Why doesn't Windows XP get patched to use the NT 5.2 kernel

[Link19]

The NT 5.2 kernel is the latest kernel version from Microsoft? Why doesn't Windows XP 32-bit get patched to use the NT 5.2 kernel?

Windows XP is the client version of Microsoft's most curent OS and Widnows Server 2003 is the server edition of Microsoft's most current OS? SO why don't they both use the most up to date NT 5.2 kernel?

Windows 2000 Professional is the client edition of Windows 2000 and Windows 2000 Server is the server edition. They both use the NT 5.0 kernel. The difference is the licensing and the server version has server features not present in WIndows 2000 Pro. But they both have the same kernel version?

Why don't Windows XP and Windows Server 2003 both use the most up to date kernel unlike Windows 2000 Professional and Widnows 2000 Server did?

I have always wondered about this, but have yet to find a a valid explanation? Are there any advanatges of the NT 5.2 kernel over the NT 5.1 kernel? Or are all improvements in the NT 5.2 kernel for server functions only?

I have heard that Windows Vista is based on Windows Server 2003 SP1. If the NT 5.2 kernel didn't provide any improvements to be used for a workstation OS, then why would Vista be based on Windows Server 2003 SP1 as WIndows Server 2003 SP1 is a server OS, and Vista is going to be a workstation/desktop OS? There must be improvements in the NT 5.2 kernel if it is what the developement of Longhorn/Vista is based on?

 

[bsobel]

Windows XP is the client version of Microsoft's most curent OS and Widnows Server 2003 is the server edition of Microsoft's most current OS? SO why don't they both use the most up to date NT 5.2 kernel?

Because XP was released before 2003 server. At the time of release they were on the 5.1 kernel. When they released 2003 server (and then xp 64) they were on the 5.2 kernel. There is no need to always have the newest thing for the sake of it being newer. Besides, the primary difference is 64bit support.

Why don't Windows XP and Windows Server 2003 both use the most up to date kernel unlike Windows 2000 Professional and Widnows 2000 Server did?

Your statement is false, only some configs are on different kernels (64bit and 2003 server are on 5.2).

I have heard that Windows Vista is based on Windows Server 2003 SP1. If the NT 5.2 kernel didn't provide any improvements to be used for a workstation OS, then why would Vista be based on Windows Server 2003 SP1 as WIndows Server 2003 SP1 is a server OS, and Vista is going to be a workstation/desktop OS? There must be improvements in the NT 5.2 kernel if it is what the developement of Longhorn/Vista is based on?

There is really only one dev branch for Windows, sure it forks for specific releases (such as xp, xp64, server, etc) but the main branch is the main branch. When you say 'based on 2003 sp1' that was the last release which respesent the newest code, hence Vista is based on that code. But it's incorrect to say Vista is based on 2003 server.

Bill

 

[Link19]

Your statement is false, only some configs are on different kernels (64bit and 2003 server are on 5.2).

I meant the 32-bit versions of Windows XP and Windows Server 2003 which are on different kernels. I know the 64-bit edition of Windows Server 2003 and Windows XP both use the 64-bit compiled NT 5.2 kernel.

It is true that Windows 2000 Professional and Windows 2000 Server versions all use the same kernel version right?

 

[bersl2]

$$$

(probably)

 

[bsobel]

Originally posted by: bersl2
$$$

(probably)

Way to add nothing to the thread.

 

[stash]

It is true that Windows 2000 Professional and Windows 2000 Server versions all use the same kernel version right?

Yes, so? They were all released simultaneously. Server 2003 was released after XP.

I'm not sure what point you're trying to make here.

 

[bsobel]

I'm not sure what point you're trying to make here.

Consider who the OP is. No point to be made...

 

[xtknight]

Erm, all he's asking is why Windows XP can not use kernel version 5.2. Is a kernel upgrade not possible (financially? technically?)? If not, why not? Would it not help anyway? Would XP's internals become incompatible with a new kernel? Kernel upgrades are routine for patches in Linux systems. Is it because Linux's kernel encompasses tons of drivers as well that need updating? Just wondering why the situation is different with Windows. Also what would be the difference between kernel 5.1 and kernel 5.2?

 

[Link19]

Originally posted by: xtknight
Erm, all he's asking is why Windows XP can not use kernel version 5.2. Is a kernel upgrade not possible (financially? technically?)? If not, why not? Would it not help anyway? Would XP's internals become incompatible with a new kernel? Kernel upgrades are routine for patches in Linux systems. Is it because Linux's kernel encompasses tons of drivers as well that need updating? Just wondering why the situation is different with Windows. Also what would be the difference between kernel 5.1 and kernel 5.2?

That is all I am asking. I would be interested in knowing the same questions you're asking.

