Daul Boot of Two WinXP. How to get "C:\WINDOWS" in both cases?

[q_q_anonymous]

:
: Timothy Daniels wrote:
: > > So I don't see what Tim was getting at talking about
: > > changing boot.ini causing a change in C. Without
: > > mentioning changing the active partition.
: > > I think you're right talking about changing the active
: > > partition. But that's not quite what Tim said.
: >
: > You're confusing me with someone else. I did not
: > in any way associate boot.ini with the assignment
: > of the partition name "C:".
: >
: > And the "active" flag on a partition has NOTHING
: > to do with whether the OS running off it calls itself
: > "C:" or not. The partition marked "active" is the
: > one where the MBR looks for a boot sector to
: > which to pass control at boot time. The boot
: > sector there expects to find ntldr in its own
: > partition so it can pass control to it. Ntldr then
: > looks for boot.ini at the same level of the file
: > structure so it can be told in what folder in what
: > partition on what hard drive to find the OS so it
: > can load that OS.
:
: but that active partition that the MBR found , is C .

C is not called C by the MBR.

agreed
I meant "will be C" (if we're talking MS Windows).

A partition gets a drive letter from the operating system.

Yeah. A proof is that linux doesn't even use letters to refer to
partitions

C does not have to be an active partition.

can you give a practical example of what you mean . A case where, "C
isn't an active partition"?

<snip>

 

[q_q_anonymous]

This is a frequently-asked question in the MS NGs,
and the answer from the MVPs is essentially "Nothing".
Some people do manage to go into the registry and
change it, but I've never had the need or the courage
to do that. It seems that the name that the OS calls its
own partition

I guess you mean the letter, as I meant. Rather than "volume name".

is set at installation time for that OS, and
Disk Management won't do it for the partition of the
*running* OS.

How are you sure that is the case (that the drive letter is set when
windows is installed)? I can't see how that can be the case at all.

Surely it is set by Windows as soon as Windows is booted on that
partition of the HDD, the partition that is found by the MBR.
Using the terminology in that article.

"Troubleshooting the Startup Process"
http://www.microsoft.com/technet/prodtechnol/winxppro/reskit/c2962167.mspx

I think,
The system partition is always going to be called C.

And I woudl argue that this isn't set when window is installed. It's
set when windows is run.

Suppose you're right that it is set when windows was installed..
And you have 2 drives on your system. Each drive has an active
partition. You boot from primary master.
You swap the drives.. Now that drive that was C, will be D. And what
was D, is now C.

So the drive letters were decided when windows ran, not when windows
was installed.
And whatever drive letters they had when windows was installed, have
now changed
<snip>

 

[Donald L McDaniel]

but that active partition that the MBR found , is C .

An amendment to the boot.ini on that active partition will not change
that. You say you agree .
So What will change it?
It seems to be determined as soon as that active partition is chosen by
the MBR . That partition is C.
But you say "the "active" flag on a partition has NOTHING to do with
whether the OS running off it calls itself "C:" or not."

Fair enough but i'm not clear what does then!


I was around in the old DOS days.. I wasn't referring to any MVPs !!

I notice that xp's "disk management" lets you change drive letter.

It will NOT let you change the drive letter of the active partition.
In other words, it will not let you change "C:"

I haven't tried it, but is that what you have in mind when you almost say
how you can set the registry to make a non active partition call itself
C? I'm not sure what you're getting at when you mention the registry.
Or (windows?) installation.

-useful links which I looked at and got something from. The latter is
thanks to your recommendation

http://ata-atapi.com/hiwmbr.htm
http://www.microsoft.com/technet/prodtechnol/winxppro/reskit/c29621675.mspx

What a bunch of blooming idiots!!!

