Re: Windows for ARM? That makes life a bit more interesting...

NTDEV
1

My guess is that these will essentially be closed platforms, with the
only customization possible being an ‘app store’ facility.

Mark Roddy

On Wed, Dec 22, 2010 at 4:09 PM, Tim Roberts wrote:
> Gregory G Dyess wrote:
>> They would really have to clean up the code and remove a lot of bloat and overhead. ?WindowsCE/Windows Mobile run well on ARM and support a very similar API, but is much lighter weight under the hood.
>
> I don’t think so. ?The ARM is a perfectly capable processor, more than
> enough to host desktop Windows. ?It will, however, be a day of
> revelation for those driver and application developers who have
> forgotten the lessons of the Alpha, MIPS, and PowerPC days of NT…
>
> –
> Tim Roberts, xxxxx@probo.com
> Providenza & Boekelheide, Inc.
>
>
> —
> NTDEV is sponsored by OSR
>
> For our schedule of WDF, WDM, debugging and other seminars visit:
> http://www.osr.com/seminars
>
> To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer
>

2

I’m actually shocked.

Mark Roddy

On Wed, Jan 5, 2011 at 5:10 PM, Lee Fisher wrote:
> On 12/22/10 2:19 PM, Tim Roberts wrote:
>>
>> Peter wrote:
>>>
>>> I guess everyone will see in January, at the CES.
>>
>> I guess everyone will see in January, at the CES.
>
> http://www.zdnet.com/blog/microsoft/ces-microsoft-shows-off-windows-8-on-arm/8339
>
> http://www.pcworld.com/businesscenter/article/215663/microsoft_to_put_next_version_of_windows_on_arm_chips.html
>
> —
> NTDEV is sponsored by OSR
>
> For our schedule of WDF, WDM, debugging and other seminars visit:
> http://www.osr.com/seminars
>
> To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
>

3

Why?

(Just curious.)

Jake Oshins
Hyper-V I/O Architect
Windows Kernel Group

This post implies no warranties and confers no rights.

“Mark Roddy” wrote in message news:xxxxx@ntdev…

I’m actually shocked.

Mark Roddy

4

That such an effort was attempted or that it is now in the chute?

I have been bemused by references to ARM in the WDK for quite some time now. The build tools have had ARM options for quite a while. It did not make sense that it was somehow ‘leakage’ in the tool sets from CE as the WDK makes no mention of the myriad other CPUs that CE supports. In some respects the writing for this has been on the wall for quite some time.

I wonder if the seemingly steady effort by MSFT to marginalize third-party driver development is because they don’t think we know how to align data properly :wink:

Cheers,
Dave Cattley

“Mark Roddy” wrote in message news:xxxxx@ntdev…

I’m actually shocked.

Mark Roddy

NTDEV is sponsored by OSR

For our schedule of WDF, WDM, debugging and other seminars visit:
http://www.osr.com/seminars

To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer

5

Microsoft can release a version of Windows on whatever (little-endian) processor they want, but that won’t mean anything if the 3rd parties don’t make applications available.

Remember MIPS, PowerPC and Alpha? The Alpha was arguably a superior processor in terms of performance, but it just couldn’t succeed without 3rd-party apps. Windows, MS Office and SQL server alone will not create commercial success. Granted, MS didn’t do much to push development of software on the non-intel processors back then. Maybe this time they will?

In any event, I welcome anything that offers the hope of openning the stranglehold Intel has over the processor market.

Fun times are coming!

Greg

— xxxxx@msn.com wrote:

From: Dave Cattley
To: “Windows System Software Devs Interest List”
Subject: RE: [ntdev] Windows for ARM? That makes life a bit more interesting…
Date: Thu, 6 Jan 2011 08:41:46 -0500

That such an effort was attempted or that it is now in the chute?

I have been bemused by references to ARM in the WDK for quite some time now. The build tools have had ARM options for quite a while. It did not make sense that it was somehow ‘leakage’ in the tool sets from CE as the WDK makes no mention of the myriad other CPUs that CE supports. In some respects the writing for this has been on the wall for quite some time.