I do know that the Windows XP kernel does get updated, but only very minor revision changes that are not displayed as part of the version number when you go to help and about in WIndows Explorer. Only if you look at the actual core OS kernel files will you be able to see the full version number. Search for the file ntoskrnl.exe and right click on it and select Properties. Go to the version tab and it will display the full file version. For instance, Windows XP SP1 was NT kernel version 5.1.2600.1176. Windows XP SP2 uses NT kernel version 5.1.2600.2180. The only part of the kernel verison number that MS sometimes updates is the last extension of the file version. You never see the 5.1.2600 part of the version number ever change. I wonder why? Why not just apply the most up to date 5.2 kernel to XP. NT kernel 5.2 I'm sure is probably more up to date than 5.1.2600 with the last part of the reviison number changing? But I could be wrong? I just don't know?

 

[bsobel]

The only part of the kernel verison number that MS sometimes updates is the last extension of the file version. You never see the 5.1.2600 part of the version number ever change. I wonder why?

App compat reasons, some folks still check version numbers so these only get bumped on major releases.

 

[Sunner]

If nothing else, bumping the version number of major system components in an industrial grade OS that you're promised to support for 5+ years is a bad thing to do.
Same deal with Redhat EL as well, 2.6.16 may be out, but Redhat keeps 2.6.9 and just backports fixes.

 

[Link19]

Originally posted by: Sunner
If nothing else, bumping the version number of major system components in an industrial grade OS that you're promised to support for 5+ years is a bad thing to do.
Same deal with Redhat EL as well, 2.6.16 may be out, but Redhat keeps 2.6.9 and just backports fixes.

Why is that?

 

[Madwand1]

$$$ is actually a good answer. If you think that MS is not interested in promoting new kernel advances, and accordingly interested in providing them for free in old OS versions, you're naive.

 

[Link19]

Originally posted by: Madwand1
$$$ is actually a good answer. If you think that MS is not interested in promoting new kernel advances, and accordingly interested in providing them for free in old OS versions, you're naive.

Then why doesn't Microsoft release a workstation version of Windows 2003? Would there be any workstation benefit to the NT 5.2 kernel?

 

[MrChad]

Originally posted by: Link19

Originally posted by: Madwand1
$$$ is actually a good answer. If you think that MS is not interested in promoting new kernel advances, and accordingly interested in providing them for free in old OS versions, you're naive.

Then why doesn't Microsoft release a workstation version of Windows 2003? Would there be any workstation benefit to the NT 5.2 kernel?

I think you're overstating the differences between the 5.2 kernel in 2003 and the 5.1 kernel in XP. There's quite a bit of overlap, as many of the same changes made in XP SP2 were also made in 2003 SP1.

Besides, Microsoft already has a new workstation OS planned ... it's called Vista.

 

[kobymu]

Originally posted by: bsobel
There is really only one dev branch for Windows, sure it forks for specific releases (such as xp, xp64, server, etc) but the main branch is the main branch. When you say 'based on 2003 sp1' that was the last release which respesent the newest code, hence Vista is based on that code. But it's incorrect to say Vista is based on 2003 server.

Bill

Not exactly but close enough
http://www.levenez.com/windows/windows.pdf

 

[Link19]

Originally posted by: MrChad

Originally posted by: Link19

Originally posted by: Madwand1
$$$ is actually a good answer. If you think that MS is not interested in promoting new kernel advances, and accordingly interested in providing them for free in old OS versions, you're naive.

Then why doesn't Microsoft release a workstation version of Windows 2003? Would there be any workstation benefit to the NT 5.2 kernel?

I think you're overstating the differences between the 5.2 kernel in 2003 and the 5.1 kernel in XP. There's quite a bit of overlap, as many of the same changes made in XP SP2 were also made in 2003 SP1.

Besides, Microsoft already has a new workstation OS planned ... it's called Vista.

So could just as many improvements be made to the NT 5.1 kernel that were already included in the NT 5.2 kernel, it is just that Microsoft decides to not change the more major part of the version number unless a new paid for product is released?

Another words, could the Windows NT 5.1.2600.1176 kernel updated to NT kernel 5.1.2600.2180 contain just as many improvements as going from the NT 5.1.2600.1176 kernel to the NT 5.2.3790 kernel?

 

[dawks]

Originally posted by: MrChad

Originally posted by: Link19

Originally posted by: Madwand1
$$$ is actually a good answer. If you think that MS is not interested in promoting new kernel advances, and accordingly interested in providing them for free in old OS versions, you're naive.

Then why doesn't Microsoft release a workstation version of Windows 2003? Would there be any workstation benefit to the NT 5.2 kernel?

I think you're overstating the differences between the 5.2 kernel in 2003 and the 5.1 kernel in XP. There's quite a bit of overlap, as many of the same changes made in XP SP2 were also made in 2003 SP1.

