Windows 8 ARM Will Not Support Legacy Software 381
An anonymous reader writes "Intel, speaking out of turn and damaging its intimate relationship with Microsoft, has revealed that legacy x86-compiled software will not work on the ARM version of Windows 8. Microsoft has promised that the Office suite will be available on Windows 8 ARM, but beyond that, nothing. While this means there won't be many compatible apps at launch, it also means this will be the first full-bodied version of Windows that won't (initially) be susceptible to viruses and malware..."
no surprise (Score:2)
Re: (Score:3)
incompatibility with existing applications will be the driving force of the inevitable "windows 8 arm edition app store" and the pile of money microsoft expects to make from it.
Re: (Score:2)
Given that its for tablets with a touch UI. I doubt there's a huge demand for x86 software from the 1990s to run natively on such devices. Most of the apps they'll run are likely to be web-based with processing performed on the remote machine.
That depends. I can see the benefit of a tablet which has a touch screen UI while you're carrying it around but which reverts to a full desktop when you dock it.
As for apps, I doubt there are many enterprise orgs which don't have at least 1 application either thick client or browser based app with an ActiveX control or hardcoded IE6/7 dependencies in it. By ignoring reality Microsoft run a serious risk of alienating people most inclined to use the device.
I hope for Microsoft's sake that they're working
The .NET Framework (Score:2)
I hope for Microsoft's sake that they're working on something akin to LLVM so that C++/C apps can be rebuilt in an architecture neutral manner.
I thought that's precisely what the .NET Framework and the C++/CLI language were for.
Re: (Score:2)
What I'm suggesting is DevStudio should have a build target analogous to LLVM. When you choose to build the app to this target the compiler spits out a low level language which is converted to native assembly at runtime. Each version of Windows 8 would provide runtime support for the virtualized hardware so the same com
Re: (Score:2)
Software from the 90s will be fine as soon as someone builds DosBox for it!
It's software from the 2000's that's not .Net that's going to be the problem.
Re: (Score:3)
So it can integrate with your Windows desktop without needing itunes or some other third party sync utility? So it can run applications developers with the familiar (to the millions of windows developers) .net framework?
If the hardware is decent, battery life is decent, and the UI/integration is slick, the OS it runs is irrelevant. Instead of "why run windows", why not? Microsoft have Windows already developed and clearly cross architecture.
Re:Why buy a Window's device... (Score:3)
...if it won't run Window's software?. Most everything else is a simple compile away from user's being able to be run on any device they own. Microsoft is really going to screw the pooch here if they don't ensure compatibility with existing software.
Re: (Score:3, Interesting)
Re: I think the point here is that... (Score:2)
...there should be no need for it. When a program proven to run on for single processor running at 100mhz can't be run on on one emulating the same at 3000mhz it is time to lay blame squarely at the foot of the emulation environment itself.
Re: (Score:2)
Speaking of WINE, I have a question I want to ask that's usually asked as a joke, but I'm asking it seriously. Will this chip run Linux? How hard would it be to port Linux to it? I've sworn off Microsoft products, so if Windows is all this chip will run it's a no-starter for me.
Re: (Score:2)
AFAIK, linux does run on ARM for close to a decade. Heck, it even runs Ubuntu !
Re: (Score:2)
Linux is very well supported on ARM. In fact, many embedded systems run on Linux/ARM.
Re: I think the point here is that... (Score:5, Insightful)
Re: I think the point here is that... (Score:5, Interesting)
The only stupid question is one that isn't asked. Nobody knows everything (and I asked the question before I had my first cup of coffee). I got my UID by being on slashdot ten years or so ago. I'm 59 years old and my synapses aren't as well oiled as they used to be.
My first computer was a slide rule. My second computer I built out of two potentiometers, a voltmeter, and a battery. When I was a teenager I made a little extra cash by converting cheap transistor radios into guitar fuzzboxes and selling them to friends.
These days it's fashionable to be a nerd, but I was a nerd back when we were pariahs.
Since Linux runs well on ARM, then I don't see what the big deal is about not being able to run legacy Windows apps in Win 8. All you'd have to do would be to install Linux dual-boot on your Windows 8 machine, and run your legacy Windows apps under Wine in Linux. Maybe I still need more coffee...