I wonder if the seemingly steady effort by MSFT to marginalize third-party driver development is because they don’t think we know how to align data properly :wink:

Cheers,
Dave Cattley

>
> --------------------------------------------------------------
> “Mark Roddy” wrote in message news:xxxxx@ntdev…
>
> I’m actually shocked.
>
>
> Mark Roddy
>
>
>
>
> —
> NTDEV is sponsored by OSR
>
> For our schedule of WDF, WDM, debugging and other seminars visit:
> http://www.osr.com/seminars
>
> To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer

—
NTDEV is sponsored by OSR

For our schedule of WDF, WDM, debugging and other seminars visit:
http://www.osr.com/seminars

To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer

6

It just seems like a rather stunning about face for a company that
basically tried to kill the netbook as too low end with their silly
windows start edition. Now they want windows on even lower end arm
based phoblets? Plus don’t you have two competing ‘windows’ things for
these platforms now?

Maybe I’m just shocked at how quickly the ‘pc’ is ending its reign.

Mark Roddy

On Thu, Jan 6, 2011 at 12:09 AM, Jake Oshins
wrote:
> Why?
>
> (Just curious.)
>
> Jake Oshins
> Hyper-V I/O Architect
> Windows Kernel Group
>
> This post implies no warranties and confers no rights.
>
> --------------------------------------------------------------
> “Mark Roddy” ?wrote in message news:xxxxx@ntdev…
> I’m actually shocked.
>
>
> Mark Roddy
>
>
>
>
> —
> NTDEV is sponsored by OSR
>
> For our schedule of WDF, WDM, debugging and other seminars visit:
> http://www.osr.com/seminars
>
> To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
>

7

I for one welcome my new ARM overlords.

The question will be how well supported ARM development will be in the WDK
and more importantly windbg. I don’t really foresee many problems with my
source ‘just working’, but learning how to debug a new architecture may be
more challenging.

In the broader sense, so long as Microsoft remembers that having an open
platform for things like 3rd party driver development is a large part of
their success this is nothing but good news for a NT driver dev. It
extends the relevancy of Windows beyond a platform that is fast loosing
relevancy. I would far rather be able to provide solutions for
mobile/embedded platforms than be locked out of iOS or Android due to
policies, licensing politics or lack of standards.

Though the annoucement makes it sound like desktop windows on
mobile/embedded platforms is what we are talking about, it doesn’t seem
entirely unreasonable that we would see something akin to Phone 7’s upper
edge running on server core either. I’m just fine with having the NT
kernel be in more places in any case. The more interesting question will
be what this does to Intel’s Atom which seems to have largely been used by
devices because it could run windows.

t.

On Thu, 6 Jan 2011, Dave Cattley wrote:

That such an effort was attempted or that it is now in the chute?

I have been bemused by references to ARM in the WDK for quite some time now. The build tools have had ARM options for quite a while. It did not make sense that it was somehow ‘leakage’ in the tool sets from CE as the WDK makes no mention of the myriad other CPUs that CE supports. In some respects the writing for this has been on the wall for quite some time.

I wonder if the seemingly steady effort by MSFT to marginalize third-party driver development is because they don’t think we know how to align data properly :wink:

Cheers,
Dave Cattley

>
> --------------------------------------------------------------
> “Mark Roddy” wrote in message news:xxxxx@ntdev…
>
> I’m actually shocked.
>
>
> Mark Roddy
>
>
>
>
> —
> NTDEV is sponsored by OSR
>
> For our schedule of WDF, WDM, debugging and other seminars visit:
> http://www.osr.com/seminars
>
> To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer

NTDEV is sponsored by OSR

For our schedule of WDF, WDM, debugging and other seminars visit:
http://www.osr.com/seminars

To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer

8

That I agree with that: new platforms is nothing but good news for us.

