Submitted by Very light on details, but this is interesting nonetheless – very interesting, and potentially one of the biggest things to have hit the operating systems business this decade. Bloomberg is reporting that Microsoft plans to announce Windows for ARM processors at CES in January 2011.
Windows on ARM has been talked about before – in fact, I firmly believe that Windows NT has been running on ARM inside Microsoft HQ in Redmond for years. I’m not sure how much of it is left, but Windows NT was originally designed to be highly portable, so much so in fact that the first versions of NT were coded on boards designed internally at Microsoft (Dazzle (Intel i860) and Jazz (MIPS)). Earlier versions of NT ran on x86, PowerPC, MIPS, Clipper, and Alpha, and a port to SPARC was under way at Intergraph.
So, NT’s roots lie in portability, but how much of those roots are left remains to be seen. However, even if Windows NT can run just fine on ARM, the bigger problem is of course the massive body of applications that do not. Then again, with ARM clearly being an embedded thing – in this case, tablets in particular – the usual Windows applications might be useless anyway.
Anywho, what’s Bloomberg reporting? “Microsoft, the world’s largest software maker, will announce a version of its Windows operating system that runs on ARM Holdings Plc technology for the first time, said two people familiar with Microsoft’s plans,” they write.
That’s it, basically. Those two people might as well be talking about Windows Mobile CE Pocket PC Embedded Compact (or whatever it’s called these days). They’re also not saying anything about how close to Windows NT-regular it would be, and how, and if, you would run regular applications on it.
Linux has quite a big foothold on ARM
MS does not have a good enough OS on ARM to complete. So NT based kinda has to come across.
Still will not change the problem that most of the Windows programs out there are x86.
But still Linux biggest advantages on arm are drivers and applications. Yes the mirror of MS biggest advantages on x86.
At least now we might be able to say by by to Windows CE and Windows Phone. So leaving MS with only 1 OS core to focus on.
… and further, even if your favorite Linux application hasn’t been released on ARM, you at least have chance of compiling it yourself. I used “scratchbox” to compile some essential command-line utilities for my ARM-based Nokia N810. It was a bit complicated, but way better than waiting for a proprietary commercial company to do it (or more likely not) for you.
How do you know if it hasn’t been released, we have almost no details and no reviews?
If you had read my full line. You would have seen this was referring to MS existing OS’s on arm. Windows CE and Windows Phone(what is basically a CE core).
CE feature set does not compare to Linux or darwin or qnx. Yes Linux the core of Android. Darwin the core of Iphone. And QNX the core the blackberry.
Most likely the fast way to expand there OS core function is port NT. But even then NT has baggage in its design.
Have you ever seen/used a *full* version of CE 4 or greater? (and *full* is subjective, as the platform builder lets you pretty much customise the OS completely.) The OS itself is a reasonably complete implementation of Win32 – albeit, the UNICODE version of the API. The userland lets it down, bet really, the restrictions were put in place when it was running on devices with about a couple of megabytes of internal storage, including the ROM space. I used to code for a handset with an RFID reader and the OS it used was extremely complete, if you ignored the fact that Microsoft didn’t seem like they were trying to make it in to a desktop OS. It was no worse than the Sharp Zaurus SL5500, which runs LINUX and was of a similar vintage. This was 5 years ago. Things don’t stay still.
Someone else mentioned drivers – well, no one seems to have issues creating drivers for other CE based devices. In industrial use, they are quite frequently used in various applications for data entry and such. They come with a variety of interfaces and add ons.
Apps – well, most Win32 API based apps are reasonably easy to port. At the point I was at 5 years ago, the main issues were the lack of non UNICODE API and that the OS was missing some the parts that the desktop OS had available. But, if you had the entire Windows codebase, porting those parts would be possible and so therefore most of the issues would go away. If you use DotNet, the differences reduced greatly, especially when you are using CE 5 based OS.
Yes I have. Its not a full OS. It lacking a proper secuirty framework that darwin, QNX and Linux has. This secuirty framework side is becoming a lot more important as devices increase in number and become possible targets.
For the simple case for a lot of items. If CE does not have the drivers we just drop it from list of OS’s to use. Yes we don’t have driver issues with CE since we just don’t use if it has.
DotNet does not address the secuirty risks that increase the means to CE based devices to be attacked.
UNICODE issue is nothing more than a smoke screen there were many third party solutions for CE to deal with the UNICODE issue. Really thinking you were not ment to be using non UNICODE code on Windows unless it was for legacy code from 1998 and before. So by CE 4 if you were running into it your coders were crappy. Sorry its the truth. So how come you were running into it 5 years ago.
Missing some parts that the desktop OS has is the serous issue. There are quite a major list of features that CE simply does not do. Yes it was built for the Limited memory space issues that don’t exist today.
I understand that, but why are you talking about CE? That is old news. This is new news. As such, I think they are talking about full blown windows 7. You haven’t seen it. I haven’t seen it.
If you are talking about a modern version of windows ( like 7 rather than NT4, the last time the product name used “NT”), then I think the market has thus far decided that its “good enough” for most desktop users, despite its flaws. And I think that’s what this is about, desktop on ARM. Really more likely netbook/tablet on ARM, but same difference.
It will likely be a juiced version of CE… I don’t know what else it could be. Unless MS has been building Windows on multiple platforms (kind of like OS X was over the years)… but MS would have no real reason to do that since the x86 market has always been all it needed.
Are you forgetting the versions of NT 4 for alpha and ppc? Windows NT was designed to be portable across processors. Have they kept it portable? At one point some one thought that an alpha powered NT machine was a good idea ( I did too, but lacked the money at the time for an alpha).
I knew about Alpha… I worked at DEC in my early years. Did not know about a PPC version! So there you have it, all things are possible.
http://en.wikipedia.org/wiki/Windows_NT#Supported_platforms
For those interested in arcane OS/chipset facts.
It’s more than that. The complete list of platforms a NT kernel has been ported to is: Intel x86, MIPS R3000/R4000, Alpha, PowerPC, Itanium, and AMD64
Edited 2010-12-22 17:20 UTC
no apps, no drivers, no devices.
This seems like a desperate move by microsoft..
Without further details we really cannot say that.
It’s not as though they go at this alone then release it to hardware vendors as a suprise expecting them to then go and produce hardware / drivers, they’d be already partnering with them.
I think it’s Microsoft responding to the market, which is what businesses do.
Desperate move? Would you say the same if support for a new architecture was added to Linux or *BSD? Probably not, right?
As has been demonstrated by both iOS and Android, new platforms can gain acceptance and applications fairly quickly, if they provide something people want.
Of course not because *BSD/Linux already run on almost every CPU architecture out there. Desktop incarnations of Windows runs on exactly one. That is a desperation move and Microsoft has a lot of work cut out for them to be even remotely competitive. From a performance perspective they’ll also face similar issues as they do with netbooks but even more so.
