On the other hand : Linux is extremely pervasive on anything else except the desktop. Both ends of the scale.
Linux is extremely popular on cluster compute nodes. All the top 500 super computers exclusively run Linux. That some heavy power that would be very handy to hacker groups if they managed to get their hands on (lots of potential abuse, from tools to helps crack leaked password salted hashes, to mundane use like mining whatever crypto coin is currently CPU-only (or GPU-only if you managed to get your hands on a CUDA cluster), to stupid use like helping compressing pirates ultra-hidef movie releases...)
Linux is extremely popular also on small embed platforms usually with busybox (Modems, set-top boxes, smart TVs, IoT, Wifi fileserver embed in compact photo camera, etc.) or Android (smartphone, tablets, etc.) userspace. Every user bashing Linux for not having any significant market share, because they use Windows Laptops at home, is probably actually having at least a dozen of such invisible devices, that they completely overlooked. That's also lots of potential abuse, speially for D-DOS and other types of service disruption.
(And that's neglecting other corner uses, like Linux being popular on cars' infotainment center - though not necessary on the controller, where QNX is a bit more popular, even if some company like Tesla are Linux-all-the-way. Again incentive to find way to hack into the car's security)
In other words, black hats have actually tons of incentive to violently target Linux (and the various user spaces).
But - compared to the fractal-catastrophe of Microsoft's own Windows 10* - it seems to actually hold not that bad given all the above.
---
*: I'm not saying that Linux is magic pixie dust above everything in terms of security. It's not. On the other hand, it's at least some decent security, like other Unix. It's microsoft's... huh... "production", that can be considered "unfit to keep existing" security-wise.
Yeah. Russian company (Kaspersky), produces product (KAV) that removes non-russian malware (e.g.: WannaCry's NSA ancestor), but perhaps spies on the users, on behalf of Russian organisation (FSB, ex-KGB). American compagny (Microsoft), produces product (Security Essentials), that removes non-american malware, but very probably spies (Windows 10's cluster fuck of telemetry) on the users, on behalf of American organisation (NSA - see Snowden files).
And you could very likely be able to say the same about chinese software.
The only question is : what happens when you install several of them together ? How do they fight ? Will each product be able to neutralise the other one's spying ? Or will you machine be open to all of the above ?
So by watching such released things, you're supporting [failed] extortionists and encouraging them to continue to do this stuff.
Extorsion with a DRMed online media just can't possibly work.
All possible scenarios will all end up bringing nearly exact same result :
Scenario 1 : - company pays ransom - the pirates happen to be honest (for a certain definition of honest) and do actually destroy their copy. (- or more likely: they'll keep their pirated copy only for themselves and not leak it) - this copy of the movie / series season doesn't end up online. - meanwhile a totally unrelated data breach happens somewhere else, or somebody else finds a new way to break the DRM, or whatever. (- this done by a completely different and independent team of pirates) - by random chance this leak also contains the same movie or season. - that other group ends up putting it online (if the company is lucky, this might happen *after* the movie/series premiere).
End result : - company lost $ransom - media is leaked online a little bit later
Scenario 2: - company doesn't pay ransom. - pirate remain true to their words and do release media online.
End result : - media is leaked too, a bit faster - company at least didn't lose ransom-money.
No matter what, the movie/series *WILL* end up online. Like every single other movie/series has been for the past decade. Very few exceptions. The company paying or not won't change much, maybe delay the online release until another team does it, and it will essentially just cost money for not much.
Thus, the companies have absolutely ZERO incentive to pay ransom. No matter what, their movie/series will end up online. So at least it makes sense to save the money and the trouble.
--
Ransom with movie and series, would be like asking a ransom to avoid speeding : it's pointless. There are a lot of people speeding. Even if one driver gets paid and does actually follow his words, there will be tons of other speeders out there and the police can't manage to catch them all.
And soon Sony will be able to pull another Root-kit scandal, but this time it will be considered as legitimate defense against the evil pirates trying to hack them (and their DRM). Too bad if a few (= tons of) users got their machines nuked by the rootkit too, even if they never attempted to circumvent DRM. It's still allowed "hack-back"!
Nuke all the machines. Kill them all and let God sort them out.
A. check the title (not TFA, not even the summary, but the title). It's *Denmark* we're speaking about. They have invested so much into turbines that in 2015 wind alone accounted for 42% of its electricity needs.
They're *very far* from the US' over reliance on burning fossil. (As are several others among the european countries). (Don't trust all the propaganda coming out of your PresiDunce entourage about clean coal, or about ecology not being economically feasible)
B. this myth has been debunked over and over, just google it: even in countries like the US, despite its strong reliance on burning fossils for electricity, due to the much better yield of a big fixed plant, and even if you factor the manufacturing of the car (yes, batteries are bit more complex to produce than a ICE, but on the other hand, manufacturer is only a small part of the total production over the whole lifetime of the vehicle), in most countries electric cars such as Tesla *are still* worth it. The only exceptions are a few countries with the most extremely unclean energy production (of the top of my head: China, Inda and Australia, if I'm not mistaken. you can easily google for the results). So yeah, US Teslas and their batteries aren't as clean as Denmark's once you factor in energy production, but they're still cleaner than ICE.
C. Yes nuke is radioactive. But the yield of energy per kg of fuel is about millions time better than coal (exact number can be googled or found in wikipedia - around 2 millions I think). Or another way to put it : a single 1 kg bar of uranium will produce as much energy as a freight train loaded with 2'000 metric tons of coal. At that crazy difference in amount, even the traces of radioactive isotope in coal start to get significant: Yes, coal is a *lot* less radioactive than uranium, but you're burning such an insanely bigger amount of it, that it ends up releasing lots of radioactive ashes in the environment over its lifetime. And that only about radioactivity, it's completely ignoring the fact that the crazy amount of pollution you're emitting will cause tons of pulmonary and cardiac diseases, too. So nuclear incident might look frightening in the TV news flash (because it's sudden and concentrated in a small recognizable region), but coal is the actual big silent insidious killer.
Either further tests by researchers will prove that the perceived effect was a mistake or due to something besides the blood itself,
In short : - Blood tranfusion from blood banks is really only red blood cells. And almost nothing more. - Hooking two mices together, is way much more than that.