==

Donald L McDaniel
Please reply to the original thread,
so that it may not become broken.
===================================================

 

[Ron Sommer]

:
: Ron Sommer wrote:
: > : > :
: > : Timothy Daniels wrote:
: > : > > So I don't see what Tim was getting at talking about
: > : > > changing boot.ini causing a change in C. Without
: > : > > mentioning changing the active partition.
: > : > > I think you're right talking about changing the active
: > : > > partition. But that's not quite what Tim said.
: > : >
: > : > You're confusing me with someone else. I did not
: > : > in any way associate boot.ini with the assignment
: > : > of the partition name "C:".
: > : >
: > : > And the "active" flag on a partition has NOTHING
: > : > to do with whether the OS running off it calls itself
: > : > "C:" or not. The partition marked "active" is the
: > : > one where the MBR looks for a boot sector to
: > : > which to pass control at boot time. The boot
: > : > sector there expects to find ntldr in its own
: > : > partition so it can pass control to it. Ntldr then
: > : > looks for boot.ini at the same level of the file
: > : > structure so it can be told in what folder in what
: > : > partition on what hard drive to find the OS so it
: > : > can load that OS.
: > :
: > : but that active partition that the MBR found , is C .
: >
: > C is not called C by the MBR.
:
: agreed
: I meant "will be C" (if we're talking MS Windows).
:
: > A partition gets a drive letter from the operating system.
:
: Yeah. A proof is that linux doesn't even use letters to refer to
: partitions
:
: > C does not have to be an active partition.
:
: can you give a practical example of what you mean . A case where, "C
: isn't an active partition"?
:
: <snip>
:

A clone to another partition (nonactive) of an XP installation (that was
installed to C).

 

[Timothy Daniels]

I guess you mean the letter, as I meant. Rather than "volume name".



How are you sure that is the case (that the drive letter is set when
windows is installed)? I can't see how that can be the case at all.

Surely it is set by Windows as soon as Windows is booted on that
partition of the HDD, the partition that is found by the MBR.
Using the terminology in that article.

"Troubleshooting the Startup Process"
http://www.microsoft.com/technet/prodtechnol/winxppro/reskit/c2962167.mspx

I think,
The system partition is always going to be called C.

(e-mail address removed), you're stuck in some
mental reference that won't let go. I don't know how or
why, but you should loosen its grip on you. For the last
time, whether the system partition (the partition that holds
the boot files) does NOT have to be called "C:" when
the OS that it cause to boot gets running. That is because
there IS no "C:" at that point in the startup, and boot.ini
has no links to the name "C:". Indeed, as I have already
pointed out, boot.ini's entries can point to any folder in
any partition on any hard drive in the system from which
to load the OS.

The name "C:" is held in the OS's registry, and if the
entry "C:" was made in that registry when the OS was
installed, that is what the OS will call it. For instance,
a clone of an OS that calls its partition "C:" will also
call its partition "C:". If other pre-existing Windows
OSes are hidden from the installer's view when the installer
is installing another Windows OS, that new OS will be
told to call its own partition "C:" (unless one, perhaps,
tells it to call it something else), and forever after, unless
its registry is diddled with, it will call its own partition "C:".
If the installer sees other OSes when it is doing an install,
it will assign an unused letter as the name of the new
OS's partition. If you believe otherwise, please test your
belief with some current experiments.

*TimDaniels*

 

[Timothy Daniels]

can you give a practical example of what you mean .
A case where, "C isn't an active partition"?

You still don't understand what the "active" flag
means. All it does is tell the boot sector that its
partition has the ntldr boot manager and the
boot.ini menu file and the ntdetect.com hardware
environment detection routine. That's ALL it means.

It has NOTHING to do with where the OS resides.
It has NOTHING to do with what the loaded and
running OS calls its own partition.

*TimDaniels*

 

[Andy]

can you give a practical example of what you mean . A case where, "C
isn't an active partition"?

Start with a disk that has no partition.
Boot from the XP CD and run setup.
Create the first partition (C), which will be primary, active.
Create a second partition (D), which will be a logical volume in the
extended partition.
Delete the first partition.
Quit setup.
Boot from the XP CD again and run setup. The second partition will now
be C, and the CD drive will be D.
Create the first partition again, which will primary, active. This
time the partition will be E.
Install Windows on E.

The key point is that if an active primary partition exists when
Windows setup is started, then it will be C. On the other hand, if
other hard disk partitions exist when Windows setup is started, then
the active primary partition, when it is created, won't be C.

 

[q_q_anonymous]

You still don't understand what the "active" flag
means. All it does is tell the boot sector that its
partition has the ntldr boot manager and the
boot.ini menu file and the ntdetect.com hardware
environment detection routine. That's ALL it means.

good definition of those files

It has NOTHING to do with where the OS resides.

I do see what you're saying.. That An OS that will call its partition
C, does not have to be on an active partition.