Your argument makes no sense. Did you also think is was a desperate move when, back in the days, Linux went from x86-only to also support its second architecture? I didn’t think so. My point here is that the reason why people like yourself want to call this a desperate move, is simply because you don’t like Microsoft and their products (btw, me neither), but that has nothing to do with them being desperate, but more about people spreading FUD because they think Microsoft is “evil” and desperately want them to fail.
It makes perfect sense. Back then Linux had zero market share to hold on to. Microsoft has everything to lose. Mobile devices are the future and Microsoft’s attempts at this so far have been extremely poor.
Right, because in the future we’re gonna be writing code and doing word processing via little on screen keyboards.
Yes, I know, this out of context. So are your claims of desperation by Microsoft because they decided to support another platform. See, I can pull sh** out of my a** too.
1. They didn’t decide anything. It’s a rumor.
2. “Support another platform” sounds so benevolent. But of course the story is a different one if they add ARM: If they do such a thing then because they feel pressed to do it (like Apple going from PPC to x86).
3. You can’t compare non-commercial projects like BSD, Linux, etc. – where support for another platform is done “because one can” or “for the fun of it” or because one needs an OS for hardware XY and don’t want to start from scratch – with the actions of a commercial entity. If a company needs to support a new platform (which is an investment, meaning you have to spend money without knowing if it ever pays off), it means they don’t see a bright future for their current assets compared with possible new ones.
If you own a platform / market like MS does with the x86 / PC market, but have to invest in new platforms instead of simple milking your old platform / market (like they did so far) is not a sign of confidence. Zune, Bing, Windows Phone 7, etc. are good example and investing in ARM would be another.
It doesn’t matter if you call such a move desperate or not – the signal stays the same: MS fears that “their” market because much less relevant because people can substitute more and more classical PC functions with smartphone and tablets where ARM rules and MS is nowhere and will not be able to simply misuse their abundance on the desktop combined with ‘we don’t care for standards and interoperability – let other adapt to Windows’ to gain market share (like they did with servers).
microsoft is evil. They were convicted of abusing their monopoly position in the market to the damage of the market.
Under the table oem deals, behind the scenes threats against manufacturers to lock out potential competition, strong armed treatment of technology partners, thats an incomplete list but reason enough why they should become a minor player in the market.
Thanks for proving my point! Which, in case you missed it, was that this isn’t about MS actually being desperate, just disliked and bashed.
Truer words have never been posted here. If anyone’s desperate here, it’s the Anything-But-Microsoft zombies and their bitterness over Microsoft’s success. Sadly, they’re all too witless to realize that that bandwagon left them behind ages. For mindless wannabe-hipsters, Google’s the company that’s now cool to hate.
That’s… not very good trolling.
Spreading FUD about MS must be really hard. I think you need people that trust MS before you can spread FUD about it. Only supporting x86 for so long that most competitors of Intel just disappeared or had to move to different markets was bad for consumers.
Windows NT kernel is built to be portable (there is a version of Windows XP for Itanium and Windows Server 2008 for Itanium also). The diagram below is the Windows 2000 architecture, I have seen similar diagrams for Windows VISTA & 7, but I couldn’t find it with a 2 second google. However it illustrates the point.
http://en.wikipedia.org/wiki/File:Windows_2000_architecture.svg
Notice at the bottom there is a Hardware Abstraction Layer. This is where I would imagine (I am a web developer not systems developer so I can’t say for certain) most of the main development will be in porting Windows to ARM (if it does indeed happen).
Secondly, Windows 7 & Windows 2008 R2 are highly modular e.g. We have Windows 2008 R2 servers which boot up without the windows shell, i.e. DOS prompt and not much else, you pretty much need to know powershell to interact with it. I don’t doubt they will customise the build to include what is necessary.
Thirdly Windows 7 already support multitouch, and they already have a touch based UI with Windows Phone 7.
I agree with the other poster I doubt this is a desperate move, I think it is a pro-active one.
Edited 2010-12-22 14:04 UTC
There is some version of Silverlight for ARM which apps can be built on. Native applications can come after. And honestly, you don’t want to run native x86 windows applications on a tablet.
Edited 2010-12-22 00:02 UTC
Silverlight on arm don’t make me laugh. That is more I want to see my battery disappear. Its not power aware so is not suitable for portable devices.
This is the big bagage NT carries none of its applications are designed to work in power effective envorments ie like where you are running on batteries.
Silverlight runs on wp7 (arm) already.
And it eats power not effective. Silverlight has failed as a web platform. To try to save MS R&D costs MS has tried to refactor it as a Mobile device development factor without retooling it for that job.
Basically there is such thing as kicking a dead horse and Silverlight is exactly that. Its a dead horse that MS does not want to admit is a screwup.
.Net is also screwed up for Mobile Devices. Google did not alter the way Java worked for Android for no good reason. The importance of power management.
MS at this stage is not tooled up correctly.
I agree that Silverlight failed as web platform.
But you don’t have any proof that it eats up battery life on WP7. Infact WP7 battery life appears to be similar to all other phones out there.
And as for your statement that Google went the Dalvik route because of battery life concerns, you have not shown any proof to back that up.
You don’t have to look to far. And its not the Dalvik where most of the alterations are.
I said altered Java that not just referring to the JVM. Dalvik improves memory effectiveness compared to the standard JVM and does help a little with powereffectiveness.
Dalvik itself does take a power hit. Little bit of coding for android and you see where android is different to normal Java and almost every major difference links back to improved power usage.
Most WP7 apps are not written in Silverlight.
XNA and XAML are not Silverlight. So of course Windows Phone 7 is going to have more apps sooner than Android. Android started off with basically a complete blank slate with a highly altered API not compadible with normal Java applications.
Now XNA and XAML. XNA was designed for games. Was not designed with the idea of being power effective. XAML is about as messy as the Android graphical solution. So in my eye Silverlight is dead MS does not want to own upto it so is trying to relabel XNA and XAML as it.
Yes WP7 phone is equally power effective as an android device until it starts having to run multi apps. Its also the reason why android has apps to measure the power effectiveness of applications so users can choose better power designed applications. Ie android better coded applications are more powereffective. WP7 all your apps are basically the same crap level of power effectiveness that lines up near the bottom edge of android apps.
To put it simply, Android just sucks at power management.
Dalvik sucks just as much as java SE, or worse. There is already a standard derivation of java SE tuned for mobile devices and that is called J2ME. Jazelle is a JVM that is wired right inside the hardware. Jazelle is an order of magnitude better than Dalvik.
Google didn’t implement Dalvik for power efficiency, or if they did they did, they failed big time and they have the stupidest engineers on earth.
