data used in wow (spells have the same characteristic as old versions of wow, same old trademarked names are used, etc.).
Fantasy Names – “Script” files and folders are named after and reference WoW fantasy names.
They're not complaining about game assets being lifted of blizzard's own software (e.g.: bitmaps, etc.)
They are complaining about the code using official Blizzard trademarked name to designate Blizzard's said trademarked characters.
(Note: e.g. it's not a trademark violation when you use microsoft's trademarked "Microsoft Windows" name to speak about Microsoft Windows itself).
They're complaining that the datamodel is very close to how it used to be in old servers :
The LightsHope spell table has identical layout and typically identical field names as the table from early WoW. We use database tables to represent game data, like spells, in WoW. In our code, we use .sql files to represent the data layout of each table (i.e. the fields of each specific table, like a spell name or the magnitude of its effect). MaNGOS, the platform off of which Light’s Hope appears to be built, uses a similar structure. The LightsHope spell_template table matches almost exactly the layout and field names of early WoW client database tables.
(Looks like the devs made their "Classic" recreation by using old dumps / backups as a referrence).
Matching Record IDs – There are “scripts” that reference database records directly by ID; there are cases where these IDs directly match the ID from WoW’s content.
"Hey, their serial numbers looks suspiciously close to our serial numbers !"
Oh, come one.
(Numbers aren't copyrightable in the US. That's basically Intel complaining that competitor's 386-compatible chip also use numbers like 386)
None of the complain is anything that looks like : "these huges chunks of code are actually a un-licensed copy of the network code of our server".
Overall : Some of the complain could almost fell under the "but these old numbers are necessary to get interoperation" exemption that exists to copyright in some jurisdiction (other /.ers have mentioned Canada as an example).
AMD considers their CPUs potentially vulnerable to Spectre Variant 2 - "Branch Target Injection". (The one were one attacker application is able to do it bidding into a completely different and innocent target application)
Some more recent AMD processors do indirect branch prediction.
But the way they do this indirect branch prediction is completely different.
Currently the Google demo code against Intel Xeon doesn't work (well, obviously).
Nobody has managed to write a successful exploit of that variant.
AMD engineer believe that it's a terribly difficult task that might not be doable.
So they might indeed end up being more or less Intel-specific-ish.
Sèectre variant 1 - "boundary check bypass" is the one affecting every CPU that those speculative execution and is basically "speculative execution working as it should/as documented, but now somebody has found a way to use it as a side-channel attack to have a software see its own data".
All the cells in all the multicellular organisms are cancerous.
For asexually reproducing organisms, there is nothing called cancer. The cells keep mutating and dividing. So what is cancer for them?
Nope. Not at all. Cancer is not simply cell division.
(Though there are some cell that do indeed not divide (e.g.: neuron, for obvious practical reasons) and whose population is repleted by progenitor cells (in that case, that would be neuroblast ; mostly happening in the amydalia region of the brain), there are other cell population were dividing cells are pretty much the norm (e.g.: cardiomyocytes in the heart do divide to replenish the population).)
Cancer is about complete uncontrolled cell divsion. Not only *unchecked* but utterly beyond any control or even coherent organisation.
Obligatory /. car metaphore:
- if cells dividing are cars driving forward
- then cancerous cells are cars witl the pedal's mechanism completely stuck, while in full throttle position. With all consequence that entail with that (including colliding everywhere)
A fully cancerous cell (once it has moved beyond hyperplasia toward full blown cancer ), will divide completely chaotically, even before it's actually ready to divide. It won't simply divide even when unneeded/unrequired by the body (that's "hyperplasia" and that's what you're thinking with your "revolution against the oppressor" logic)
It will divide even when it doesn't make any sense :
- Before having checked that the duplicated genetic material is correct (cancerous cells accumulate mutation at an alarming rate. On a global scale they are terribly inefficient : lots of them die just because they've completely destroyed their DNA. It's just that, on scale of division speed happening in a tumor, there are still a few that miraculously manage to be still semi functionnal enough to keep reproducing. It's survival of the fittest, but with the production of unfits turned up to eleven).
- Before even having correctly duplicated its genetic material (chromosomic aberration are abundant in a tumor, leading to lots of dysfunctional cell)
- Before even having accumulated enough resources to be functional. That's why you don't see cancer in single-cell organism : a cancerous cell isn't even able to function anymore, and require an organism on which to parasytically to rely in order to sustain. (A cancerous amoeba would be unable to eat and will die after a couple of divisions - cancerous cell are defined by their loss of function). (Just like the metaphorical car would run out of fuel pretty fast. From that point of view a cancer is closer to a trolley-bus : able to tap into the city's resources (electric grid) to still drive. And just like the cancer cell, a lot of the trolley bus would derail their electrical feed and die of)
A cancerous cell isn't simply "reverting to a pre-multicellular state". A cancerous cell is going batshit insane about division in a way that could not be survivable outside a multicellular organism.