Mark Roddy

On Thu, Jan 6, 2011 at 9:53 AM, Tracy Camp wrote:
> I for one welcome my new ARM overlords.
>
> The question will be how well supported ARM development will be in the WDK
> and more importantly windbg. ?I don’t really foresee many problems with my
> source ‘just working’, but learning how to debug a new architecture may be
> more challenging.
>
> In the broader sense, so long as Microsoft remembers that having an open
> platform for things like 3rd party driver development is a large part of
> their success this is nothing but good news for a NT driver dev. ?It extends
> the relevancy of Windows beyond a platform that is fast loosing relevancy.
> ?I would far rather be able to provide solutions for mobile/embedded
> platforms than be locked out of iOS or Android due to policies, licensing
> politics or lack of standards.
>
> Though the annoucement makes it sound like desktop windows on
> mobile/embedded platforms is what we are talking about, it doesn’t seem
> entirely unreasonable that we would see something akin to Phone 7’s upper
> edge running on server core either. ?I’m just fine with having the NT kernel
> be in more places in any case. ?The more interesting question will be what
> this does to Intel’s Atom which seems to have largely been used by devices
> because it could run windows.
>
> t.
>
> On Thu, 6 Jan 2011, Dave Cattley wrote:
>
>>
>>
>> That such an effort was attempted or that it is now in the chute?
>>
>> I have been bemused by references to ARM in the WDK for quite some time
>> now. ? The build tools have had ARM options for quite a while. ? ?It did not
>> make sense that it was somehow ‘leakage’ in the tool sets from CE as the WDK
>> makes no mention of the myriad other CPUs that CE supports. ? ?In some
>> respects the writing for this has been on the wall for quite some time.
>>
>> I wonder if the seemingly steady effort by MSFT to marginalize third-party
>> driver development is because they don’t think we know how to align data
>> properly ?:wink:
>>
>> Cheers,
>> Dave Cattley
>>
>>>
>>> --------------------------------------------------------------
>>> “Mark Roddy” wrote in message news:xxxxx@ntdev…
>>>
>>> I’m actually shocked.
>>>
>>>
>>> Mark Roddy
>>>
>>>
>>>
>>>
>>> —
>>> NTDEV is sponsored by OSR
>>>
>>> For our schedule of WDF, WDM, debugging and other seminars visit:
>>> http://www.osr.com/seminars
>>>
>>> To unsubscribe, visit the List Server section of OSR Online at
>>> http://www.osronline.com/page.cfm?name=ListServer
>>
>> —
>> NTDEV is sponsored by OSR
>>
>> For our schedule of WDF, WDM, debugging and other seminars visit:
>> http://www.osr.com/seminars
>>
>> To unsubscribe, visit the List Server section of OSR Online at
>> http://www.osronline.com/page.cfm?name=ListServer
>
> —
> NTDEV is sponsored by OSR
>
> For our schedule of WDF, WDM, debugging and other seminars visit:
> http://www.osr.com/seminars
>
> To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
>

9

Fabulous news, actually.

Note that the Microsoft announcement for ARM was limited to supporting ARM SOC systems. So, it’s not clear if we’ll see ARM desktops or workstations (as in NVidia Project Denver) for Win8.

I heartily agree with Mr. Camp’s comments, specifically:

Recall that the vision for Windows has traditionally been “Windows Everywhere” – The whole slate-based computing thing caught Microsoft with its pants down. Going back a bit further, the whole move of doing useful work on smart phones has caught Microsoft with its pants down. Windows CE, even when named Windows embedded, blows. These announcements signal that Win8 will aggressively address the slate/phone market, and return to Windows Everywhere.

OTOH, I can’t HELP but wonder three things:
(a) how many ARM variants Windows 8 will support,
(b) how effectively Win8 will deal with the power challenges inherent in devices that use SOCs,
(c) will it all be too late

There are real iOS slates and phones TODAY. There are real Android phones and slates that use what are essentially Android phone operating systems today, and there will be a real Android operating system for slates tomorrow (Honeycomb). This HAS to make you wonder if Windows 8 will just be too little too late… tiles or no tiles (MoSH)… gestures or no gestures… ARM or no ARM.

I guess we’ll just have to wait and see.

But for US guys, who work with hardware and devices? Bone-up on your ARM architecture, everyone. It’s all good news.

Peter
OSR

10

It may not be too late to be in the SOC space or for that matter into slates/tablet/phone/other gadgets. This space has a steeper positive slope than desktop. Also Apple and Google would love to capture some desktop/laptop area. Their growth is phenomenal - so all in all sounds like a good move.