Re:Why buy a Window's device... (Score:5, Informative)
Would not help at all. Wine is two things, its an implementation of the Windows api and a loader. If you have the source you can compile your windows api application for other some architectures using winelib. So you might be able to port your program to ARM Linux with it. You would not need winelib on Windows because Windows will provide the windows api.
You can't use wines loader and server functions to run x86 code on ARM period, it does not provide a virtual machine. All it can do is let you run binaries build for x86 windows on other x86 platforms. So wine is useless for running legacy software on ARM Windows.
Re: (Score:2)
http://wiki.winehq.org/WineOnWindows [winehq.org]
Working on it apparently
This page is about trying to get Wine to run in Windows. Many Wine DLLs can be cross-compiled with MinGW already, but Wine itself doesn't work yet.
Why would we want to get Wine running in Windows? Newer versions of Windows fail to support old applications that are still supported by Wine. So Wine for Windows would supply useful backward compatibility for users.
Re: (Score:2)
Does anybody really think virus writers don't own compilers?
Me? I'm bet the virus writers will have their stuff working long before the anti-virus companies have theirs.
PS: Won't Word macro viruses work without even a recompile?
Re: (Score:2)
ANS: Office, Exchange, .NET (Score:2)
...if it won't run Window's software?.
Well, you can be sure that it will run "official" MS Office and"official" Microsoft Exchange clients (...and probably Exchange servers if ARM servers take off). That will be attractive to some people (even though it might not sway the average Slashdotter).
Also, as others have pointed out here, it could potentially run any Windows software developed in .NET, which compiles to bytecode (like Java) rather than native machine code.
That's assuming netbooks and servers - tablets and phones are a different kettl
Re: (Score:2)
Re:they already have windows for arm (Score:4, Interesting)
The problem with those cheap wince based laptops, is they're advertised as running windows, which means people buying them often expect that they run the same windows they may already have on a desktop, or have at work etc... Once they get it, they are usually severely disappointed and this usually results in a very high return rate.
Another ARM version of windows is likely to do the same thing, disappoint users, fragment the brand and end up with lots of returns...
An ARM based version of linux on the other hand could sell very well, if its properly marketed... Users would have no preconceptions about it, and take the devices for what they are. Just make sure there is a proper linux distro, not the crippled versions that came with the first round of x86 netbooks... And make sure the benefits of linux are well advertised to users, especially the package manager.
Re: (Score:2)
People didn't want Linux netbooks, because they couldn't run the applications that they wanted on them. That's why so many were returned, and Linux quickly disappeared from the netbook scene. Why would you think it would be different this time around?
Re: (Score:3)
Because this move resets the software scene. There won't be "applications that they wanted on them" anyway, so it really is a golden opportunity to throw a proper, Joe User-oriented Linux distro out there.
Re:they already have windows for arm (Score:4, Insightful)
People who bought netbooks didn't really wanted a netbook - what they wanted was a very small, mobile and cheap version of the notebook computer they used before (hence in the end most netbooks came with Windows). Unfortunately this combination makes for very poor usability.
This is what killed the netbook market and why the iPad is reigning supreme. It takes a while for people to figure out that what they originally wanted from a particular product does not perform as they originally envisaged.
Re: (Score:2)
People who bought netbooks didn't really wanted a netbook - what they wanted was a very small, mobile and cheap version of the notebook computer they used before
And I got it. Twice the speed, twice the ram, half the size, and only 168 fewer pixels vertically.
Unfortunately this combination makes for very poor usability.
I'll have to disagree with this one.
Re: (Score:2, Insightful)
And I got it. Twice the speed, twice the ram, half the size, and only 168 fewer pixels vertically.
This just indicates that your hardware replacement rate is a lot lower than that of most others.
There's a reason the current Macbook Air and the iPad are both doing well, they cover two slightly different use cases that people tried to cover with netbooks. The Air covering the "A laptop that's actually mobile, looks nice, has enough power to run all my regular apps" and the iPad the simple mobile entertainment niche (watching a movie in the bathroom/on the train, browsing the web from the couch, etc.).
(Yes,
Re: (Score:2)
That and they are actually promoted, a great machine is of little use if noone knows it exists.
I have a HP mini 5101 with the high res screen option and I love the thing for giving me a usable screen resoloution (which IMO was the biggest issue with netbooks for general desktop tasks) in a small package but afaict HP never marketed it. So the only people who would have been likely to end up with one are those carefully researching the market for a machine with those charactersitcs.