Out of the top of my head: - Means that the old mouse's blood is processed by the organs of the younger one : kidneys, liver - organs which are in charge of eliminating/chemically processing toxins.
- There's more that simply red cells that the old mouse is getting from the young one : platelets, white blood cells and antibodies (ie.: imune system), hormones and growth factors (the original study mentions quite a few of them), etc.
Oh, and - mentionned in TFA : these results haven't been replicated successfully yet.
Disclaimer: I*A*AMD, I just don't dabble into casual vampirism.
Whenever some people are really discriminated against, you find others that are just trying to get a free ride on this. Pretty bad.
The free rider are pretty bad indeed. Even more so, because they contribute to reduce the trust in actual victims.
The couple of stupid women claiming "rape" just to get some money, will make it all more difficult for all the *actual real* rape victims out-there to speak, because the victims will fear they won't be believed.
It's a sort of Girl who cried wolf, except that the consequences of "excessive wolf-crying" will fall on someone else.
SecureBoot has been since long worked around. Most distributions use signed shim (signed with the official Microsoft keys) which can then in turn whatever they want (UEFI SecureBoot -> shim -> grub -> Linux or whatever else you picked from grub's menu).
Drivers ? Meh...
You'll probably be stuck with some old 3.xx kernel version that Qualcom provides to manufacturer of smartphones that use the same chipset. and/or You'll need to use libhybris to get the drivers (normally designed by Qualcom for android) to be useable in a full blown GNU/Linux distribution. (This laptop will be similar to the efforts of getting any full blown GNU/Linux (Jolla's Sailfish OS, Canonical's Ubuntu Touch, etc.) ported to smartphones/tablets).
With some luck, if this chip gets popular enough, there will be efforts to support it in vanilla kernels, but then you'd only could use Mesa's Freedreno for driving its GPU.
At least the "SecureBoot" part of the equation is more or less solved.
Most distro can use signed shims to chain into their bootloaders (eg.: UEFI SecureBoot -> shim ->.grub -> Linux kernel or whatever you pick from the grub menu).
driver on the other hand..
You'll probably be still stuck with a very old 3.xx kernel that qualcom provides to manufacturer of smartphones and tablets built around the same chipset. and/or you'll need to use libhybris to leverage the android drivers on a normal full blown GNU/Linux.
So it's not going to be trivial.
On the other hand, if that chips get realy widespread, you might end up seeing support in vanilla kernel (but would probably need then to stick to Mesa's Freedreno driver for the GPU).
Other than asking if this hardware will run Linux (I know the basic answer is "yes", but I would like to see the network driver release plan for Linux)
It's not so sure: It's a Qualcom chip.
If cyanogen/lineageos has taught us anything, it's that these chips have some weird design. The summary says:
Snapdragon 835 has a higher level of integration than Intel's mobile chips,
That sometimes means that on some chips, the wireless modem is the CPU's northbridge. i.e.: the device that - by laws (specifically laws around licensed frequencies) - runs 3rd party closed binary firmware, some of which gets automatically updated by the service provider, is in charge of bringing up the RAM, and has uncomfortably lots of access everywhere. (e.g: HP Pre3 has such a configuration. Meaning that some firmware automatic updates could crash the whole phone).
You can also expect that the chip *will* have some linux support (because it's going to be used on countless android smartphones) but you'll have to stick to a specific kernel version provided by qualcomm (probably even a 3.xx). And you'll probably need to use Jolla's "libhybris" adaptation layer to leverage the rest of the drivers and libraries as you want to use a full blown GNU/Linux layer above the kernel instead of the Android stack for which said kernel was devised. (See the porting of any GNU/Linux stack (Jolla's Sailfish OS, Canonical's Ubuntu Touch) on other smartphone).
Though if the chip gets popular enough, you'll probably eventually see support for it in vanilla kernels. (Though you might need to fall back to Mesa Freedeno if you want to use the GPU on a kernel different than that provided by Qualcom to smartphone manufacturer).
Also, because it's a non x86 chip, you can bet that microsoft will be using UEFI as a firmware and only in signed/secure mode, tough that one has been tackled with signed shims for quite some time.
So in short : Linux is technically possible, but it's going to be as cumbersome as on any smartphone using the same chips as this laptop. In other words: You're better off getting a Pyra.
I have to wonder about Microsoft pushing a Snapdragon solution in terms of apps.
Yup. Microsoft will indeed probably port some of their code base, but... legacy x86 applications are going to run as crap due to slow emulation (and some even break the emulator).
And forget about games. Emulation wouldn't be useful most of the time, and WSL already failed as an attempt to open Windows to the android apps & games ecosystem. (but at least brought us bash in windows)
Which renders the whole purpose of having Windows as an OS moot.
You'd better hope for Google to make Chromebooks around the same platform. Those at least are going to be easier to hack into full blown GNU/Linux.
Indeed, there could be an incentive to plant false proof just for the seek of getting the money.
Then there are other problems: - The files might not have been his. (2nd hand drive ? Not very likely on a doctor's salary, but still...)
- It's a *Medical doctor* bringing his machine for repairs. There might be information falling under protection of medical data. (e.g.: the doctor could have been in middle of work, when his computer crashed, and not have been able to sanitize it, before bringing it for repairs). By scanning beyond what it needed to fix the computer, they might discover patient private data, which they shouldn't.
- It's a *gynecologist*. He might have post-puberty, not yet adult, teen patients. The geek squad might accidentally find things that they consider "child pornography" (pictures of pussy belonging to a girl clearly under 21), whereas the Doctor might be documenting the evolution under treatment of a very weird rash of one of his younger patients (it's private patient data that nobody outside the said patient and his doctor should ever access) (or there are even less happy reasons for a gynecologist to store picture documentation).
- etc.
Also: regarding FBI's behaviour.
the FBI is completely idiotic in actually using this in court. At worst, they should parallel construct : consider this only as a tip attracting attention to some potential problem instead of directly acting on it, and do some of their own police work until they gather enough *legally obtained* information that confirms above suspictions. They build the entire case, even request for warrants, out of the legally obtained information. They only use the paid informant's tips as a suggestion in which direction to look to.