Android sucks battery life like there is no tomorrow. Very few (any?) Android device can last more than 2 days. Some Symbian devices last several weeks. And Android is not even comparable to Symbian in features. It can be compared to S40, which is also much more efficient both in power management and in performance.
Edited 2010-12-22 09:07 UTC
Almost all of them do.
The exception being OEM apps which use Silverlight and native code interop to do their thing.
Of the 4,500 apps, 99% of them use either pure Silverlight or pure XNA which is also .NET.
So? You can’t dismiss a competitor advantage because its an advantage ..
Yes it was. The .NET Compact Framework was designed to be power effective , which XNA builds on.
Stop making shit up. None of this is true. You’re throwing a bunch of acronyms and words together into an incoherent thicket of bullshit. Don’t do this.
Again, you make absolutely no sense. Your lack of understanding, which is probably willful ignorance is egregious. Please, please, research what you speak about.
As a Silverlight dev, I can write a faster, prettier , more functional app in less than half the time it takes anyone else on any platform.
WP7 is .NET nirvana. It has an army of devs behind it. MSFT has the most vibrant dev ecosystem on the planet with .NET, and at this rate, we will outpace the other platforms faster than Android did iOS
…because you’re familiar with it and it has a complete graphics library.
A Java dev might say the same thing.
For a custom server-side application, a CL dev might say the same thing…
If you’re familiar with the platform, and it’s got the needed libraries, you can work faster on that platform than any other.
That’s just… basic logic.
For art, I know clay and glazes the best, so I can work fastest with those tools.
No tool is the best tool, the best tool for the job varies from job-to-job, person-to-person.
Absolutely right, to an extent, familiarity helps, but especially with Java you’re going to run into some limits as to what you can do with the platform. There’s really not a presentation platform on Android anywhere near as rich as Silverlight, and this is coming from someone who desperately wishes there was.
Don’t get me wrong, Android is very nice and the platform is very well thought out, but it’s just not the same.
Who said anything about Android?
I did? Is this a trick question?
*sigh*
This is a common idiom in the English-speaking world, indicating confusion at a non-sequitur.
I really don’t think its too far of a reach to name another mobile platform in a discussion about mobile platforms. Especially since I first made the assertion that I am more productive with one, having developed for three. Not by a small margin either.
Additionally, it just speaks to the point about knowing the language and the platform well, but still running into practical limits of the platform. I could just as easily have named iOS or to a much lesser extent, webOS.
When did this become a discussion about mobile platforms?
I thought we were talking about _development_ platforms and operating systems.
Even that’s a digression from operating systems and processor architectures.
You mentioned silverlight, I mentioned Java.
No one brought up Android/Dalvik until you stuck it in there like the stuffy-guy it is.
Edited 2010-12-22 23:34 UTC
I first mentioned them in an off the cuff remark about Silverlight at the end of my first post. In fact, it’s the part of my post you ran with.
Now, why are you wasting time with technicalities and semantics instead of commenting on the substance of my post.
Because I only replied to a single sentence. A sentence which states a fallacy born of familiarity and comfort.
You’re the only one self-important enough to be saying ” NONONO!! I WANNA TALK ABOUT THE THING I WANNA TALK ABOUT! NYEH!”
I disagree, and I’ve explained why, which you’ve neglected to address.
No, you persist in going on about specificities and Dalvik VS WP7, when I’m talking of the generalities of development platforms.
I punched a gigantic hole in one of your arguments, so you decided to pretend I was talking about something else.
Charyou-tree. Come, reap.
No, I won’t explain my abstruse references. You’re in front of a computer, you have a world of information to access.
Edited 2010-12-23 00:16 UTC
No, the premise of the counterargument I made is simple. Familiarity only helps productivity to an extent, after that limit is reached, you run into the inherent problems in the platform itself.
My argument was that using Silverlight, the threshold for the limitation is much higher, and as a result I am more productive with Silverlight compared to iOS, Android, and webOS (which I also am very familiar with).
Also, Dalvik is the name of the VM for Google’s Java implementation .. it’s not the name of the Android platform, and in fact, my gripes are not even with the language so much as they are with the libraries.
Yeah, you’re so familiar with them that you didn’t even know that Dalvik/Android libs aren’t the same as Java libs.
That the API is different in numerous ways…
Look out, billy goats, this one’s gotten loose.
Where did I say that? Lol. Also, while its true some of the libraries are different, there’s also a lot of libraries that are mostly the same.
That is, the core Java class libraries are mostly the same with small changes, and the Android SDK proper is completely different, for the better.
Dalvik is just the bytecode format/VM and Jitter that power Android’s Java .. , it’s akin to calling C# “IL”. Stop doing that.
I was talking about _Java_. You know, the language with a large standard library and cross-platform implementations? Android/Dalvik are similar to Java, most of the syntax is the same.
However, it’s not Java.
Red hands, charyou-tree.
The VM works in a totally different manner, so code optimisations can and should be done differently.
Again, this should be obvious.
You’re either a poseur, (I’ve never claimed to be anything but a collector of facts, to note) or a troll.
Well, perhaps you’ve forgotten the wisdom of Sow Crates.
“That’s us, dude!”
I never claimed it was Java? Wtf? I corrected you in saying that they didnt share a set of standard libraries, because even having an elementary understanding of the Android platform, you’d see that they actually do share a lot of similarities.
I never compared it to Sun’s VM, I merely stated that Dalvik was the name of the VM and bytecode, so calling it “Dalvik/Android” is annoyingly stupid.
I thought I alluded to it enough in a post (which you no doubt neglected to read, or couldnt find something to nitpick at so you simply moved on) that I actually do have extensive experience with both the Android SDK proper, and with doing low level interop between native code and Java on Android.
You know, a prerequisite for being a smart ass know it all, is to actually have a clue about what you’re talking about.
That’s admitting differences. Fail.
Stating a difference, when a difference exists is stupid… yeaah…
I’ve been calling this bullshit.
Irony.
Edited 2010-12-23 00:59 UTC
And wake up and smell the nukes. Dalvik/Android would be the correct term for the normal SDK in the Android platform.
This is part of your problem. Talking about Android you would normally not directly point at Dalvik. Instead you would be saying using the Android SDK. Now when not using Dalvik and going direct you are a talking about the Android NDK.
Its like performance modding a car. Each part you add will give a performance boost. But if you add up the parts boost individually you will find that the combined boost to performance is greater.
There is no single bit of magic that explains Android power effectiveness. There are a lot of small bits spreed through the complete SDK and just like car performance moding the improvement they provide is greater than them individually added up would give.
There are even bits spreed threw the NDK that also give power effectiveness.
Dalvik is also highly platform locked. Since it reaches straight into the kernel itself to a api that only exists normally in the android version of the Linux kernel. Yes for power effectiveness reasons kernel powered IPC exist designed particularly for android.