Besides, Microsoft already has a new workstation OS planned ... it's called Vista.

Yup, the differences are pretty small. Point releases such as 5 .1 to 5 .2 dont offer many changes. As said before much of it is just tweaks. Windows 2000 = 5.0 Windows XP 5.1 Windows Server 2003 5.2. The changes from 2000 to XP brought a few things like increasing driver size limits, and tweaking boot-times and the application pre-fetch system. The same is true with Server 2003. While I dont know the precise details I'm assuming there were a few changes to optimize the kernel for large server loads. Not too many things that would benefit workstation users.

 

[Sunner]

Originally posted by: Link19

Originally posted by: Sunner
If nothing else, bumping the version number of major system components in an industrial grade OS that you're promised to support for 5+ years is a bad thing to do.
Same deal with Redhat EL as well, 2.6.16 may be out, but Redhat keeps 2.6.9 and just backports fixes.

Why is that?

Compatibility reasons for one.
Say you have app x running on OS y, if your OS vendor decided to start upgrading major system components(and the kernel certainly is), you'd have a lot of compatibility testing to do.
Sure, you'll do that with regular patches too, but at least you'll know they won't change the behaviure of the system with security patches and such, while a new version might have new features, old ones changed or even removed, etc, basically you're getting a new system, and you don't want that in a live production environment.

 

[Canterwood]

Originally posted by: Sunner
Compatibility reasons for one.
Say you have app x running on OS y, if your OS vendor decided to start upgrading major system components(and the kernel certainly is), you'd have a lot of compatibility testing to do.
Sure, you'll do that with regular patches too, but at least you'll know they won't change the behaviure of the system with security patches and such, while a new version might have new features, old ones changed or even removed, etc, basically you're getting a new system, and you don't want that in a live production environment.

But isn't that what we basically had with XP SP2?
A fair ammount of apps had to be re-coded to work properly with it.

 

[Link19]

Originally posted by: Canterwood

Originally posted by: Sunner
Compatibility reasons for one.
Say you have app x running on OS y, if your OS vendor decided to start upgrading major system components(and the kernel certainly is), you'd have a lot of compatibility testing to do.
Sure, you'll do that with regular patches too, but at least you'll know they won't change the behaviure of the system with security patches and such, while a new version might have new features, old ones changed or even removed, etc, basically you're getting a new system, and you don't want that in a live production environment.

But isn't that what we basically had with XP SP2?
A fair ammount of apps had to be re-coded to work properly with it.

I was thinking the same thing.

 

[dawks]

Originally posted by: Link19

Originally posted by: Canterwood

Originally posted by: Sunner
Compatibility reasons for one.
Say you have app x running on OS y, if your OS vendor decided to start upgrading major system components(and the kernel certainly is), you'd have a lot of compatibility testing to do.
Sure, you'll do that with regular patches too, but at least you'll know they won't change the behaviure of the system with security patches and such, while a new version might have new features, old ones changed or even removed, etc, basically you're getting a new system, and you don't want that in a live production environment.

But isn't that what we basically had with XP SP2?
A fair ammount of apps had to be re-coded to work properly with it.

I was thinking the same thing.

Alot of it was just becuase many programmers were lazy and were doing things they shouldn't have, security wise. When XP tightend down, many of these apps 'broke'.

Some problems were also due to the firewall being enabled by default.

One of the big changes we're going to see with Vista is users arent going to be running as administrator now. The vast majority of applications on the market today will not work properly unless the user is running as admin. And as a network admin, thats huge PITA.

 

[Sunner]

Originally posted by: dawks

Originally posted by: Link19

Originally posted by: Canterwood

Originally posted by: Sunner
Compatibility reasons for one.
Say you have app x running on OS y, if your OS vendor decided to start upgrading major system components(and the kernel certainly is), you'd have a lot of compatibility testing to do.
Sure, you'll do that with regular patches too, but at least you'll know they won't change the behaviure of the system with security patches and such, while a new version might have new features, old ones changed or even removed, etc, basically you're getting a new system, and you don't want that in a live production environment.

But isn't that what we basically had with XP SP2?
A fair ammount of apps had to be re-coded to work properly with it.

I was thinking the same thing.

Alot of it was just becuase many programmers were lazy and were doing things they shouldn't have, security wise. When XP tightend down, many of these apps 'broke'.

Some problems were also due to the firewall being enabled by default.

One of the big changes we're going to see with Vista is users arent going to be running as administrator now. The vast majority of applications on the market today will not work properly unless the user is running as admin. And as a network admin, thats huge PITA.

Well, Sp2 did cause a bit of a ruckus in the corporate world, but MS did everything in their power to keep the problems away.
Nevertheless, it was badly needed.