But the devices that would come from will have to have something extra. I’ve seen and used - iPhone, Android, and latest palm. I’m not a good user, hate computer games, so I ask my kids. Their first choice is palm, then android, then iphone. Not sure why, but they don’t like win phone.

Amazon came with Kindle, now Apple is eating away a bit, and others are getting into this space, even Android OS…

So this never ends, so there is no timeout soon !!!

-pro

On Jan 6, 2011, at 7:56 AM, xxxxx@osr.com wrote:

Fabulous news, actually.

Note that the Microsoft announcement for ARM was limited to supporting ARM SOC systems. So, it’s not clear if we’ll see ARM desktops or workstations (as in NVidia Project Denver) for Win8.

I heartily agree with Mr. Camp’s comments, specifically:

Recall that the vision for Windows has traditionally been “Windows Everywhere” – The whole slate-based computing thing caught Microsoft with its pants down. Going back a bit further, the whole move of doing useful work on smart phones has caught Microsoft with its pants down. Windows CE, even when named Windows embedded, blows. These announcements signal that Win8 will aggressively address the slate/phone market, and return to Windows Everywhere.

OTOH, I can’t HELP but wonder three things:
(a) how many ARM variants Windows 8 will support,
(b) how effectively Win8 will deal with the power challenges inherent in devices that use SOCs,
(c) will it all be too late

There are real iOS slates and phones TODAY. There are real Android phones and slates that use what are essentially Android phone operating systems today, and there will be a real Android operating system for slates tomorrow (Honeycomb). This HAS to make you wonder if Windows 8 will just be too little too late… tiles or no tiles (MoSH)… gestures or no gestures… ARM or no ARM.

I guess we’ll just have to wait and see.

But for US guys, who work with hardware and devices? Bone-up on your ARM architecture, everyone. It’s all good news.

Peter
OSR

NTDEV is sponsored by OSR

For our schedule of WDF, WDM, debugging and other seminars visit:
http://www.osr.com/seminars

To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer

11

I’m with Mark – I’m shocked. It has been 10 years since the last Alpha
support was terminated. The kernel is now 3 or 4 times larger than it
was at that point, and that’s a lot of new code. Even with the very
best of intentions, I have no doubt that a large number of x86
assumptions slipped into the kernel code. The effort required to go
through EVERY kernel function and assess whether it is ARM-safe must
have been enormous.

xxxxx@osr.com wrote:

Recall that the vision for Windows has traditionally been “Windows Everywhere” – The whole slate-based computing thing caught Microsoft with its pants down. Going back a bit further, the whole move of doing useful work on smart phones has caught Microsoft with its pants down. Windows CE, even when named Windows embedded, blows.

I’m not sure I agree that Microsoft was caught with its pants down.
It’s more like Microsoft simply bet on the wrong horse. Microsoft has
been exploring the slate and tablet market for nearly 10 years. I
remember Ballmer trying to take notes on a tablet at an MVP Summit a
bunch of years ago. However, Microsoft treated the form factor as a
limited laptop, and the marketplace was not interested. Apple’s flash
of insight was to treat the form factor as a more powerful cell phone,
and that, apparently, is exactly what the people wanted.

If “Windows 8 ARM” continues the “limited laptop” model, we will have to
see if the world is now ready for that.

OTOH, I can’t HELP but wonder three things:
(a) how many ARM variants Windows 8 will support,
(b) how effectively Win8 will deal with the power challenges inherent in devices that use SOCs,
(c) will it all be too late

(b) and (c) are very interesting questions indeed. Today’s ARM chips
support an unbelievably intricate power management paradigm – FAR more
intricate than any x86 chip. Some of these chips have a dozen or more
separate subsystems, each of which can has an array of different steps
of adjustment in its clock and power levels. The state transition
diagrams are mind-boggling, and managing all of that requires a fair
amount of complicated heuristics. The sheer volume of code in the Linux
kernel that handles ARM power management is enough to make your head spin.

–
Tim Roberts, xxxxx@probo.com
Providenza & Boekelheide, Inc.