They double-extra won't ever give it to you with a hypervisor which gives access to all resources, but the truth is that Other OS made it too easy to rip Blu-Ray.
How so ?
I vaguely understand the logic behind the game piracy scare : - apparently some frustrated tinkerer found an unofficial way around the hypervisor limitations, and that would in theory open up way to devise solution to run commercial games under OtherOS without the copy protection kicking in.
But how would a PS3 running OS help ripping Blurays differently than a USB3 Bluray drive connected to a laptop ? Or did you meant "ripping the encryption key out of the player's RAM" ? (something along the way of : play Bluray in console, then switch to otherOS, dump the memory, search the memory for the AACS key)
Disclaimer : I'm a bit out of touch with the whole "ripping" thing, since I last did backups of important DVD with K9Copy and DVDShrink.
Why even bother? At this point for most practical purposes something like a Raspberry Pi is a much better hardware.
Why is there homebrew community for DreamCast, NeoGeo, and even older (Megadrive, etc.) consoles when a modern Pi will pack more power than any of them ? Because some people have fun tinkering with hardware. Some people actually enjoy the intellectual challenge on trying to make something work on constrained hardware. (See the demoscene on very old machines: 8-bits computers like C64, original IBM PC, etc.)
OtherOS was shut down by Sony on the grounds of helping piracy. Now that they won't produce anymore official content, there's less impact of potential piracy and would make the home-brew/tinkerer communities happier.
But as mentioned by others : fat chance. Sony seems to be as much anti-opening as possible.
Speaking of which, it would be nice if their last patch before ditching the console did re-enable the "other os" option, this time with the hyper visor officially giving access to all resources.
OK... but there is no way it's running ASR and NLU locally on a device of this form factor.
I would say, it depends on the expected performances. There a difference between an (cloud-based) AI that can listen and answer nearly in realtime. And an AI that react slower, requires a simpler vocabulary, etc. (but thusly works even if the connection is down). With the advance in moore's law, the latest gen of small form-factor hardware might be able to run locally some significant deep neural-nets.
It's why I still can't take my water bottle on any flight despite there having never in the history of aviation ever been a credible threat related to liquids.
There *IS* a credible threat related to liquids...
...a threat to the profits of the businesses selling liquids at a steep price on the other side of the security checks.
The Army used to be made of REAL MEN. It's a disgrace what its become. Could you imagine someone like Chelsea Manning storming the beaches of Normandy?
Actually, speaking of D-day, there wren't only REAL MEN storming on the beaches of Normany - e.g.: Martha Ghellhorn, Ernest Hemingway's ex-wife (though a natural-born woman, but still definitely not a REAL MAN) managed to be among the first waves on the beach (even before her ex-husband) by first hiding on a boat and then disguising as a combat medic (though her actual profession was war journalist). There *WAS DEFINITELY* a pair of boobs under one of the uniforms running on the beaches of Normandy.
But indeed, the bitmasking capability of older WRAM and SGRAM have been made redundant by the much more general-purpose capabilites offered by the GPU coupled with the much more complex modern interface. (i.e.: It's opengl running your Linux Compiz / Apple Quartz and whatever was the windows equivalent).
Those where great processors for the money that you paid for them. I believe it used Pentium pro instructions instead of the pure Pentium.
According to wikipedia : - the Cyrix MII - was Pentium Pro / Pentium II compatible, as you mention. - before that : Cyrix 6x86 MX - was Pentium MMX compatible. - even before : the previous Cyrix 6x86 & 8x86L - were more or less, but not entirely, Pentium compatible. (They officially identified themselves as "486") (I remember that bit)
Also: - their FPU was less optimized, because most of the typical software workload was integer back then. (Also rings a bell)
For a linux server they would very likely have been quite descent, because : - FPU is indeed irrelevant. - GCC and Linux kernel *do handle* 6x86 (it's not considered as a pure 486, they can use the pentium-compatible parts).
But not as bad as described. Again, PCs contemporary to the Amiga where able to 16 colors palette (still less than Amiga's 32 colours, but more than the 4 reported by Ars).
On my A1000 I could listen to mods, while copying files, and writing a term paper at the same time all in a graphical, full-color windowed user interface. No stuttering between windows and apps.
Which has more to do with the very nice AmigaOS with very nice support for multi-tasking. Meanwhile on PCs (contemporary processor : Intel 286 on PC/AT) you had to use DOS (single-tasking) or Windows (not that brillant ad multi-tasking).
Though CPU helped a bit : motorola 68k 32bits/16bits hybrid featured in Amiga were quite a beast. You had to wait until the 386sx to get similar capabilities (and no OS that good at multi-tasking).
And depending on the video mode the Amiga was capable of 32,64, or 4096 colors. Orders of magnitude more than the PC.
Yes, the first generation of Amiga was able to do 32 colors (out of a large palette) Yes, that was better than the PC.
I'm just signaling that the PC was already at 16 colors, not 4 colors as mentionned by ars.
All of this changed fairly quickly once soundblaster came along as well as the video card manufacturers stepped up there game.
Note that regarding digital audio playback, soundblaster only really added DMA to the mix. Still the capability to play a single digital channel (as the original PC speaker did), but entirely automatically, without the main CPU needing to work much. (Also, additionally it had full AdLib compatibility for multiple synth voices).
Feature parity with Amiga (capability to play multiple digital channels) only came with Gravis Ultra Sound / Sound Blaster 32 AWE. Before that, digital audio was mixed in software. (e.g.: all modplayer on PC) but given the ever increasing power in PC CPUs, that was acually doable.
Once windows 95 and 3-d video cards emerged 10 years later, the Amiga's fate was sealed.
I would think it came a little bit earlier. Amiga could count on its co-processors, PC tended to count on the raw processing power.
Amiga could easily do flat shaded 3D thanks to its coprocessor (which could offload the drawing of polygons).
On the other hand, the power CPU found on PC could do textured environment. Thus enabling easily 3D games such as wolfenstein 3D and then Doom, or games with pre-rendered 3D models like Commander (there were attemps at ports of WC on Amiga, but with sucky performance). Now move further to Quake and it starts to be beyond the reach of the contemporary amiga.