Re:they already have windows for arm (Score:4, Insightful)
Re:they already have windows for arm (Score:4, Informative)
Windows CE only shares the name with desktop Windows.
It's basically a different OS.
Windows 8 (as I unterstand) will be the same OS compiled for a different platform.
But yes, Microsoft has experience with mulitplattform OS': Windows NT ran on Alpha and other architectures.
Re: (Score:3)
While this is obviously not aimed at desktops, couldn't this version of windows run on ARM-enabled desktops?
And if not desktops, maybe TVs, info-kiosks, ATMs etc. = semi-embedded machines with a limited scope and not really needing 100% backward compatibility. It's enough for a few critical pieces of software to be recompiled. Smart, internet enabled TVs and refrigerators ruining Android might become a huge growth market in the near future, and Microsoft clearly wants a piece.
Also, let's not forget about dynamic translation. Android applications themselves aren't naive ARM, rather JIT-compiled bytecode. I don't s
Re: (Score:2)
Re: (Score:2)
Apple did port OSX from 68x to Power and then from Power to x86. Every time, legacy apps did keep working.
Re: (Score:2)
Re: (Score:3)
The point is not whether it was emulated or not. The day they shipped it, all apps kept working.
With windows, MS is announcing that the day they'll ship it, no apps at all are going to work.
That's a hell of a difference. Plus the GP was bitching about the fact that "It's a different architecture, of course apps will not work." It could. It won't.
Re: (Score:3)
Your not understanding, every time apple changed platform they increased the performance and speed. In this case its the opposite (one day arm may be at a similar performance but not now). Much much harder to get an emulator to work on slower equipment. My analogy should of said its like trying to play a PlayStation game on a Nintendo 64. There is a reason apple used ios on the iphone and its not touch.
What is preventing ARM from competing performance wise? Is there something inherent to the architecture that prevents a company like AMD or Intel from licensing the architecture and producing a six-core beast running at 4 GHz? I understand that ARM is optimized for low power consumption, but if you were to slap a big, honkin heat-sinc and fan on one, like you have to do with a x86 chip, and give up any powersaving features that rob performance, is there any reason why an ARM chip wouldn't outperform an x8
They went further than that (Score:4, Insightful)
Intel went so far as to say that legacy software would "not ever" run on ARM. To do that they have to have to have the stick of software patents to prevent an ARM->x86 emulator.
This is not good for Microsoft. It means their relationship with Intel is irretrievably broken. The WinTel alliance is no more.
As consumers we can win from this. Without the constraint of making the bloated Windows OS run on their chips, Intel can dive into low power. Without the glacial software development lifecycle in Redmond Intel can bring out new stuff faster. That's good stuff.
The distant threat is that when Intel seeks a market they want all of it. They're late to this game and their Atom chips don't cut it yet - their promises are some 24-36 months out, and ARM and Microsoft are not going to be standing still in the meantime. They're promising "best in class mobile video tech" but I swear to God if they buy Imagination Technologies to cut out ARM mobile chipset vendors I'm going to fucking do everything in my power to kill them. That would shift Intel from the "Invention of technologies" camp to the "prevention of technologies" camp. I'm not OK with that.
But if what Intel means is that they're going to let the legacy go and deliver the best low-power chips they can, that's a good thing. Your PC doesn't have to burn the watts it does. There are lots of folk in the third world with valuable input who don't have watts. It does not take a kilowatt gaming rig to work spreadsheets any longer.
Re: (Score:2)
The trouble with Intel is that they are tying themselves to x86, which carries with it a lot of legacy cruft that ARM doesn't have to deal with... The end result is that, in order to remain competitive with ARM Intel have to keep a step ahead on fabrication technology, since an ARM fabbed on the same process will always have an advantage.
Re: (Score:2)
If you have to increase the amount of L2 and L3 cache and implement pipelining to get the type of integer performance that x86/AMD64 processors are achieving for this type of workload, then all of the savings that you should automagically get from ARM pretty much dissipated. The translations between the CISC x84 instruction set and the internal RISC core of modern x86 CPU are relatively efficient.
The only reason I see that MS is moving to ARM is because from a power consumption and scalability standpoint, i
Re:They went further than that (Score:4, Informative)
As consumers we can win from this. Without the constraint of making the bloated Windows OS run on their chips, Intel can dive into low power. Without the glacial software development lifecycle in Redmond Intel can bring out new stuff faster. That's good stuff.