When your autocorrect's auto trained dictionnary fixes "USD" as "USB", you know that you've been geeking way to much.
I think it's time that we categorize the "BTC to USB exchange rate" officially as a "strong random number generator" and call it a day ~
:-D
(NOTE: Jokes aside, that doesn't preclude that idea of the cryptocoins' decentralized protocols can be useful).
Are you also saying we as a society shouldn't try and shut down sketchy con-artist retailers because you're not stupid enough to fall for what they're selling and should be able to waste your money if you want to?
Or maybe, we should reconsider the education system a bit :
Maybe instead of blocking targets that are semi arbitrary labelled as "con artists" on the grounds of trying to shelter stupid people,
we should instead educate stupid people and teach them how to use their brain and do their due information search to not fall for cons/snake oil/conspiracy theory/etc. ?
Maybe if people just didn't forget how to use their brains, we'll have a lot less problems with people falling to stupid tricks.
Just saying...
On the other end, none of my media players (Xbox One, Nvidia Shield, Chromecast, AppleTV, etc) support it or my 4k projector (All support MPEG2/h.264/HEVC).
The idea of AV-1 creators, is to use GPGPU approaches for the hardware acceleration for the first devices supporting it.
So your Xbox, Shield and Chromecast will likely get the support added in a future OS upgrade, probably managing to get fluid play (but maybe at a higher power consumption than MPEG-AVC/H264).
After all, AV-1 is also going to be IETF's NetVC, and all these device boast web browsers.
Apple, being one of the HEVC licensor is probably going to drag their feet much longer.
HEVC is already entrenched in the UHD standard, so that isn't going away any time soon.
From being used in TV ? Indeed, it's going to stay because of various UHD DVB-xxx standards, UHD BlueRay, etc.
From being present in gadgets ? Well, you'll have to check the bargain bin at your local electronics store : not all dead-cheap no-name asian tablet feature actual hardware MPEG-HEVC/H265 acceleration.
That's actually a complain of the chipset manufacturer :
licenses fee for H265 are so high, that they become a significant fraction of the price of the crappy cheap SoC that are used in such tablets. (Hint: the SoC isn't the most expensive component in a crappy tablet. The screen and the battery are. The SoC are actually quite cheap. When a no-name table manufacturer shops for parts, they'll try to get the cheapest possible part, and H265-less SoCs are actually significantly cheaper at the volume that interests them).
I'm not making this up, it's actually been cited as a rationnal behing starting the whole Daala, AV-1, NetVC story.
So unless they can get the content creators (TV Shows and movies) to switch, then AV-1 will be a "web-only" technology unfortunately or we will have to wait for a UHD 2.0 standard.
Yes, for the close future, AV-1 / NetVC is going to be an internet-only technology (mostly web and apps), just like OPUS is currently (Digital Radio Mondial unofficial experiment non withstanding).
(Though I wouldn't be surprise if there are similar experiments in using AV-1 in some long range digital media showing eventually, as part of small scale local , experiments or in less wealthy region of the world).
But given amount of media consumed online, that's actually the lion's share of the audiovisual world.
By volume, *that* is what is going to be relevant.
There was a /. entry not so long ago that even in the US, the tipping point is being reach where more media is consumed over the net than over TV.
And again, like Vorbis before it, because it's free to implement, lots of no-name gadgets are going to implement it eventually, just to add another bullet point on their feature list, even if it's not going to be advertised on the stickers on the box.
Eventually, if TV standards are still relevant by then, it might show up in some UHD2.0 standard (DVB-T3... ?) but the whole question if TV is going to remain relevant by then remains.
It might even show up in some media descendant of BlueRay disc if such a thing still exists by then.
Streaming is probably going to be the most widespread media consumption mean, be it on some service provider's box (cable box, except relying more on IPv6 than on DVB-C), or with website / apps (think Youtube, Hulu, Netflix, etc.)
UCS has been evaluating & tracking how much mpg is needed to match an EV on a grid-level basis
http://blog.ucsusa.org/dave-re...
Thank you for the useful link.
The above AC would have noticed what difference in Sun vs. Moon eclipse means, and would have told you that you actually can't go blind from a lunar eclipse,
but he cannot read anymore because hE WENT BLIND! by attempting to use eyes during the last solar eclipse.