As soon as game designers decided to push 3D beyond the classic flat shaded polygons, strong CPUs were required.
Someone old enough to remember the original PC here. CGA was not a standard feature; the base model (which most people had) used something called MDA (Monochrome Display Adapter), which supported text output only without any pixel-addressable graphics. {...} Almost nobody got CGA on the PC.
Fun because on this side of the ocean, CGA was quite widespread, and as soon as EGA was realeased by IBM, we started also seeing it here around. (Maybe differences in buying power between countries ?)
The thing which *was* rarer here was composite output : local colors standards where PAL and SECAM (depending on country, but most TV where dual standard). Not that many TV monitors did support NTSC too, meaning that TV-output wasn't trivial with the CGA's composite out. So most often it was either monochrome monitors or expensive RGB monitors.
When Lotus 1-2-3 came along, the need for plotting drove the adoption of a proprietary monochrome graphics technology called "Hercules Graphics Adapter", which was much, much more popular than CGA ever was.
My father's business did use Hecrules graphic cars because of that plotting (though Hercules had hardware-assisted emulation of CGA, rendering the low-resolution 320x200 4 colors as high resolution halftones to simulat grayscale levels). Our schools was CGA (some with even expensive color RGB monitors), then rather early EGA.
As a side note, even then sound wasn't a standard feature on personal computers. The most they could do was beep. You had to add a proprietary sound card to get anything more.
Nope. Again : the PC speaker had PWM capabilities. It wasn't used in *games* (At least not before Access on much more powerful computers), because the PC Speaker didn't have DMA, and the CPU needed to push samples manually in software.
But it was used in a few novelty tools (small DOS program wich could synthetize speech by stringing together pre-recorded words). And a few games used it for music during title screen (when the CPU isn't busy by the game itself).
But yes, PC sound only really started taking of in 1987 with the apparition with sound cards (AdLib, CMS, IBM MF, etc.) (Though not everything was proprietary. In 1984 already the PC could do the openstandard MIDI with a specific card. But it didn't get used in games until MT-32 become popular, again by Sierra Online, starting with SCI text games, first used by King's Quest IV )
This is why Macintosh became common in schools. As a developer you could count on every Mac having the same set of very rudimentary capabilities. As a school administrator, you just unboxed the thing and fed it floppies; there was no opening the case and installing optional boards that had to have their address and interrupt vectors chosen by the user and configured by jumpers.
Different countries, different situation. Here around IBM PC were nearly as popular. Probably because of the reputation of the magic "three letters".
There's more to CGA than that too. When used with a composite video output games could exploit the namesake of NTSC (that is Never Twice the Same Colour) to create artefacts that actually produced quite fantastic graphics (compared to 4 colour CGA)
Yup, one of the official methods to get more colors relied on the NTSC color clash. This one was massively used by Sierra Online (lots of their CGA games used 16 colors on NTSC composite out - nearly all of the AGI text adventure game did)
The other one (working with RGB monitors too) and more or less officially documented, relied on the text mode. The CGA is switched in 80x25 mode (640x200 pixels). Text lines are compressed vertically until each caracter line covers only 1 or 2 pixels. (displaying 100 or 200 text of lines per screen - at least the top of each line as only the top 1 or 2 pixels of each caracter cell are display). There are too characters in the table - 221 and 222 - that are half block. e.g.: the left half of the caracter cell is foreground, the right half of the cell is background. By playing with that one can make a 160x100 (more frequently, because square pixels) or 160x200 (highest resolution, but single page) with colors. (Technically it's 80x100 or 80x200 text modes, with each cell split in 2 halfs one with the foreground the other with the background color of the char cell).
with a wide range of colours in 320x240 mode with 16 possible colours on the screen at once out of a palate of 64 possible colours (in 4 groups)
BTW: it's 320x200, that the most vertical scanline you could output backthen.
This time, the technique is to turn on the 640x200 monochrome CGA mode. BUT to activate the colour output on the composite output (can be make-shift hacked by setting a specific border colours which will be interpreted as the NTSC color burst), so a composite monitor will interpret it as color signal. Then output specific black-white pattern (on-off pixels) on the monochrome screen, they will be interpreted by the NTSC as colors. That more or less gives 16 colors to play with. (but at a 160 horizontal resolution) That's how nearly all AGI text adventure games of Sierra did work.
Alternative consisted of using the various palettes of 320x200 4 colors CGA, and the horizontal bluring of colors. (remember : NTSC offers only 160 horizontal color resolution), you have 16 ways to combine two 4 colors neight boors in composite mode, giving you 16 colors. With a different set of 16 colors, depending on the CGA color palette selected.
The only problem is single pixels with contrast got blurred which made reading text very difficult in these modes.
Blur brought the resolution down to 160x200. but this "blur" is where the "color mixing" enabling 16 colours comes from.
Sierra solved it by having a 160x200 16 color playfield, but use only monochrome (black-on white) text to increase readability, and do all the critical text-only steps (like inventory management) in monochrome texte-mode instead of graphics.
Then, there's the modern "8088mph" demo. It combines both above approaches.
It creates a 80x200 text modes, and use all possible patterns (including also the alternating pixels patterns, "!!" ascii character, etc.) together with all possible foreground/background colors combination, to create thousands of different outputs on a composite monitor.
I've found WinSCP to be better than FileZilla especially since so many providers offer SFTP now anyway.
Note that Filezilla support SFTP too.
I don't store my passwords so the master password thing is not an issue to me. Don't store passwords if you don't want them to be found.
Even better : don't use passwords. Use Public Keys pairs.
(Filezilla supports them, and can use Putty's key agent to handle them) (I'm sure that WinSCP can too, just didn't bother to check).
Best part : you can then completely switch off the support for password on the SSH/SFTP server. Your server is then (obviously) immune to brute force / password guessing.
Slashdot Mirror
On the other hand :
Linux is extremely pervasive on anything else except the desktop.
Both ends of the scale.
Linux is extremely popular on cluster compute nodes. All the top 500 super computers exclusively run Linux.