Yeah, it was Windows holding them back, not the laws of physics. Nice catch.
Oh, and Windows is better with power than any of the Linux distros I've used.
Re: (Score:2)
With a [url=http://en.wikipedia.org/wiki/Maemo]little optimization[/url], Linux can be /far/ better with power than Windows ever can.
Just because you're running a desktop-oriented distribution doesn't mean the underlying system's poor.
Re: (Score:2)
Just because you're running a desktop-oriented distribution doesn't mean the underlying system's poor.
Whelp, better not compare it to Windows like symbolset did then. It's a bit unfair to compare a desktop OS to an OS running on a phone or something, surely you can agree?
Re: (Score:2)
Re: (Score:2)
{Oops, html mode makes aweful looking plain text)
Well, I think we're talking about the use case of laptops. With a laptop, windows power consumption may be lower than Linux /in some cases/(others swear it's the opposite), it's not the fault of Linux, just of how optimized it is for the device. Maemo's a good example of what you can do with a little optimization, and it doesn't really give much up compared to a desktop-optimized distro.
When it comes to tablets and phones, if we take the example of Maemo agai
Re: (Score:2)
Because Microsoft is moving to ARM, known for its high power draw. Clearly, Windows is the reason that Intel chips have been devouring power like that is the main focus of their chips.
Microsoft had one OS slip: Vista (note: this is not referring to flops, such as Windows ME). They corre
Re: (Score:2)
Re: (Score:2)
Acorn computers have had x86 emulators for ARM around 20 years also. Nothing to stop Intel building ARM processors though, they have a license for StrongARM they inherited from DEC all those years ago.
Phillip.
.NET (Score:3, Insightful)
Re: (Score:2)
OpenOffice is not written in Java (or .NET), so it won't work.
I would say a vast majority of Windows applications are *not* .NET (notably games, old-as-dirt product lines)
Re:.NET (Score:4, Interesting)
Unfortunately, as people trying to get .NET apps to run on Mono have found, a very significant portion of those .NET apps do actually call unmanaged code.
Re: (Score:2)
Re: (Score:3)
If microsoft provides the Win32 API on ARM it will only take a recompile to make them run on ARM.
Keep dreaming. It will take much more than a simple recompile.
It doesn't really matter that the x86 binaries that exist today do not run on ARM.
You are discounting legacy software that will never be recompiled for anything. Also, you are presuming that ISV's are going to care about Windows 8 on ARM. That is extremely speculative at this point particularly in light of the tepid response WP7 has had.
Simple solution... (Score:4, Insightful)
Re: (Score:2)
Re:Simple solution... (Score:5, Informative)
Re: (Score:2)
Hate to burst your bubble, but Windows CE did and still does (there are still new products with it) run on ARM. Endianess is not a hard problem in high level languages.
Additionally I had to laugh at this:
While this means there won't be many compatible apps at launch, it also means this will be the first full-bodied version of Windows that won't (initially) be susceptible to viruses and malware...
All the malware that ran under Windows CE can probably be ported to Windows 8 ARM with little to no effort.
Re: (Score:3)
Re: (Score:2)
ARM supports both endiannesses, and most if not all heavyweight mobile devices these days are configured in little-endian mode. To user-level C software not using inline asm, ARM looks very similar to x86. The main difference is ARM support for unaligned accesses is nonexistent in earlier chips and incomplete in newer ones, but you can have the OS trap those and emulate them (at a significant performance penalty, but depending on how often they happen that might not matter).
Re: (Score:2)
Manual bit-packing is mostly unnecessary if you can target a specific compiler, anyhow. This is why God gave us structs, bit fields, and __attribute__((packed)).
The only tricky part with code units > 8 bits going over the wire, you need to match endian.
Really? (Score:5, Informative)
This just in, x86 and ARM instruction sets are NOT compatible! Everyone panic! Blame MS! No, wait... Sony must have had a hand in this!
File this under no shit, Sherlock.
Re: (Score:2, Informative)
Then what does Rosetta do on Mac?
Re: (Score:2)
It's a VM, naturally.
One could conceivably write a VM for Windows 8 to emulate x86 on the ARM, but that wouldn't be native. It would be a VM.
I see this as some Intel FUD sowing about the whole ARM craze, really...