12

Basically most of us on this list are “low-level” guys, and there is one
major problem that is an impediment to our clan being successful on these
“new” platforms and OSes:

BRANDING

If Microsoft support of these new platforms is similar to that of CE/Mobile,
(as well as Google/Android) then branding will be done by device
manufacturers and by the network carrier. Part of the “branding” is to make
proprietary modifications to the OS and to place limitations on software
installation. New per-device qualification and signing requirements, etc.

In my experience a low-level guy can develop a really nice product that will
run successfully on a reference platform (e.g., “CE PC”) but cannot be
installed on platforms in commercial distribution. …Unless you are an
insider with the device/network vendor…

I already see this in the Android environment. One can’t install an AT&T
U-Verse application on a Verizon Android device. (Ok, there are actually
other reasons for that…)

Possibly great new world for app/game developers. Going to be challenging
for low-level guys.

Thomas F. Divine

From:
Sent: Thursday, January 06, 2011 10:56 AM
To: “Windows System Software Devs Interest List”
Subject: RE:[ntdev] Re: Windows for ARM? That makes life a bit more
interesting…

> Fabulous news, actually.
>
> Note that the Microsoft announcement for ARM was limited to supporting ARM
> SOC systems. So, it’s not clear if we’ll see ARM desktops or workstations
> (as in NVidia Project Denver) for Win8.
>
> I heartily agree with Mr. Camp’s comments, specifically:
>
>


>
> Recall that the vision for Windows has traditionally been “Windows
> Everywhere” – The whole slate-based computing thing caught Microsoft with
> its pants down. Going back a bit further, the whole move of doing useful
> work on smart phones has caught Microsoft with its pants down. Windows
> CE, even when named Windows embedded, blows. These announcements signal
> that Win8 will aggressively address the slate/phone market, and return to
> Windows Everywhere.
>
> OTOH, I can’t HELP but wonder three things:
> (a) how many ARM variants Windows 8 will support,
> (b) how effectively Win8 will deal with the power challenges inherent in
> devices that use SOCs,
> (c) will it all be too late
>
> There are real iOS slates and phones TODAY. There are real Android phones
> and slates that use what are essentially Android phone operating systems
> today, and there will be a real Android operating system for slates
> tomorrow (Honeycomb). This HAS to make you wonder if Windows 8 will just
> be too little too late… tiles or no tiles (MoSH)… gestures or no
> gestures… ARM or no ARM.
>
> I guess we’ll just have to wait and see.
>
> But for US guys, who work with hardware and devices? Bone-up on your ARM
> architecture, everyone. It’s all good news.
>
> Peter
> OSR
>
> —
> NTDEV is sponsored by OSR
>
> For our schedule of WDF, WDM, debugging and other seminars visit:
> http://www.osr.com/seminars
>
> To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer

13

Before going down the “Windows has only supported x86 for the past 10 years!” road, keep in mind that up through Win7, there was a fully supported IA64 kernel, if you’re going to consider x86 and amd64 mostly identical. (I’ve even got a machine running it in my office.)

  • S (Msft)

-----Original Message-----
From: xxxxx@lists.osr.com [mailto:xxxxx@lists.osr.com] On Behalf Of Tim Roberts
Sent: Thursday, January 06, 2011 9:47 AM
To: Windows System Software Devs Interest List
Subject: Re: [ntdev] Re: Windows for ARM? That makes life a bit more interesting…

I’m with Mark – I’m shocked. It has been 10 years since the last Alpha support was terminated. The kernel is now 3 or 4 times larger than it was at that point, and that’s a lot of new code. Even with the very best of intentions, I have no doubt that a large number of x86 assumptions slipped into the kernel code. The effort required to go through EVERY kernel function and assess whether it is ARM-safe must have been enormous.

xxxxx@osr.com wrote:

Recall that the vision for Windows has traditionally been “Windows Everywhere” – The whole slate-based computing thing caught Microsoft with its pants down. Going back a bit further, the whole move of doing useful work on smart phones has caught Microsoft with its pants down. Windows CE, even when named Windows embedded, blows.