Sorry Nelson trying to find a magic bullet was one of the things that said you were a newbie to embedded development. You learn that there are no magic bullets to power effectiveness just a huge stack of minor appearing bugs that all add up to massive power savings when fixed. The idea that Dalvik was the difference between .net on Windows Phone and Android is basically plan wrong. Dalvik is only a small part of the difference.
Yes power effectiveness does take studying the complete API and looking for the existence of a huge set of features. More of those features the more likely the platform will be power effective. Please note I said likely. This is why optionising for power effective is a pain in ass. Since you might be inside one patch of having power effectiveness but it appears to be worse than where you started from because that last patch add a feature that makes all the other features that you have added work and since they features are not working its now more power eating.
.net in Windows Phone is missing massive numbers of the features required to get power effectiveness.
Fixed.
Problem here. OpenKode. Works on iOS. Also works meego and symbian and windows … Basically everything bar the Xbox 360 (unless you pay for development kit for native program then you can make OpenKode stuff work) Thinking it worth while paying for native on the Xbox 360 since native runs a lot better than XNA, blackbery and the Windows Phone 7.
QT also has the same kind of everywhere coverage and its growing. QT is a broader API than OpenKode.
The simple point here you are productive in the most non cross platform solution. For me spittng out a OS X/Windows/OSi/Windows mobile 6.5/Android/sybmian application is basically code all the same things and just build for each platform.
Blackbery and Windows Phone 7 are the two that are complete crap to me. Since I would have to go out of my way to make custom code to support the platforms.
I don’t call you productive. In fact I call you non productive since you cannot build effectively for 90 percent of the market out there. Sorry 90 percent of desktop is less than 50 percent of the total market out there for graphical apps.
I don’t class webos as a competitor to worry about. Market share is too small. .net skills don’t help you at all with webos. HTML 5 skills do. So to support it you support web users again. Not coding for a single set of platforms from 1 company.
If what you are coding in is not proper cross platform you are wasting your dollars these days.
Except that OpenKode is built upon the same set of dodgy standards which have variable support across devices. I take it you’ve never tried to implement OpenGL ES across more than one Android device? I have. Surely you’re aware of the missing vendor extensions, different texture depths to accomondate, and outright quirky behavior of some popular handsets, right?
You’ve seen the differences in the OGLES implementation of Tegra SoCs vs the Snapdragons or TI SoCs right? No? Okay then.
It’s a pain in the ass, a big fucking pain in the ass.
QT is decent, and I’ve dabbled in Qt Quick, but you’re still going to run into a practical limit of productivity. A lot of the Qt stuff is in its infancy when it comes to rich media presentation, and even so, the maturity of the platforms, not to mention the tooling, is woefully inadequate.
I think that’s a huge problem. Sure, you’re cross platform, but you’re forced to shoe horn experience into a least common denominator across all of them.
You’re not as cross platform as you think, and even if you are, quality suffers.
The slice of the market we target is not as relevant as is the amount of returns we get on our investment. We had to spend very little money to get our existing code to go from working on Windows and OSX to working on Windows Phone 7. So in one month we were able to create a new stream of revenue at a cost of very little man hours.
And not just that, we were able to, due to the power of Silverlight, deliver a consistently compelling experience to our customers.
webOS was actually trivial to do, since web technologies mesh well with how our service pushes data to our clients. Again, a testament to the interoperability of Wcf and its web services.
I think that’s really a cop out, and it’s cutting corners. I’d rather be really good on a handful of platform, than mediocre on all of them.
Our returns relative to our investment are good enough for it to be considered a success story in house.
I have no idea what any of this post is trying to say, because it reads like copypasta from a corporate buzzword cookbook.
Haha, you hang around people in my field you start to use their lingo “experiences, immersive, compelling” but the substance of my post is still sound.
Things may be cross platform, ideally, but they’re also pieces of crap on each and every platform they exist on.
I’d rather write something good on a few platforms vs writing something bad on a bunch of platforms.
And I had the same quirkness with Windows Mobile as well. Its normal chipset hell. Windows Phone 7 will have equal issues from the chipset if it ever gets large enough production. You think .net some how protects you from defective batches of hardware where most of the quirkiness comes from.
Thing is I am tooled up to deal with those differences you are not. Its only a pain in the ass once per problem if you code is implemented right. So all those defects are to me these days is a minor annoyances.
Really I don’t find that the case. Most likely because I do come from a KDE background and know what I can pinch from it to cover all the weaknesses .net users have ever listed to me. QT is only the starting point there are a lot of libraries that add features on top of it. Just like .net there are a lot of features that if you don’t know the right library to add you cannot do simply.
Really sorry to say it does not. But yes it does require being prepared to spend some money in the past years on decent compilers that could solve out C++ so removing most of the QT bloat and any other feature interface wrapper you do. So its not shoe horn experience. Its the simple case that you have not been doing this so its scary. At lest the next generation Gcc will near decently optimize C++ so I might be able to stop paying for commercial compliers.
And you mentioned half assed crossplatform before. Not dealing with vendor extensions to hardware when doing particular operations means that using .Net you don’t use the hardware correctly. The exact thing you accused me of earlier. You don’t have to use vendor extentions to opengl se to go cross platform. You many need to use them to correct a defective chipset. Now if that is under windows phone 7 the device is basically lemon since the control options will not be exposed. Yes this is another reason why I want Windows Phone 7 dead. Its going to come a big problem as the number of chipsets windows phone 7 increase due to the general chipset defectiveness.
When android was first release we had a nice holiday of no chipset hardware issues. Same with when Windows CE was first released. But with years you know in 12 months the normal hell will be back. Only platforms were we normally don’t see this. Is like blackberry and iphone. Where the OS maker is making the device and does not cut corners other wise the dollar signs tempt the makers to use cheaper chips.
Thinking I have been generating a side stream on sybmbian and Windows mobile from before .net existed with very little effort on my part. And the future holds for me android iOS and meego as my side streams in the mobile world. I am really annoyed to lose phone due to MS idea of not providing native API’s or means to build C++ effectively.
OS X and Windows production is a basic given. Linux platforms where requested and without legal questions.
Yes I have years of it and my in house stuff is very advanced at dealing with the problems.
Remember webOS is a platform by HP. And what you just described does not work with it well at all. No .net at all and applications have to be able to work without web access since they might be a printer or something else strange. I guess you are missing a space. Web OS the term. Not webos the trademarked OS name for the decent-ant of palm os’s. That HP is talking about releasing phones containing.
Yes I hate terms that a insanely close to.
GTK and QT both have worked out there mistake on missing website support and are working on correcting that. But for websites I normally use PHP anyhow. Hosting is cheaper.
Really all you are developing is mediocre on a few of them. Where I can develop mediocre on most of them quickly. Focus a bit deeper and and improve particular ones where the most customers are. Yes the very thing you first complained about allows me to focus deeper. Basically you are in a younger firm without enough experience to know better.