Re: (Score:2)
Rosetta is a binary emulator/translator. You are not running non-x86 instructions on x86, they are translated from their native PowerPC instructions to x86 equivalents. Rosetta is far from comprehensive and does not magically work with every PowerPC app, unlike what Apple would like you to believe. Many basic apps which don't do anything complicated generally work, but once you go outside that small area, things fall apart pretty quickly.
It's also rumored to be discontinued in the upcoming OS X Lion.
Re:Really? (Score:4, Informative)
I would beg to differ with regard to Rosetta not working with anything complicated, and I have a perfect example - Mac Office 2004 on a Core Duo Macbook Pro, verses iWork Numbers on the same platform.
I had a spreadsheet with about 200 data points, of which I wanted to make three graphs - in Numbers, running natively on Intel, it dragged along for tens of minutes when rebuilding the graphs. With Mac Office 2004, running under Rosetta, Excel had the whole thing done in a couple of seconds.
I haven't used Numbers since.
Its not just about instruction sets (Score:4, Interesting)
This just in, x86 and ARM instruction sets are NOT compatible! Everyone panic! Blame MS! No, wait... Sony must have had a hand in this!
File this under no shit, Sherlock.
I think what intel is saying is that MS are:
that Windows 8 is going to drop some of legacy API support available in WIndows 7 - and while Win8 x86 is going to offer a "classic" mode this won't be available on ARM (...I wonder if this is a reference to the existing virtualization-based legacy mode in Win7/Vista?)
Of course, what Microsoft gets and Intel apparently doesn't is that Win8/ARM's main competitors will not be other Windows machines (as was the case when Windows NT briefly supported other processors such as Alpha) but against iOS and Android in the mobile world and Linux in the server world. If Win8/ARM netbooks can run "geniune" MS Office and Win8/ARM servers talk "genuine" Active Directory and Exchange Server, along with lots of "modern" windows software written in .NET, some people will choose them over iOS, Android or Linux. Intel will surely be the solution of choice for corporates wanting to run their 1990-era dBaseII systems - but even that market will eventually fade away.
As for tablets and smartphones - they'll need custom-designed software anyway so legacy is irrelevant.
(* Hell, I was running x86 PC software via an emulator on my ARM3-based desktop back in 1990 - but the ARM3 was a desktop superchip that smoked the 286s of the day... maybe ARM will make a triumphant return to the desktop, but it will need a 64-bit makeover and a FPU).
Re: (Score:3)
In related news, Exxon announces that Windows 8 ARM computers will not run on gasoline. "At last a full-blooded Windows computer that is guaranteed to be non-flammable!" gushes an anonymous /. contributor.
Re: (Score:2)
OS X -> PPC, Intel x86, Intel x64, ARM (iOS)
Windows -> x86, x64, and previously MIPS, Alpha, PPC
Just because its not released, doesn't mean its not maintained internally "just in case" (no doubt, as is the case with MS on ARM).
Re: (Score:2)
Yes, actually i am quite certain that all big OS companies will have some internal demo version running on the major processors families. If you dont support a lot of HW, recompiling a OS is not so time consuming. And i personally found compiling/running a program on different architectures a quite useful metric of code quality. It keeps weird assumptions localized to specific places.
Re: (Score:2)
A challenge, which was mastered before, like in the Basic for the commodore 64.
But it could be a performance hit.
Re: (Score:2)
I'm guessing Windows 8 is probably only going to run on the models of ARM chips that do actually have an FPU, then. Most of the ones used in stuff like tablets and smartphones tend to anyway, though I think the FPUs in some of the slightly older ARM chips are a bit slow. (Kit like ARM-based routers and the various plug computers out there doesn't tend to have one though. There are even some low-end ARM chips with no MMU, which tends to make running any modern OS impractical.)
FPU Re:Really? (Score:2)
What the frack are you talking about?
Every ARM worth talking about these days (past ARMv6) has a mandatory FPU coproc on 14 (single-precision) and 15 (double-precision), it's a vector FPU, and NEON multimedia in the later Cortex versions. All of the iPhones are ARM and have a FPU.
I should know. I'd written a FPU driver for WinMob6 that remapped all the software library calls into hardware calls, as well the fact MSVC generates ARM inline FPU calls, if set with the right flags.
Why do you think Froyo jumped
Re: (Score:2)
Acorn had a PC Emulator in - ooh, 1987? 1988? which ran DOS quite happily (albeit rather slowly ;) on an 8MHz ARM2.