I’m not sure I agree that Microsoft was caught with its pants down.
It’s more like Microsoft simply bet on the wrong horse. Microsoft has been exploring the slate and tablet market for nearly 10 years. I remember Ballmer trying to take notes on a tablet at an MVP Summit a bunch of years ago. However, Microsoft treated the form factor as a limited laptop, and the marketplace was not interested. Apple’s flash of insight was to treat the form factor as a more powerful cell phone, and that, apparently, is exactly what the people wanted.

If “Windows 8 ARM” continues the “limited laptop” model, we will have to see if the world is now ready for that.

OTOH, I can’t HELP but wonder three things:
(a) how many ARM variants Windows 8 will support,
(b) how effectively Win8 will deal with the power challenges inherent
in devices that use SOCs,
(c) will it all be too late

(b) and (c) are very interesting questions indeed. Today’s ARM chips support an unbelievably intricate power management paradigm – FAR more intricate than any x86 chip. Some of these chips have a dozen or more separate subsystems, each of which can has an array of different steps of adjustment in its clock and power levels. The state transition diagrams are mind-boggling, and managing all of that requires a fair amount of complicated heuristics. The sheer volume of code in the Linux kernel that handles ARM power management is enough to make your head spin.

–
Tim Roberts, xxxxx@probo.com
Providenza & Boekelheide, Inc.

NTDEV is sponsored by OSR

For our schedule of WDF, WDM, debugging and other seminars visit:
http://www.osr.com/seminars

To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer

14

Skywing wrote:

Before going down the “Windows has only supported x86 for the past 10 years!” road, keep in mind that up through Win7, there was a fully supported IA64 kernel, if you’re going to consider x86 and amd64 mostly identical. (I’ve even got a machine running it in my office.)

True. I never had to work with an Itanium, but since it has an Intel
pedigree, I have tended to assume that it shares most of its basic
assumptions with the x86.

–
Tim Roberts, xxxxx@probo.com
Providenza & Boekelheide, Inc.

15

No, the IA64 is a very different machine than x86.

  • S

-----Original Message-----
From: xxxxx@lists.osr.com [mailto:xxxxx@lists.osr.com] On Behalf Of Tim Roberts
Sent: Thursday, January 06, 2011 10:11 AM
To: Windows System Software Devs Interest List
Subject: Re: [ntdev] Re: Windows for ARM? That makes life a bit more interesting…

Skywing wrote:

Before going down the “Windows has only supported x86 for the past 10
years!” road, keep in mind that up through Win7, there was a fully
supported IA64 kernel, if you’re going to consider x86 and amd64
mostly identical. (I’ve even got a machine running it in my office.)

True. I never had to work with an Itanium, but since it has an Intel pedigree, I have tended to assume that it shares most of its basic assumptions with the x86.

–
Tim Roberts, xxxxx@probo.com
Providenza & Boekelheide, Inc.

NTDEV is sponsored by OSR

For our schedule of WDF, WDM, debugging and other seminars visit:
http://www.osr.com/seminars

To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer

16

> Note that the Microsoft announcement for ARM was limited to supporting ARM SOC systems.

So, it’s not clear if we’ll see ARM desktops or workstations (as in NVidia Project Denver) for Win8.

So are you speaking about SOC-class or handheld-class devices??? Please note that ARM processors of the same family may come in two major flavors - “embedded processors” that don’t necessarily provide even MMU support, and " application processors" that offer MMU support, multiprocessor support and, in general, are comparable to 32-bit x86 in terms of functionality. The ones of the former type are normally used in SOC systems, while the latter ones used in handheld devices like mobile phones, PDAs, netbooks, etc. Which of them will Win8 support?

Concerning ARM desktops or workstations…well, I am not sure there is any need for them, in the first place. One of the main advantages of ARM over x86 is more efficient energy use (which, of course, comes at the cost of lower performance), which puts ARM at advantage over x86 when it comes to the systems that are powered by battery (i.e. netbooks, phones,etc) because of the longer battery life. However, as long as the target system is meant to have AC supply this advantage is almost of zero importance, right. Therefore, it does not really seem to make sense to use ARM in desktop, don’t you think - you will get performance degradation as a price of gaining something you don’t really care about anyway…

Anton Bassov

17

With all due respect, there’s not many actual proprietary modifications to the OS. It’s mostly all window-dressing and stuff done at the application layer.