I should have read this first. Really Android NDK is currently richer than Silverlight due to the toolkits it can host. Code in Java like language is not your only option. Existing QT in the NDK is way more rich than Silverlight presentation platform.
Interesting that you were bagging that it was not ready. All it proves to me is that you had not tried it. Also proves that you don’t know Android. Android has two levels NDK and Android form of the Java language.
Android is well thought out. Java like language for simple applications. NDK for complex ones needing fancy displaying. The NDK is well on its way to matching the requirements of http://www.khronos.org/openkode/ opengl es, opensl es and openvg are all exposed only openmax is missing for android ndk to be a openkode platform,
Yes native code for android is mostly not as platform locked as it use on almost every platform that allows building of native binaries that has openkode.
Extra note. http://code.google.com/p/nativeclient/ is also almost an openkode platform. So really NDK android is more a web platform than what silverlight is these days
Edited 2010-12-22 22:57 UTC
You do realize that you sound pretty desperate, don’t you?
People with a time / technology horizon that goes beyond the walled landscape of MS will not care much about stellar productivity promises: they have never materialize and neither WP7 nor Silverlight is any different.
Really? We went from drawing board to app store in less than a month and are getting some decent returns out of an ad supported application. You cant tell me that there are no productivity gains in developers being able to use familiar languages and frameworks with competent design and development tools.
I’ll be waiting in the real world when you decide to join it.
In fact, we achieved a pretty decent amount of code sharing from our desktop solution due to the common language and framework. Basically all of our backend model and binding/validation logic was reused, we only wrote a new UI to accomondate the smaller display and look and feel of WP7.
You can doubt, you can deny, but my checks sure seem to materialize in my bank account.
Yes, ability to easily create revenue is indeed the best way of measuring the quality of a software platform.
iFart is proof that ObjC/XCode is the most fantastic development language/environment.
O_O
I don’t much like Silverlight either; yet another Flash/Java runtime compiler to have installed offering vulnerabilities.. however, “dead horse” is a bit exaggerated. It’s currently the best performer for streaming media; outperforms Flash I hear. You’ll not be watching your Netflix without it. I’m also seeing it required for hosted conference services. It’s also a microsoft product, you know they’re going to push it as hard as .NET.. MS already has it required for enterprise apps.
I do hope it lays down beside old Bessy but it seems alive and well so far.
Where have you been Google aquire Netflix drm provider.
Yes You can view Netflix without .Silverlight it works on the PS3 does it not. Normal MS troll without out of date information.
Honestly you don’t want to run the crapware you find on iTune or Android market on any platform either.
Edited 2010-12-22 09:11 UTC
The app part is why they developed .net (and to compete with Java of course), technically all the should have to do is port the Compact Edition .Net framework from whatever phones it’s already running on to this.
Edited 2010-12-22 01:34 UTC
No devices? Seems to me that is why they are doing this. The hardware market has clearly declared ARM the platform of choice for tablets. MS wants to run on those tablets.
It seems more like a pragmatic response to the market than anything.
No applications? Well, the Win32 API is the most supported platform on earth. Many apps and libraries might work with a simple recompile. There are millions of developers with the skills to target Windows on ARM.
Microsoft may even have the chance to distribute ARM apps through an app store like Apple has done. This would be a whole new revenue stream.
No drivers? Why would a wireless card in a tablet need a different driver than the one in a laptop? Even if they have to be written, the Windows driver model is not exactly uncharted territory for hardware vendors. I bet drivers will not be a problem.
It would be a huge win for Microsoft if they could retire Windows CE and deploy a common code-base from tablets up (as Linux has done). They retired the Windows 95 code and moved everything to Windows NT (Windows 2000) when desktops became powerful enough. This could be much the same.
Remember, the NT core used to ship on x86, MIPS, PowerPC, and Alpha. Those are the ones I can remember anyway. It has run on PowerPC more recently in gaming. So, it seems like cross-platform capability has really been part of the plan for a very long time.
Also, this could be a warning shot across Intel’s bow. Intel has been toying with Linux which reduces the power of the Wintel duopoly.
Microsoft using their market dominance to reduce Intel’s market power long before Intel has a chance to use theirs against Microsoft is a proactive move, not a desperate one.
All-in-all, I cannot think of a reason not to do it. What exactly are they risking? Desperate moves are usually not nothing but potential upside for the party making them. What is the downside?
Well a fairly recent NT kernel currently runs on a PowerPC core in the XBox 360, so it seems Microsoft have been quite careful to maintain that portability over the years.
You’re confused. The original Xbox is x86. The 360 is PowerPC, but doesn’t use any NT code as far as I know (it’s entirely custom).
My understanding is that it runs a heavily modified NT 5 kernel.
Anyway, Windows Server 2008 R2 runs on Itanium, so it’s a better, more recent example. IIRC, Windows needs a CPU capable of running in little-endian mode, but that’s the main restriction.
The 360 runs a modified NT kernel.
I’m pretty sure it doesn’t.
http://blogs.msdn.com/b/xboxteam/archive/2006/02/17/534421.aspx
Sounds to me like a modified NT. It’s not running Windows 2000, but it is definitely NT, albeit heavily customized. The article only states that it is not Windows 2000, but a different OS built specifically for the XBox. So a NT-based customized OS.
Yeah… But didn’t they just do major changes of NT kernel in Windows Vista/7? I mean, a modified NT5 is not NT7(is it 7 on Win7? I’m on Linux, so I can’t check).
Here’s a break down of NT version numbers:
NT 5.0 -> Windows 2000 / Windows 2000 server
NT 5.1 -> Windows XP
NT 5.2 -> Windows 2003/Windows XP x64
NT 6.0 -> Windows Vista / Windows 2008
NT 6.1 -> Windows 7 / Windows 2008 R2
Windows 7 would probably have been called “Vista R2” but given the damage done to Vista name it made since to use a completely new name.
Really notice that Windows 2008 users did not have to pay to get 2008 R2.
Yet poor people who paid for Vista have to pay to get 7.
Its normal MS money grubing double standards.
So Linux/*BSD running WINE must be a modified NT too.
Yes, so you are wrong then Thom – that article implies it shares common heritage with the Windows codebase. It might not be anywhere close to the current NT Kernel, but it is based on a common ancestor. Or you believe that Microsoft threw everything away, started again with a clean slate and then reimplemented the same API?
Euh?
Wouldn’t be the first time.
By my (rough) count, we have two separate implementations of Win16 (Windows, and Windows NT), and at least five separate implementations of Win32 (Windows NT, Win32s, Windows 9x, Windows CE, and Xbox / Xbox 360).