And I was saying, that the "system partition", the active partition
that the MBR goes to, the partition that Does not necessarily have to
have an OS. It could just have a boot manager with boot menu and
hardware detection. I was saying that that partition is always C.

You suggested a way for it to not be C. I will try it.

I will try your and Ron's experiment regarding Cloning a partition.

It has NOTHING to do with what the loaded and
running OS calls its own partition.

well, "nothing to do with" is a weak expression.

I haven't yet tried your experiment, but I was saying that "if" the
system partition calls itself C, then the boot partition, when it isn't
the system partition, will be DEF e.t.c. or EF.

You're arguing that if I do this cloning thing I'll see otherwise,,
I'll have to try it. I will.

I have no problem seeing what you are saying. that is precisely why I
asked you for a demonstration of your point. And you've provided one
with the cloning idea. I have to test it.

 

[q_q_anonymous]

:
: Ron Sommer wrote:
: > : > :
: > : Timothy Daniels wrote:
: > : > > So I don't see what Tim was getting at talking about
: > : > > changing boot.ini causing a change in C. Without
: > : > > mentioning changing the active partition.
: > : > > I think you're right talking about changing the active
: > : > > partition. But that's not quite what Tim said.
: > : >
: > : > You're confusing me with someone else. I did not
: > : > in any way associate boot.ini with the assignment
: > : > of the partition name "C:".
: > : >
: > : > And the "active" flag on a partition has NOTHING
: > : > to do with whether the OS running off it calls itself
: > : > "C:" or not. The partition marked "active" is the
: > : > one where the MBR looks for a boot sector to
: > : > which to pass control at boot time. The boot
: > : > sector there expects to find ntldr in its own
: > : > partition so it can pass control to it. Ntldr then
: > : > looks for boot.ini at the same level of the file
: > : > structure so it can be told in what folder in what
: > : > partition on what hard drive to find the OS so it
: > : > can load that OS.
: > :
: > : but that active partition that the MBR found , is C .
: >
: > C is not called C by the MBR.
:
: agreed
: I meant "will be C" (if we're talking MS Windows).
:
: > A partition gets a drive letter from the operating system.
:
: Yeah. A proof is that linux doesn't even use letters to refer to
: partitions
:
: > C does not have to be an active partition.
:
: can you give a practical example of what you mean . A case where, "C
: isn't an active partition"?
:
: <snip>
:

A clone to another partition (nonactive) of an XP installation (that was
installed to C).
--

So, thanks to the cloning, I will have 2 partitions.. Both will have
had win xp installed on it with them taking the C drive letter.

I'll just define system partition so we're clear which is the system
partition, which are the boot partition(s). There is an overlap - by
definition.
The system partition (by definition, it being the one the MBR goes to,
will be the active one). It will have win xp on it ('cos it was the
original partition you speak of). The boot loader will run off it, NT
always has those 3 files. And the partition that the boot manager
chooses to boot off of, is the boot partition.

For your experiment, i'll try first letting the system partition be
the boot partition, as it was. Thus, leaving its boot.ini alone. And
then i'll try making "the clone" the boot partition by amending the
system partition's boot.ini accordingly.

thank you

 

[q_q_anonymous]

Start with a disk that has no partition.
Boot from the XP CD and run setup.
Create the first partition (C), which will be primary, active.
Create a second partition (D), which will be a logical volume in the
extended partition.
Delete the first partition.
Quit setup.
Boot from the XP CD again and run setup. The second partition will now
be C, and the CD drive will be D.
Create the first partition again, which will primary, active. This
time the partition will be E.
Install Windows on E.

The key point is that if an active primary partition exists when
Windows setup is started, then it will be C. On the other hand, if
other hard disk partitions exist when Windows setup is started, then
the active primary partition, when it is created, won't be C.

interesting, innovative, very nice !

thanks

 

[Timothy Daniels]

So, thanks to the cloning, I will have 2 partitions.. Both will have
had win xp installed on it with them taking the C drive letter.
Yes.


I'll just define system partition so we're clear which is the system
partition, which are the boot partition(s). There is an overlap - by
definition.

There is an overlap in the sense that they can be the
same partition, but there is nothing that requires them to be
the same partition.

The system partition (by definition, it being the one the MBR
goes to, will be the active one).