Probably not terribly relevant these days as I would be astonished if any modern version of Windows even runs on something without some sort of FPU, but there you go.
Well, of course (Score:2)
I think Microsoft—after repeatedly failing in the tablet market with Windows—has finally noticed that precisely what allows the current tablets to succeed is that they don't try to act like a touch-screen desktop. There's no point in them bringing compatibility with old apps!
Initial Viruses (Score:3, Insightful)
won't (initially) be susceptible to viruses and malware
Well, now, I wouldn't speak too soon. There will undoubtedly be a beta release or a leak which will give malware authors ample time to develop zero-day viruses. And with Windows 8 exploring very different terrain this time around, there's bound to be a plethora of exploits just waiting for someone to coax them out of hiding (or plain sight).
Not complete accurate (Score:4, Informative)
The article isn't completely accurate. It fails to specify that it will not natively run x86-based code on Win8 ARM. There's no valid reason why x86 code won't be able to run inside a virtual x86 machine running on top of the ARM architecture.
The summary also makes this statement which is not accurate to the version in the article:
it also means this will be the first full-bodied version of Windows that won't (initially) be susceptible to viruses and malware
The actual quote is that it won't be susceptible to existing viruses and malware.
They also assume that all code will have to be re-written from the ground up, which is completely false. Most application code will need to be ported, and in many cases security holes which are due to fundamental design flaws (as opposed to coding mistakes) will simply be ported along with it. So yes, a lot of existing malware will break but that's no reason to lay down and assume that developers who made crappy software in the past will suddenly cease their shitty practices.
GNU/Linux (Score:2)
Oblig (Score:4, Funny)
-- Steve Ballmer, 2012
Crap summary (Score:3)
I think I speak for many here when I say that the editors need to, well, edit a bit more. The summary is full of bias which should be reserved for the comments. Can we please just have factual summaries in future?
Re: (Score:2)
I think I speak for many here when I say that the editors need to, well, edit a bit more. The summary is full of bias which should be reserved for the comments. Can we please just have factual summaries in future?
Your Id is below 200k, you should know better not to ask.
Re: (Score:2)
Microsoft already commented on this (Score:4, Informative)
Re: (Score:2)
MS has integrated instruction set emulation into windows before (they certainly did it on alpha and I think they did it on itanium too since the hardware level emulation was appalling and IIRC it was even removed in later versions of the itanium line), there is no reason they couldn't do it again. Intel are claiming they won't but then it's in intels interests to claim that given that they are trying to compete with ARM.
a-duh (Score:2)
Re: (Score:2)
There will be two versions of Windows 8, Legacy and ARM - so pick one and stop whining.
If the Register's report on Intel's claim is correct, there will be one for each SoC. So it would be Windows 8 x64 edition, Windows 8 Tegra edition, Windows 8 OMAP edition, Windows 8 Snapdragon edition and so on...
Re: a-duh (Score:4, Interesting)
And that is the really shocking thing that will actually kill the platform - fragmentation. All of these different versions will be incompatible with each other, forwards and backwards. Intel must be laughing their asses off.
The lack of a standard "ARM platform" has already been a big problem for Linux netbooks. They're all x86 because each ARM platform is different and requires a different BSP, making ongoing support a complete nightmare. I have to say, I really expected Microsoft to force the ARM SoC makers to standardise.
The lack of any sort of x86 emulator is really the icing on the cake. The big advantage of Windows is gone. But I suppose there is still a possibility of a third-party emulator like the original Virtual PC for Mac.
Re: (Score:2)
To be fair, with linux on ARM, the binaries seem to be pretty interoperable, at least at the application level. AFAIK - and I will have to double-check that - I was able to run an armv5 binary built with Debian on an AC100 (Tegra2). Certainly this was not much of a problem with CE and PocketPC, and the Android NDK seems to be pretty robust too.
At the kernel and bootstrap level, it's a freaking mess, though.
Re: (Score:2)
I'd agree absolutely, it is not so difficult to have compatibility in "user mode", unless something really crazy is going on (like, some SoCs have VFP, NEON and Thumb while others don't.. which seems unlikely).