The Xbox / Xbox 360 OS only implements a tiny part of the same API as desktop Windows does. It only needs the really low-level stuff, like the filesystem or memory management.
Even those parts have to work differently on a console. For example, games don’t just have direct access to the filesystem. They only get access to the disc (if the OS allows it), and a small part of the hard drive to store save games. As another example, there’s no virtual memory on an Xbox.
It’s really not worth it. Most of the features Windows NT brings are irrelevant at best, or pure overhead at worst. Basing the Xbox OS on Windows NT just doesn’t make sense.
Instead, they implemented a compatible low-level API, ported bits and pieces (like the C / C++ runtime, or .NET) over to the Xbox OS, and implemented their own APIs (the Xbox 360 version of Direct3D, XInput, XAudio) on top of that new OS.
Makes sense to me.
Or not.
“According to Microsoft, it is a common misconception that the Xbox and Xbox 360 use a modified Windows 2000 kernel. They claim that the Xbox operating system was built from scratch but implements a subset of Windows APIs. The idea that it does, indeed, run a modified copy of the Windows kernel still persists in the community.”
http://en.wikipedia.org/wiki/Windows_NT
According to me the idea that Microsoft wrote a new operating system for the Xbox (which is identical to an x86 PC) is ridiculous.
It is the only option actually.
Game machine doesn’t need thick HAL or drivers. Everything is fixed and works directly with hardware.
OS is tuned for performance, memory consumption.
It has no swap. It must occupy fixed, predictable amount of ram.
You can’t just allocate “a couple more of megabytes” YOU FAILED. You can’t consume much more CPU time – YOU FAILED.
Edited 2010-12-22 13:19 UTC
Why is it ridiculous? When they already had DOS/Windows, they wrote a new OS (WindowsNT) for higher-end, 32-bit systems.
For palm-top computers, they wrote a new OS (Windows CE) rather than trim-down an existing OS. (according to the wikipedia entry)
The NT kernel has way too much abstraction to be useful for a console system. Microsoft is also a very large company, with more than enough resources to spin a new kernel, especially a small one useful for a console. I’d be willing to wager it would be more work to make NT suitable for the XBox than just write a new kernel for it, especially when the XBox came out.
Yes, I remember a fellow classmate who was perfectly happy running NT4 on aplha. The Poor Bastard.
Please go and look at the meego and android sites about targeted platforms.
You will notice these OS’s are attempting to create 1 OS for everything.
MS will be forced to play the same game.
AFAIK, CE has been running on ARM for a long time, in the Zune for example. Since they’ve just released their new phone OS which is probably just a new CE with a new user interface (just speculating here) then it seems that a big announcement would be either a standard Windows desktop OS or a netbook/tablet version of win phone 7 because Windows 7 is just not able to power an ipad competitor.
In any case, they already have experience on the platform and with the .Net framework they also have an excellent API that a lot of developers are familiar with. Plus Visual Studio is one of the best IDEs around and it already supports several platforms. It would probably not be long before apps started showing up.
We have to see what really happens but I’d not call this a failure just yet.
Wake me when .net gets retooled to allow power effective optionals. Visual Studio is not the best IDE around for crossplatform development. That is a toss up between QT Designer Kdevelop and Eclipse.
Number of platforms Visual Studio supports is basically nothing.
Lets not forget that you can develop on Windows Mobile, Windows Phone 7, Zune, Xbox 360 (via XNA) and then there is the .Net Micro-framework, which can run on a wristwatch.
Anyway, the original posters point was that there are plenty of developers with relevant experience already for .Net, which will certainly be included with any ARM version of Windows.
So the development skills are already there.
Again compared to modern market basically nothing.
Qt does Windows Mobile so Kdevelop does as well. Zune is basically a dead product. Zune is basically been watered down into being a Music store for Windows Phone 7.
Xbox 360 XNA and Windows Phone 7 are basically the same platform.
Qt does iPhone applications. Qt can also be made to do Android based devices. Same code on all 3 with no alterations.
If nokia is going to support Windows Phone 7. Thinking Qt is nokia product. Windows Phone 7 better find a way of supporting QT. Notice some here. Being nokia friendly removes any reason to be developing .net applications.
Lucas_maximus sorry the say there is also a huge army of QT developers.
No the claim was not that there was plenty of .net developers out there. The claim what that Visual Studio was the best tool for cross platform development. This is clearly false.
Yes there are tones of .net developers out there. Just like there where tones of Visaul basic developers out there. When we are talking embedded you need quality not numbers. Visual Basically developers were the worst to higher for embedded work if you wanted battery live and .net developers are not much better.
At no point did he claim it was. He just said it was the best development IDE … something which I happen to agree with. Eclipse isn’t anywhere near as good IMO, though it is a pretty decent IDE (some really annoying bugs), which ended up with me dropping into notepad.
I haven’t used the IDE for QT so I couldn’t possibly comment.
Just because someone develops in Visual Basic (I did because my company happen to have everything written in VB.NET) does not mean they are a crap developer.
Also could you please check your spelling and grammar in future because it is awful.
I didn’t say that VS was the best IDE for cross platform, I said it’s one of the best around. It’s also highly extensible and if MS decides that it’s important for VS to support something, they definitely have the ability to do it right.
I’m not sure why you have a fixation on embedded though, if you reread my original post, we know MS already have an embedded OS so if there is a big announcement then it’s unlikely to be about embedded. Or it’s unlikely to be a big announcement.
Also your claim about many devs vs. good devs has little backing. Think about all of the people developing for iphone or android, I doubt they are all experts about mobile development and battery friendly techniques. Yet it’s the amount of apps available that’s propping those two platforms to the top. It doesn’t take a genius to write a fart app. If you look at android, they’ve achieved large speed and battery usage improvement in version 2.2 and that was an update of the system, not the apps.
Finally you seem to have a fixation on Qt and that’s cool, but it seems that it’s a different thing from .Net. You’re talking about Nokia and all of that but MS has already released their phone OS and the development tools are already there. I see little reason for them to listen to you and migrate everything to Qt.
All I was saying is there are quite a lot of .Net developers around and it makes sense that MS would count on them to try and quickly get apps out for some hypothetical new Windows on ARM.
“Qt does iPhone applications. Qt can also be made to do Android based devices.”
What? That’s just bogus. Qt may run on Android at somepoint, but right now it’s extremely pre-alpha. I wouldn’t call it “running on Android” at all.
Android QT surprising stable. I am running the development branch for android. Note I said made that does not mean recommended.
Android QT at this stage is just the qt embemded branch with a wrapper in native Android byte code so it accepts it.
The embemded branch of QT is only 6 years old so is highly stable and mature. Biggest issue is that its power hungery compared to normal Android apps. But compared to .net applications on mobile devieces about the same.
Bogus claim without trying it and knowing what it is. tbcpp. http://code.google.com/p/android-lighthouse/issues/list 24 outstanding bugs. Out of those ones critical go well behaving android apps.