That some heavy power that would be very handy to hacker groups if they managed to get their hands on
(lots of potential abuse, from tools to helps crack leaked password salted hashes, to mundane use like mining whatever crypto coin is currently CPU-only (or GPU-only if you managed to get your hands on a CUDA cluster), to stupid use like helping compressing pirates ultra-hidef movie releases...)
Linux is extremely popular also on small embed platforms usually with busybox (Modems, set-top boxes, smart TVs, IoT, Wifi fileserver embed in compact photo camera, etc.) or Android (smartphone, tablets, etc.) userspace.
Every user bashing Linux for not having any significant market share, because they use Windows Laptops at home, is probably actually having at least a dozen of such invisible devices, that they completely overlooked.
That's also lots of potential abuse, speially for D-DOS and other types of service disruption.
(And that's neglecting other corner uses, like Linux being popular on cars' infotainment center - though not necessary on the controller, where QNX is a bit more popular, even if some company like Tesla are Linux-all-the-way. Again incentive to find way to hack into the car's security)
In other words, black hats have actually tons of incentive to violently target Linux (and the various user spaces).
But - compared to the fractal-catastrophe of Microsoft's own Windows 10* - it seems to actually hold not that bad given all the above.
---
*: I'm not saying that Linux is magic pixie dust above everything in terms of security. It's not.
On the other hand, it's at least some decent security, like other Unix.
It's microsoft's... huh... "production", that can be considered "unfit to keep existing" security-wise.
Yeah.
Russian company (Kaspersky), produces product (KAV) that removes non-russian malware (e.g.: WannaCry's NSA ancestor), but perhaps spies on the users, on behalf of Russian organisation (FSB, ex-KGB).
American compagny (Microsoft), produces product (Security Essentials), that removes non-american malware, but very probably spies (Windows 10's cluster fuck of telemetry) on the users, on behalf of American organisation (NSA - see Snowden files).
And you could very likely be able to say the same about chinese software.
The only question is :
what happens when you install several of them together ? How do they fight ?
Will each product be able to neutralise the other one's spying ?
Or will you machine be open to all of the above ?
So by watching such released things, you're supporting [failed] extortionists and encouraging them to continue to do this stuff.
Extorsion with a DRMed online media just can't possibly work.
All possible scenarios will all end up bringing nearly exact same result :
Scenario 1 :
- company pays ransom
- the pirates happen to be honest (for a certain definition of honest) and do actually destroy their copy.
(- or more likely: they'll keep their pirated copy only for themselves and not leak it)
- this copy of the movie / series season doesn't end up online.
- meanwhile a totally unrelated data breach happens somewhere else, or somebody else finds a new way to break the DRM, or whatever.
(- this done by a completely different and independent team of pirates)
- by random chance this leak also contains the same movie or season.
- that other group ends up putting it online (if the company is lucky, this might happen *after* the movie/series premiere).
End result :
- company lost $ransom
- media is leaked online a little bit later
Scenario 2:
- company doesn't pay ransom.
- pirate remain true to their words and do release media online.
End result :
- media is leaked too, a bit faster
- company at least didn't lose ransom-money.
No matter what, the movie/series *WILL* end up online.
Like every single other movie/series has been for the past decade.
Very few exceptions.
The company paying or not won't change much, maybe delay the online release until another team does it, and it will essentially just cost money for not much.
Thus, the companies have absolutely ZERO incentive to pay ransom.
No matter what, their movie/series will end up online. So at least it makes sense to save the money and the trouble.
--
Ransom with movie and series, would be like asking a ransom to avoid speeding : it's pointless.
There are a lot of people speeding. Even if one driver gets paid and does actually follow his words, there will be tons of other speeders out there and the police can't manage to catch them all.
And soon Sony will be able to pull another Root-kit scandal, but this time it will be considered as legitimate defense against the evil pirates trying to hack them (and their DRM).
Too bad if a few (= tons of) users got their machines nuked by the rootkit too, even if they never attempted to circumvent DRM.
It's still allowed "hack-back"!
Nuke all the machines.
Kill them all and let God sort them out.
So, if workshation-mode gives us peak performance and reliability, then what the hell are we receiving now?
Peak performance and reliability too... for the buyer to whom they are selling all the telemetry data.
Meanwhile, movie industry starts planning to make 2 trilogies of 150min films out of this book.
Three things :
A. check the title (not TFA, not even the summary, but the title).
It's *Denmark* we're speaking about.
They have invested so much into turbines that in 2015 wind alone accounted for 42% of its electricity needs.
They're *very far* from the US' over reliance on burning fossil. (As are several others among the european countries).
(Don't trust all the propaganda coming out of your PresiDunce entourage about clean coal, or about ecology not being economically feasible)
B. this myth has been debunked over and over, just google it:
even in countries like the US,
despite its strong reliance on burning fossils for electricity,
due to the much better yield of a big fixed plant,
and even if you factor the manufacturing of the car (yes, batteries are bit more complex to produce than a ICE, but on the other hand, manufacturer is only a small part of the total production over the whole lifetime of the vehicle),
in most countries electric cars such as Tesla *are still* worth it.
The only exceptions are a few countries with the most extremely unclean energy production (of the top of my head: China, Inda and Australia, if I'm not mistaken. you can easily google for the results).
So yeah, US Teslas and their batteries aren't as clean as Denmark's once you factor in energy production, but they're still cleaner than ICE.
C. Yes nuke is radioactive.
But the yield of energy per kg of fuel is about millions time better than coal (exact number can be googled or found in wikipedia - around 2 millions I think).
Or another way to put it : a single 1 kg bar of uranium will produce as much energy as a freight train loaded with 2'000 metric tons of coal.
At that crazy difference in amount, even the traces of radioactive isotope in coal start to get significant: Yes, coal is a *lot* less radioactive than uranium, but you're burning such an insanely bigger amount of it, that it ends up releasing lots of radioactive ashes in the environment over its lifetime.
And that only about radioactivity, it's completely ignoring the fact that the crazy amount of pollution you're emitting will cause tons of pulmonary and cardiac diseases, too.
So nuclear incident might look frightening in the TV news flash (because it's sudden and concentrated in a small recognizable region), but coal is the actual big silent insidious killer.
Either further tests by researchers will prove that the perceived effect was a mistake or due to something besides the blood itself,
In short :
- Blood tranfusion from blood banks is really only red blood cells. And almost nothing more.