Yes. The MBR passes control to the partition (on the
hard drive at the head of the BIOS's hard drive boot order)
that is marked "active". First it looks for the hard drive
designated by the HD boot order, then it looks for the
Primary partition that is marked "active".

It will have win xp on it ('cos it was the original partition you speak of).

No. It will (should) have ntldr/boot.ini/ntdetect.com on it.
Where the WinXP is can be anywhere.

The boot loader will run off it, NT always has those 3 files.

Yes. NT/2K/XP all use ntldr to load the OS.

And the partition that the boot manager chooses to boot off of,
is the boot partition.

No. The boot manager (i.e. ntldr) resides on the "active"
partition, and that partition is called (non-intuitively) by Microsoft
the "system partition".

Also non-intuitively, Microsoft calls the partition that contains
the OS the "boot partition" - for historical reasons.

So "system partition" and "boot partition" are the opposite
of what you'd expect them to be.

For your experiment, i'll try first letting the system partition be
the boot partition, as it was. Thus, leaving its boot.ini alone. And
then i'll try making "the clone" the boot partition by amending the
system partition's boot.ini accordingly.

If you're making a clone, spare yourself any grief by making
the clone on another hard drive - not on the same HD as the
original OS. That is because the NT/2K/XP family of Windows
OSes appear to do some randomly-distributed cross-referencing
of their files if they see their "parent" OS when they start up for
the 1st time after being created by cloning. For some files, when
you think you're seeing and editing it, you're really seeing and editing
the same-named file in the "parent" - which eventually leads to grief.
The "parent" OS will be OK, the clone will be OK after it is started
for the 1st time in isolation from its "parent". But remember to
disconnect the "parent" OS's hard drive before firing up the clone
for its first run. If the clone is on the same HD as its "parent" OS,
you will have to use a 3rd-party utility (such as Partition Magic) to
hide the "parent" OS's partition before booting the clone for its first
run.

*TimDaniels*

 

[Timothy Daniels]

I do see what you're saying.. That An OS that will call its partition
C, does not have to be on an active partition.
Right.


And I was saying, that the "system partition", the active partition
that the MBR goes to, the partition that Does not necessarily have to
have an OS. It could just have a boot manager with boot menu and
hardware detection. I was saying that that partition is always C.

Yes, and now you realize that it doesn't have to be C:.

I have to test it.

Good. I like your approach. Experimentation is integral
to education.

*TimDaniels*

 

[Donald L McDaniel]

(e-mail address removed), you're stuck in some
mental reference that won't let go. I don't know how or
why, but you should loosen its grip on you. For the last
time, whether the system partition (the partition that holds
the boot files) does NOT have to be called "C:" when
the OS that it cause to boot gets running. That is because
there IS no "C:" at that point in the startup, and boot.ini
has no links to the name "C:". Indeed, as I have already
pointed out, boot.ini's entries can point to any folder in
any partition on any hard drive in the system from which
to load the OS.

The name "C:" is held in the OS's registry, and if the
entry "C:" was made in that registry when the OS was
installed, that is what the OS will call it. For instance,
a clone of an OS that calls its partition "C:" will also
call its partition "C:". If other pre-existing Windows
OSes are hidden from the installer's view when the installer
is installing another Windows OS, that new OS will be
told to call its own partition "C:" (unless one, perhaps,
tells it to call it something else), and forever after, unless
its registry is diddled with, it will call its own partition "C:".
If the installer sees other OSes when it is doing an install,
it will assign an unused letter as the name of the new
OS's partition. If you believe otherwise, please test your
belief with some current experiments.

*TimDaniels*

You guys are just arguing over technicalities.

For all intents and purposes, there will never be a time when two C:
partitions are active at the same time, since Windows makes no
distinction between a physical disk and a partition (in giving them
drive letters). Windows will not give two disks or partitions the
same drive (or logical drive) letter at the same time.

Therefore, while there may theoretically be two "C:" partitions (if
they are both hidden from each other -- and I'm not entirely convinced
that even that is even theoretically possible), for all practical
purposes, there can be only ONE "C:" shown in "My Computer" (or
"Computer", in Vista).

And the Windows Boot manager also will not show two "C:"
partitions/drives if one is dual-booting.

Please, give it a rest?

==

Donald L McDaniel
Please reply to the original thread,
so that it may not become broken.
===================================================