21 QtMobility Lack of this is why well designed QT apps are power vampires.
55 57 both should be fixed but can be worked around both releate to incorrect screen size information getting into QT and are basically the same bug.
1 Virtual Keyboard one Yes its nice to use the Android virtual keyboard. But QT can provide its own.
4 Use opengl ES for acceleration was not able due to Opengl ES not being exposed to native apps. Fixed in last lot of google sdk updates. Really that bug should be closed soon. Since from my last build with opengl on it works perfectly. But I might have missed testing something.
5 Audio fault also gone in the last google sdk update providing OpenSL that QT already knows from other platforms.
And if you think 34 is important read the bug.
Yes the bug list is running a little behind. The amount working is more Beta level with 2 major bugs left. Cut and paste works. So compared to what MS calls production release Windows Mobile phone QT for android passes.
Bullshit, the best IDE is Emacs.
Whatever floats your boat dude
CE and WP7 are awful. Wake me when they have real options that are not a joke.
I am not very knowledgeable in this area, so I am just kind shooting in the dark here, but is it possible for them to run windows 7 and native apps on some type of hypervisor? I dont know if it’s possible to emulate or virtualize x86 code on ARM, but that would be interesting.
Microsoft may not be the best in the virtualization sector, but the hypervisor technology in Windows 2008 is actually pretty good and is not too resource hungry.
I am not very knowledgeable in this area, so I am just kind shooting in the dark here, but is it possible for them to run windows 7 and native apps on some type of hypervisor? I dont know if it’s possible to emulate or virtualize x86 code on ARM, but that would be interesting.
Of course it is _possible_ to emulate one or another architechture on it, but it’s simply not feasible. It requires a sh*tload of CPU-time meaning the apps wouldn’t run all that well and that it’d eat battery like there was no tomorrow. The less the application actually does and the simpler it was then sure, the less it’d require, but as soon as the app is more complicated than a simple calculator the system would have to spend way too much time translating the code that it’d bog everything down.
No. Emulation is possible, but not virtualization. Consider virtualization to be “hardware-accelerated emulation”, and x86 doesn’t have any built-in hardware acceleration to emulate ARM. The virtualization you know is emulating x86 on x86, which *is* hardware-accelerated, for example using similar features that OSes normally use to control userland applications.
No. Emulation is possible, but not virtualization. Consider virtualization to be “hardware-accelerated emulation”, and x86 doesn’t have any built-in hardware acceleration to emulate ARM. The virtualization you know is emulating x86 on x86, which *is* hardware-accelerated, for example using similar features that OSes normally use to control userland applications.
To nitpick, one could virtualize parts of the OS: like for example redirect calls to OS-specific libraries and system calls to actual ARM equivalents so that only the application code needs to be translated, not the whole OS. But that’s different kind of virtualization, and while entirely possible it still wouldn’t be in any way or form feasible on a mobile platform.
There actually exists a somewhat similar system for the N900; some people have one or another need to run an x86-specific application and use qemu to translate the application code on the fly but the application itself has access to the underlying OS, instead of the whole OS running inside qemu. The problem is that while it works, it’s really slow and hogs CPU, thus hogging battery.
Just pointing these things out in case the OP is interested in knowing — or someone else for that matter — not to insult you or insinuate anything.
No offense taken
In fact, running native-ported libraries with an emulated or software-translated application was new to me, although in hindsight it is obvious… another tool in my toolbox, should I ever need it.
Look at qemu, it provides some support for that on the syscall level, bringing with it reasonable speed for many applications.
Automated thunking of libraries between architectures is much harder, because pesky issues like data structure alignment come into play then.
WINE actually has played with this. WINE is a server/client model. The client of course needs to run in a x86 -> ARM emulator/translator. But the server can be native ARM.
The Wall Street Journal also has an article about this and their article says it is two years away from release. I don’t know if that was a mistake but if it’s true then I think it will be way too late. http://online.wsj.com/article/SB10001424052748704851204576034051605…
If Windows on ARM is due for release two years away, then IMO it is not targetted at the mobile space at all … it is targetted at the server space.
The ARM Cortex-A15 MPcore component is also a couple of years away, and that component is also targetted at desktops and servers.
http://www.engadget.com/2010/09/09/arm-reveals-eagle-core-as-cortex…
Cortex-A15 MPcore will be capable of quad-core computing at up to 2.5GHz, at much lower power consumption levels than comparable x86/x86_64 chipsets today. Server machines, which are on all the time, will be designed with one or two or even more of the ARM Cortex-A15 MPcore CPUs. Linux as it is today will run like a dream on these machines of the near future. If Windows cannot run on such machines, Microsoft will virtually hand the server market over to Linux.
PS: Linux servers today can run Samba 4, Alfresco, CUPS and OpenChange.
http://www.alfresco.com/
http://www.cups.org/documentation.php/network.html
http://www.computerworld.com.au/article/273515/active_directory_com…
http://www.openchange.org/index.php/home/what-is-openchange
So, IMO, Windows-on-ARM is actually targetted at the low-power high-performance quad core Cortex-A15 MPcore based servers that will be available on the market in a few years time.
Edited 2010-12-22 12:23 UTC
Yes, that makes a hell of a lot more sense to me as well (I’ll write a new article on this today, with the WSJ article and your comment as sources). I mean, I’m sure Microsoft plans to move WP7 to the tablet space, while keeping NT for desktops/laptops/servers. Since a Microsoft server usually consists of Microsoft software (which MS can compile for ARM itself) combined with open source software (which is usually available for ARM already), the application situation wouldn’t apply.
NT on ARM is just that, NT on the ARM architecture. People can do whatever they want with that. You can run it on a server or on a tablet.
Windows 7 already runs on tablets. Some people think that you don’t need Windows software on a tablet but that is wrong. I know people who bought tablets just to run Photoshop on it.
And how are these people faring?
Very well actually. This is how you should use Photoshop. Photoshop works better on a tablet with a stylus than on a screen with a mouse. I believe a tablet with Photoshop is pretty much a must have for any artist.
Edited 2010-12-22 15:01 UTC
I second that. Properly set up graphics editors don’t require a keyboard nor a mouse to work since Windows got a proper visual keyboard which shows up automatically when entering a text input field (was it in Vista or 7 ?). A pen tablet is sufficient. In fact, for some tasks like painting, it’s better.
Then, of course, not all graphics editors are equal in this field. Photoshop’s super-tiny controls are annoying when you want to use it with a tablet, Gimp is better in this area.
Edited 2010-12-22 17:06 UTC
Yeah, but you’re talking about the early tablets that usually have a stylus and are supposed to be used as input devices only like those manufactured by Wacom and Genius unlike the current tablets that are akin to a small computer with a touch display even though some of the products of the former category do have displays in its higher and more expensive range. These are completely different devices intended for different use cases which I believe is the point that the parent poster was trying to make.