- Hooking two mices together, is way much more than that.
Out of the top of my head:
- Means that the old mouse's blood is processed by the organs of the younger one :
kidneys, liver - organs which are in charge of eliminating/chemically processing toxins.
- There's more that simply red cells that the old mouse is getting from the young one :
platelets, white blood cells and antibodies (ie.: imune system),
hormones and growth factors (the original study mentions quite a few of them),
etc.
Oh, and - mentionned in TFA :
these results haven't been replicated successfully yet.
Disclaimer: I*A*AMD, I just don't dabble into casual vampirism.
Whenever some people are really discriminated against, you find others that are just trying to get a free ride on this. Pretty bad.
The free rider are pretty bad indeed. Even more so, because they contribute to reduce the trust in actual victims.
The couple of stupid women claiming "rape" just to get some money, will make it all more difficult for all the *actual real* rape victims out-there to speak, because the victims will fear they won't be believed.
It's a sort of Girl who cried wolf, except that the consequences of "excessive wolf-crying" will fall on someone else.
Regarding DRM ?
Yes.
SecureBoot has been since long worked around. Most distributions use signed shim (signed with the official Microsoft keys) which can then in turn whatever they want (UEFI SecureBoot -> shim -> grub -> Linux or whatever else you picked from grub's menu).
Drivers ?
Meh...
You'll probably be stuck with some old 3.xx kernel version that Qualcom provides to manufacturer of smartphones that use the same chipset.
and/or
You'll need to use libhybris to get the drivers (normally designed by Qualcom for android) to be useable in a full blown GNU/Linux distribution.
(This laptop will be similar to the efforts of getting any full blown GNU/Linux (Jolla's Sailfish OS, Canonical's Ubuntu Touch, etc.) ported to smartphones/tablets).
With some luck, if this chip gets popular enough, there will be efforts to support it in vanilla kernels, but then you'd only could use Mesa's Freedreno for driving its GPU.
At least the "SecureBoot" part of the equation is more or less solved.
Most distro can use signed shims to chain into their bootloaders (eg.: UEFI SecureBoot -> shim -> .grub -> Linux kernel or whatever you pick from the grub menu).
driver on the other hand..
You'll probably be still stuck with a very old 3.xx kernel that qualcom provides to manufacturer of smartphones and tablets built around the same chipset.
and/or
you'll need to use libhybris to leverage the android drivers on a normal full blown GNU/Linux.
So it's not going to be trivial.
On the other hand, if that chips get realy widespread, you might end up seeing support in vanilla kernel (but would probably need then to stick to Mesa's Freedreno driver for the GPU).
Other than asking if this hardware will run Linux (I know the basic answer is "yes", but I would like to see the network driver release plan for Linux)
It's not so sure: It's a Qualcom chip.
If cyanogen/lineageos has taught us anything, it's that these chips have some weird design. The summary says :
That sometimes means that on some chips, the wireless modem is the CPU's northbridge.
i.e.: the device that - by laws (specifically laws around licensed frequencies) - runs 3rd party closed binary firmware, some of which gets automatically updated by the service provider, is in charge of bringing up the RAM, and has uncomfortably lots of access everywhere.
(e.g: HP Pre3 has such a configuration. Meaning that some firmware automatic updates could crash the whole phone).
You can also expect that the chip *will* have some linux support (because it's going to be used on countless android smartphones) but you'll have to stick to a specific kernel version provided by qualcomm (probably even a 3.xx).
And you'll probably need to use Jolla's "libhybris" adaptation layer to leverage the rest of the drivers and libraries as you want to use a full blown GNU/Linux layer above the kernel instead of the Android stack for which said kernel was devised.
(See the porting of any GNU/Linux stack (Jolla's Sailfish OS, Canonical's Ubuntu Touch) on other smartphone).
Though if the chip gets popular enough, you'll probably eventually see support for it in vanilla kernels.
(Though you might need to fall back to Mesa Freedeno if you want to use the GPU on a kernel different than that provided by Qualcom to smartphone manufacturer).
Also, because it's a non x86 chip, you can bet that microsoft will be using UEFI as a firmware and only in signed/secure mode, tough that one has been tackled with signed shims for quite some time.
So in short :
Linux is technically possible, but it's going to be as cumbersome as on any smartphone using the same chips as this laptop.
In other words: You're better off getting a Pyra.
I have to wonder about Microsoft pushing a Snapdragon solution in terms of apps.
Yup. Microsoft will indeed probably port some of their code base, but...
legacy x86 applications are going to run as crap due to slow emulation (and some even break the emulator).
And forget about games. Emulation wouldn't be useful most of the time, and WSL already failed as an attempt to open Windows to the android apps & games ecosystem. (but at least brought us bash in windows)
Which renders the whole purpose of having Windows as an OS moot.
You'd better hope for Google to make Chromebooks around the same platform.
Those at least are going to be easier to hack into full blown GNU/Linux.
Indeed, there could be an incentive to plant false proof just for the seek of getting the money.
Then there are other problems:
- The files might not have been his. (2nd hand drive ? Not very likely on a doctor's salary, but still...)
- It's a *Medical doctor* bringing his machine for repairs.
There might be information falling under protection of medical data.
(e.g.: the doctor could have been in middle of work, when his computer crashed, and not have been able to sanitize it, before bringing it for repairs).
By scanning beyond what it needed to fix the computer, they might discover patient private data, which they shouldn't.
- It's a *gynecologist*. He might have post-puberty, not yet adult, teen patients.
The geek squad might accidentally find things that they consider "child pornography" (pictures of pussy belonging to a girl clearly under 21),
whereas the Doctor might be documenting the evolution under treatment of a very weird rash of one of his younger patients (it's private patient data that nobody outside the said patient and his doctor should ever access) (or there are even less happy reasons for a gynecologist to store picture documentation).
- etc.
Also: regarding FBI's behaviour.
the FBI is completely idiotic in actually using this in court.
At worst, they should parallel construct : consider this only as a tip attracting attention to some potential problem instead of directly acting on it, and do some of their own police work until they gather enough *legally obtained* information that confirms above suspictions. They build the entire case, even request for warrants, out of the legally obtained information. They only use the paid informant's tips as a suggestion in which direction to look to.