But I don't like the idea of being reliant on the hardware vendor for OS updates, or (for that matter) being locked in to a particular OS because nothing else works on this highly specific platform. Android users have already suffered whenever telcos, vendors and manufacturers have failed to provide
Re: (Score:2)
Re: (Score:2)
No, it really isn't. Because the (16-bit) BIOS is not used after the (32/64-bit) kernel boots, the kernel has to include the BSP. And there is only one generic BSP for all x86 PCs.
For instance, when the x86 Linux (or Windows) kernel boots, it finds devices on the PCI/PCIe buses using IO ports 0xcf8 and 0xcfc [wikipedia.org]. This is common to all PCs with PCI buses. The location is always the same, the access protocol is always the same. Combined with similar standards for graphics, the keyboard, the interrupt controller,
Re: (Score:3)
You misunderstand. You appear to think I am talking about drivers, but I am really not. The BSP (board support package) is a much lower-level entity than that. While it does contain drivers, they are drivers for things that the OS needs immediately and cannot load at a later stage, such as the interrupt controller and the timer. The location and size of usable RAM come from the BSP, which will also specify the interrupt mapping and the locations of peripheral controllers such as the PCI configuration regist
It's nice (Score:2)
How? (Score:3)
How does releasing Windows 8 for arm and not supporting x86 apps equate to "the first full-bodied version of Windows that won't (initially) be susceptible to viruses and malware...?"
Does that mean that Windows 8 is closing all of the security holes that allow for viruses and malaware? If so, why would it just be the arm version that is protected? On the other hand, if they are not closing the security holes, how does using arm protect it? Just because a virus will have to be rewritten to execute on the arm platform does not mean the platform won't be susceptible to viruses and malware, unless the OS is changed to protect against it.
Microsoft already denied this yesterday (Score:3)
"Intel’s statements during yesterday’s Intel Investor Meeting about Microsoft’s plans for the next version of Windows were factually inaccurate and unfortunately misleading. From the first demonstrations of Windows on SoC, we have been clear about our goals and have emphasized that we are at the technology demonstration stage. As such, we have no further details or information at this time."
Re: (Score:2)
But it was Intel who said this! And no, it can't be that Intel just pointed out the obvious. After all, it's Intel. They are in bed with Microsoft. They are obviously evil. I'm sure there's an x86 core hidden somewhere in ARM which is only disabled because of Intel! After all, how could a processor work without an x86 core? :-)
Re:first full bodied nonx86? (Score:4, Insightful)
PPC, MIPS, Alpha, IA64 and i860 i believe...
What do all these have in common? Noone used them.
At the time, these architectures offered vastly superior performance to x86, but couldn't run legacy windows apps or legacy apps designed for other OS that typically ran on the hardware. Since there were so few users, virtually no commercial software was ever ported to non x86 windows and very few people ever even bothered to port open source code to them.
MS' biggest strength - proprietary lockin, is also their biggest weakness...
If your going to move to an incompatible hardware platform, and lose access to your legacy software in the process then you'd be a fool to run windows... Linux already runs on ARM, will not lock you in like windows is designed to, costs nothing, and already runs 99% of the same software the x86 version does.
And ofcourse if everyone is running open source code, the architecture becomes irrelevant and we can switch again very easily if something better than ARM comes along.
It's also possible to have a range of architectures for different purposes, ARM or MIPS for low power devices, perhaps x86, IA64 or Alpha for high performance devices where power usage isn't a concern.
Re: (Score:3)
I used NT4 on Alpha - there were even beta copies of Windows 2000 for the Alpha.
It also had FX!32 which was fairly decent at allowing x86 binaries to run under Alpha NT4 which meant companies didn't need to port software specifically for the Alpha.
Re: (Score:2)
Exactly! Historically, Intel does not have a great track record at succeeding in the marketplace with new architectures.
If we just go with Intel:
i432 Super Object Oriented - lots of transistors, super CISCy. Very slow too, but probably could have been sped up decently (half the problem was the compiler and lack of caches)
i860 - Very fast for it's day as long as you don't need to respond to a trap or interrupt... Evil to program. Great for an accelerator but not a great general purpose CPU
i960 - This on
Re: (Score:2)
You're never gonna give this up are you Rik?
Re: (Score:2)
It will take two more days for the /. editor to pick up this "news".
Re: (Score:3)
Who wrote the article is an idiot. Properly written managed .NET code will _at worst_ require recompilation. Ask yourself why MS has been pushing managed code like hell...
At worst, it will call down to some ancient DLL for which the source code has been lost or for which the vendor decides it's not worth porting.