I can imagine that with the tablet version of an application like ArtRage an artist can accomplish a lot and I have even seen some of my favorite comic book artists such as Jim Lee taking a shot at it with surprisingly good results for pictures that look like finger painting! But make no mistake, these tablets are meant to be consumption devices.
I have yet to see anyone claiming that he/she bought an iPad to work with Photoshop in it…
Edited 2010-12-22 17:15 UTC
The problem is that everybody thinks a tablet is an iPad.
I was talking about tablet PC. There are many many different models in different shapes. There are convertible tablets like the lenovo thinkpad X60, there are small tablets like tha Archos 9, tiny tablets like the N900 and the ones you talk about that are only input devices.
Windows on ARM may be used in servers, desktops, tablets or your fridge. My point was that Microsoft designs and sells an OS. It does not sell a final product. Well, it does, but that is not its main business. I was replying to the post that said it was designed for server and not for tablet. Windows does already runs on x86 tablets. If they make an ARM version, why not run it on a tablet? (Not talking about the iPad at all here, the iPad runs iOS and you would be sued to hell by Apple if you tried to ran anything else on it)
Then this world would finally get rid of this horrible architecture called x86 (Since the whole world basically runs on microsoft products. Sad but true).
And now excuse me. I’ll go dreaming of ARM based servers, running silently in the server room. Sans Windows, of course.
Edited 2010-12-22 09:00 UTC
I have a gigantic UltraSparc IIIi server in the other room.
Sucks power, so I don’t run it; I just… hug it from time to time.
😀 I know how you feel… Oh I SO wish I had one of the old uVAX II machines from the early ’90’s!! I’d hug it every day, too!
I wish I had room for a PDP-11 setup… It’s just a cool architecture, historically.
Well, x86 has its quirks, for sure. Enough to be called “horrible”, I’m not sure, especially since AMD made some good spring cleanup job with AMD64.
The more I learn about other architectures, the less I’m convinced that everything would be so much better if we suddenly switched to something else. Some things would certainly get better (like power management), but some would also get worse (not all architectures offer automatic TLB and cache management, OSs would become harder to write without this kind of help from the hardware). It wouldn’t be a perfect win-win situation in an ideal world.
Edited 2010-12-22 13:08 UTC
Android, IOS, QNX, and Palm Pre OS stand unimpressed, better late than never I suppose
x86 application might be run through virtualisation or MS might release a type of QT like development platform.
ARM based processors is fast enough and will have more memory support with 64bit versions coming out in a few weeks.
http://www.zdnet.co.uk/news/processors/2001/10/16/arm-to-unveil-64-…
Intel and AMD have a lot of x86 patents though and might not like it.
off topic. If this turns out to be full windows 7 support Nvidia would at last be able to shift their business from discrete graphics to CPU’s also. It always had this ambition but could’nt.
http://arstechnica.com/hardware/news/2009/02/intel-sues-nvidia-over…
The company could’nt compete on the x86 market because Intel and AMD patents and new GPU integration into CPU’s could spell the beginning of hard times for discrete graphics in years to come.
http://www.zdnet.com/blog/computers/nvidia-to-make-the-case-for-dis…
I was thinking about this already by first Android phones released with Arm like cpu.
To me it is inevitable that biggest sufferer is Intel,
and it is no strange that life plays this kind of twists.
Intel newer felt the breath on the neck, and suddenly it is here, and it looks to me Intel has lost the game in the long turn.
While Intel was doubling the Hz and GHz, and at some point artificially stalling the progress, going form standard PC to mobile, Arm has gone different approach from embedded, industrial and military to consumer and mobile. Nor physical space, nor power consumption and heat, nor performance on limited physical space was ever on Intel agenda.
Arm has developed the performance CPU platform on small space via algorithms, on limited power resources and has wide range of applications all the line from military grade to consumption market.
To me any android phone could be extended to a slate or tablet and further to the ordinary notebook. I do not see the laptop of net 18 – 24 – 36 months with the set of Intel chips, but Arm.
Can not wait ho hold my future notebook with some 16 mobile work hours.
Why Intel is not going to the mobile market? It is simple.
It has nothing really to offer.
In addition Interview with arm CEO you can find on http://www.cio.com.
“Very light on details, but this is interesting nonetheless – very interesting, and potentially one of the biggest things to have hit the operating systems business this decade.”
Do you mean compared to IOS? That’s a joke right? I mean, more than joke that Ballmer is…
Steve Ballmer is no joke
Most people will never use iOS. Most people use, have used or will use Windows.
Making that platform available for a new CPU architecture is a bigger deal than a phone OS that shipped with nice launcher icons and good framerate. It would “validate” ARM for many desktop/laptop use cases (e.g. for people that just need MSOffice).
Fact is that most people use more then one OS. It’s just your narrow historic view that rates desktop OSes as more important than their smaller brethren running on phones, mp3-players, tablets, PDAs, etc.
But luckily for users, technology instead of user centric classical desktop OSes are loosing ground in more and more use cases. Gone are the days where you need your tech savvy neighbour for everyday tasks just because the OS sucks ..
And thus, a rumor about a dinosaur OS extending to another CPU architecture is maybe interesting but surely not groundbreaking or the biggest thing in a decade. The coming of iOS and Android are ..
Windows CE was always there for ARM processors, whats the big news in this? They want to compete with nokia and the others now lol? I cant imagine my n900 multibooting with grub to windows :S
you have tunnelvision
I really fail to see what should be ground breaking about that? Maybe for a Windows fanboy realizing that there’s a world outside x86.
Going with an OS and the whole application landscape from PPC to x86 without much trouble or from x86 to x64 without hampering users is ground breaking ..
(And will we see ‘one of the biggest thing since bread and butter’ headlines in OS news from Thom whenever Linux compiles and runs on another hardware platform?)
How about you stop being a dick and just look at what;s (maybe) happening here: Microsoft breaking up WinTel (or WinAMD). THAT is the big news, and yes, that is VERY BIG. It’s not the technological achievement – it’s the symbolic value of it all. Windows NT on ARM, as a real product?
After all these years – that’s massive.
Why? Because is looks like an act of desperation, catching once more a train that has already left the station?
Not impressed. Not even by the symbolism ..
While I stand by the meaning of my statements, I also admit that by joining in a battle of ‘wits’ with this humanoid I have dragged this thread down into madness and nonsense.
A manic depressive on steroids (eczema and muscle spasms) isn’t the best of judgment, though I did have my facts straight.
Ah, being both obtuse and abstruse simultaneously.
*kisses fingertips and opens hand*
Lovely.
Why do people vote some things down…
An apology to all but the fool… sure, why not…