They double-extra won't ever give it to you with a hypervisor which gives access to all resources, but the truth is that Other OS made it too easy to rip Blu-Ray.
How so ?
I vaguely understand the logic behind the game piracy scare :
- apparently some frustrated tinkerer found an unofficial way around the hypervisor limitations, and that would in theory open up way to devise solution to run commercial games under OtherOS without the copy protection kicking in.
But how would a PS3 running OS help ripping Blurays differently than a USB3 Bluray drive connected to a laptop ?
Or did you meant "ripping the encryption key out of the player's RAM" ? (something along the way of : play Bluray in console, then switch to otherOS, dump the memory, search the memory for the AACS key)
Disclaimer : I'm a bit out of touch with the whole "ripping" thing, since I last did backups of important DVD with K9Copy and DVDShrink.
Why even bother? At this point for most practical purposes something like a Raspberry Pi is a much better hardware.
Why is there homebrew community for DreamCast, NeoGeo, and even older (Megadrive, etc.) consoles when a modern Pi will pack more power than any of them ?
Because some people have fun tinkering with hardware.
Some people actually enjoy the intellectual challenge on trying to make something work on constrained hardware.
(See the demoscene on very old machines: 8-bits computers like C64, original IBM PC, etc.)
OtherOS was shut down by Sony on the grounds of helping piracy.
Now that they won't produce anymore official content, there's less impact of potential piracy and would make the home-brew/tinkerer communities happier.
But as mentioned by others : fat chance. Sony seems to be as much anti-opening as possible.
Speaking of which, it would be nice if their last patch before ditching the console did re-enable the "other os" option, this time with the hyper visor officially giving access to all resources.
OK... but there is no way it's running ASR and NLU locally on a device of this form factor.
I would say, it depends on the expected performances.
There a difference between an (cloud-based) AI that can listen and answer nearly in realtime.
And an AI that react slower, requires a simpler vocabulary, etc.
(but thusly works even if the connection is down).
With the advance in moore's law, the latest gen of small form-factor hardware might be able to run locally some significant deep neural-nets.
It's why I still can't take my water bottle on any flight despite there having never in the history of aviation ever been a credible threat related to liquids.
There *IS* a credible threat related to liquids...
...a threat to the profits of the businesses selling liquids at a steep price on the other side of the security checks.
The Army used to be made of REAL MEN. It's a disgrace what its become. Could you imagine someone like Chelsea Manning storming the beaches of Normandy?
Actually, speaking of D-day, there wren't only REAL MEN storming on the beaches of Normany - e.g.: Martha Ghellhorn, Ernest Hemingway's ex-wife (though a natural-born woman, but still definitely not a REAL MAN) managed to be among the first waves on the beach (even before her ex-husband) by first hiding on a boat and then disguising as a combat medic (though her actual profession was war journalist).
There *WAS DEFINITELY* a pair of boobs under one of the uniforms running on the beaches of Normandy.
But then everything moved onto GPU's.
Modern GDDR retains the capability to clear buffers by itself.
But indeed, the bitmasking capability of older WRAM and SGRAM have been made redundant by the much more general-purpose capabilites offered by the GPU coupled with the much more complex modern interface. (i.e.: It's opengl running your Linux Compiz / Apple Quartz and whatever was the windows equivalent).
Those where great processors for the money that you paid for them. I believe it used Pentium pro instructions instead of the pure Pentium.
According to wikipedia :
- the Cyrix MII - was Pentium Pro / Pentium II compatible, as you mention.
- before that : Cyrix 6x86 MX - was Pentium MMX compatible.
- even before : the previous Cyrix 6x86 & 8x86L - were more or less, but not entirely, Pentium compatible. (They officially identified themselves as "486") (I remember that bit)
Also:
- their FPU was less optimized, because most of the typical software workload was integer back then. (Also rings a bell)
For a linux server they would very likely have been quite descent, because :
- FPU is indeed irrelevant.
- GCC and Linux kernel *do handle* 6x86 (it's not considered as a pure 486, they can use the pentium-compatible parts).
PC's were as shitty as that...
Worse than Amiga ? yes.
But not as bad as described.
Again, PCs contemporary to the Amiga where able to 16 colors palette (still less than Amiga's 32 colours, but more than the 4 reported by Ars).
On my A1000 I could listen to mods, while copying files, and writing a term paper at the same time all in a graphical, full-color windowed user interface. No stuttering between windows and apps.
Which has more to do with the very nice AmigaOS with very nice support for multi-tasking.
Meanwhile on PCs (contemporary processor : Intel 286 on PC/AT) you had to use DOS (single-tasking) or Windows (not that brillant ad multi-tasking).
Though CPU helped a bit :
motorola 68k 32bits/16bits hybrid featured in Amiga were quite a beast.
You had to wait until the 386sx to get similar capabilities (and no OS that good at multi-tasking).
And depending on the video mode the Amiga was capable of 32,64, or 4096 colors. Orders of magnitude more than the PC.
Yes, the first generation of Amiga was able to do 32 colors (out of a large palette)
Yes, that was better than the PC.
I'm just signaling that the PC was already at 16 colors, not 4 colors as mentionned by ars.
All of this changed fairly quickly once soundblaster came along as well as the video card manufacturers stepped up there game.
Note that regarding digital audio playback, soundblaster only really added DMA to the mix. Still the capability to play a single digital channel (as the original PC speaker did), but entirely automatically, without the main CPU needing to work much.
(Also, additionally it had full AdLib compatibility for multiple synth voices).
Feature parity with Amiga (capability to play multiple digital channels) only came with Gravis Ultra Sound / Sound Blaster 32 AWE.
Before that, digital audio was mixed in software.
(e.g.: all modplayer on PC)
but given the ever increasing power in PC CPUs, that was acually doable.
Once windows 95 and 3-d video cards emerged 10 years later, the Amiga's fate was sealed.
I would think it came a little bit earlier.
Amiga could count on its co-processors, PC tended to count on the raw processing power.
Amiga could easily do flat shaded 3D thanks to its coprocessor (which could offload the drawing of polygons).