But in any case, you can be QUITE certain that Microsoft isn’t going to let people make random modifications to Windows. The ONLY time that I have EVER seen this allowed in the history of Windows was with IBM’s support of the PowerPC. Not only did IBM create their own HAL, but they modified the Windows OS itself. I don’t expect that to be allowed ever again. Today, even custom HALs are no-no.

All that’s public is that ARM SoC architectures will be supported.

http://www.microsoft.com/presspass/press/2011/jan11/01-05SOCsupport.mspx

There has, however, been considerable interest concerning ARM in the SERVER space. Believe it or not. Think things like disk farms.

If you’re a device guy, this all SHOULD lead to some cool opportunities.

Peter
OSR

18

> There has, however, been considerable interest concerning
> ARM in the SERVER space.

But isn’t ARM still 32-bit? And isn’t NT server stack (or at least most
of it) reliant on 64-bit stack? Or is there a “Green IT” subset of the
server stack that works well on 32-bit?

19

I hope that this is true:

“With all due respect, there’s not many actual proprietary modifications to
the OS. It’s mostly all window-dressing and stuff done at the application
layer.”

It wasn’t that ay on CE/Mobile IIRC. Device vendor installed filters that
could not be removed.

Remember that on systems like Android the “owner” doesn’t have root
(administrator) access. The network vendor (e.g., Verizon) does. Big crimp
in one’s style. Of course you can root it, but then you risk your phone
(Tablet) not working.

Then there are the “stores”… Will Toshiba-branded device work only with
Toshiba store? Motorola with Moto?

I think there is more than application window dressing to be considered.

And I’m not saying it’s bad.

Thomas F. Divine

From:
Sent: Thursday, January 06, 2011 1:55 PM
To: “Windows System Software Devs Interest List”
Subject: RE:[ntdev] Re: Windows for ARM? That makes life a bit more
interesting…

>


>
> With all due respect, there’s not many actual proprietary modifications to
> the OS. It’s mostly all window-dressing and stuff done at the application
> layer.
>
> But in any case, you can be QUITE certain that Microsoft isn’t going to
> let people make random modifications to Windows. The ONLY time that I
> have EVER seen this allowed in the history of Windows was with IBM’s
> support of the PowerPC. Not only did IBM create their own HAL, but they
> modified the Windows OS itself. I don’t expect that to be allowed ever
> again. Today, even custom HALs are no-no.
>
>


>
> All that’s public is that ARM SoC architectures will be supported.
>
> http://www.microsoft.com/presspass/press/2011/jan11/01-05SOCsupport.mspx
>
>


>
> There has, however, been considerable interest concerning ARM in the
> SERVER space. Believe it or not. Think things like disk farms.
>
> If you’re a device guy, this all SHOULD lead to some cool opportunities.
>
> Peter
> OSR
>
>
> —
> NTDEV is sponsored by OSR
>
> For our schedule of WDF, WDM, debugging and other seminars visit:
> http://www.osr.com/seminars
>
> To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer

20

>

Microsoft can release a version of Windows on whatever (little-endian)
processor they want, but that won’t mean anything if the 3rd parties
don’t
make applications available.

*cough* app store *cough*

Remember MIPS, PowerPC and Alpha? The Alpha was arguably a superior
processor
in terms of performance, but it just couldn’t succeed without
3rd-party apps.
Windows, MS Office and SQL server alone will not create commercial
success.
Granted, MS didn’t do much to push development of software on the
non-intel
processors back then. Maybe this time they will?

My exposure to Windows on Alpha was limited to one machine, but wasn’t
there some x86 emulation software available for it? Or even bundled with
the machine when you bought Windows?

And if .NET is supported on ARM (and of course it will be), then any
.NET app that uses only managed code will ‘just work’. Quite a lot of
software is .NET these days, and this will most likely only increase
that market share - easier to build one .NET app than an x86, amd64, and
arm version.

James