On the other hand, the power CPU found on PC could do textured environment.
Thus enabling easily 3D games such as wolfenstein 3D and then Doom,
or games with pre-rendered 3D models like Commander (there were attemps at ports of WC on Amiga, but with sucky performance).
Now move further to Quake and it starts to be beyond the reach of the contemporary amiga.
As soon as game designers decided to push 3D beyond the classic flat shaded polygons, strong CPUs were required.
Someone old enough to remember the original PC here. CGA was not a standard feature; the base model (which most people had) used something called MDA (Monochrome Display Adapter), which supported text output only without any pixel-addressable graphics. {...} Almost nobody got CGA on the PC.
Fun because on this side of the ocean, CGA was quite widespread, and as soon as EGA was realeased by IBM, we started also seeing it here around.
(Maybe differences in buying power between countries ?)
The thing which *was* rarer here was composite output : local colors standards where PAL and SECAM (depending on country, but most TV where dual standard).
Not that many TV monitors did support NTSC too, meaning that TV-output wasn't trivial with the CGA's composite out.
So most often it was either monochrome monitors or expensive RGB monitors.
When Lotus 1-2-3 came along, the need for plotting drove the adoption of a proprietary monochrome graphics technology called "Hercules Graphics Adapter", which was much, much more popular than CGA ever was.
My father's business did use Hecrules graphic cars because of that plotting (though Hercules had hardware-assisted emulation of CGA, rendering the low-resolution 320x200 4 colors as high resolution halftones to simulat grayscale levels).
Our schools was CGA (some with even expensive color RGB monitors), then rather early EGA.
As a side note, even then sound wasn't a standard feature on personal computers. The most they could do was beep. You had to add a proprietary sound card to get anything more.
Nope. Again : the PC speaker had PWM capabilities.
It wasn't used in *games* (At least not before Access on much more powerful computers), because the PC Speaker didn't have DMA, and the CPU needed to push samples manually in software.
But it was used in a few novelty tools (small DOS program wich could synthetize speech by stringing together pre-recorded words).
And a few games used it for music during title screen (when the CPU isn't busy by the game itself).
But yes, PC sound only really started taking of in 1987 with the apparition with sound cards (AdLib, CMS, IBM MF, etc.)
(Though not everything was proprietary. In 1984 already the PC could do the openstandard MIDI with a specific card. But it didn't get used in games until MT-32 become popular, again by Sierra Online, starting with SCI text games, first used by King's Quest IV )
This is why Macintosh became common in schools. As a developer you could count on every Mac having the same set of very rudimentary capabilities. As a school administrator, you just unboxed the thing and fed it floppies; there was no opening the case and installing optional boards that had to have their address and interrupt vectors chosen by the user and configured by jumpers.
Different countries, different situation. Here around IBM PC were nearly as popular.
Probably because of the reputation of the magic "three letters".
There's more to CGA than that too. When used with a composite video output games could exploit the namesake of NTSC (that is Never Twice the Same Colour) to create artefacts that actually produced quite fantastic graphics (compared to 4 colour CGA)
Yup, one of the official methods to get more colors relied on the NTSC color clash.
This one was massively used by Sierra Online (lots of their CGA games used 16 colors on NTSC composite out - nearly all of the AGI text adventure game did)
The other one (working with RGB monitors too) and more or less officially documented, relied on the text mode.
The CGA is switched in 80x25 mode (640x200 pixels).
Text lines are compressed vertically until each caracter line covers only 1 or 2 pixels. (displaying 100 or 200 text of lines per screen - at least the top of each line as only the top 1 or 2 pixels of each caracter cell are display).
There are too characters in the table - 221 and 222 - that are half block. e.g.: the left half of the caracter cell is foreground, the right half of the cell is background.
By playing with that one can make a 160x100 (more frequently, because square pixels) or 160x200 (highest resolution, but single page) with colors.
(Technically it's 80x100 or 80x200 text modes, with each cell split in 2 halfs one with the foreground the other with the background color of the char cell).
with a wide range of colours in 320x240 mode with 16 possible colours on the screen at once out of a palate of 64 possible colours (in 4 groups)
BTW: it's 320x200, that the most vertical scanline you could output backthen.
This time, the technique is to turn on the 640x200 monochrome CGA mode.
BUT to activate the colour output on the composite output (can be make-shift hacked by setting a specific border colours which will be interpreted as the NTSC color burst), so a composite monitor will interpret it as color signal.
Then output specific black-white pattern (on-off pixels) on the monochrome screen, they will be interpreted by the NTSC as colors. That more or less gives 16 colors to play with. (but at a 160 horizontal resolution)
That's how nearly all AGI text adventure games of Sierra did work.
Alternative consisted of using the various palettes of 320x200 4 colors CGA, and the horizontal bluring of colors. (remember : NTSC offers only 160 horizontal color resolution), you have 16 ways to combine two 4 colors neight boors in composite mode, giving you 16 colors.
With a different set of 16 colors, depending on the CGA color palette selected.
The only problem is single pixels with contrast got blurred which made reading text very difficult in these modes.
Blur brought the resolution down to 160x200. but this "blur" is where the "color mixing" enabling 16 colours comes from.
Sierra solved it by having a 160x200 16 color playfield, but use only monochrome (black-on white) text to increase readability, and do all the critical text-only steps (like inventory management) in monochrome texte-mode instead of graphics.
Then, there's the modern "8088mph" demo.
It combines both above approaches.
It creates a 80x200 text modes, and use all possible patterns (including also the alternating pixels patterns, "!!" ascii character, etc.) together with all possible foreground/background colors combination, to create thousands of different outputs on a composite monitor.
I've found WinSCP to be better than FileZilla especially since so many providers offer SFTP now anyway.
Note that Filezilla support SFTP too.
I don't store my passwords so the master password thing is not an issue to me. Don't store passwords if you don't want them to be found.
Even better :
don't use passwords. Use Public Keys pairs.
(Filezilla supports them, and can use Putty's key agent to handle them)
(I'm sure that WinSCP can too, just didn't bother to check).
Best part : you can then completely switch off the support for password on the SSH/SFTP server.
Your server is then (obviously) immune to brute force / password guessing.