Reqst: How get Fedora to boot - Redhat

This is a discussion on Reqst: How get Fedora to boot - Redhat ; "Hamilcar Barca" wrote in message news:20050107180012.930$t1@news.newsreader.com... > In article (Fri, 07 Jan 2005 07:47:58 > -0500), Nico Kadel-Garcia wrote: > >> "Hamilcar Barca" wrote in message >> news:20050107013429.844$wp@news.newsreader.com... >>> In article (Fri, 07 Jan 2005 14:27:17 >>> +0800), >>> Matthew ...

+ Reply to Thread
Page 3 of 3 FirstFirst 1 2 3
Results 41 to 55 of 55

Thread: Reqst: How get Fedora to boot

  1. Re: Question about Linux


    "Hamilcar Barca" wrote in message
    news:20050107180012.930$t1@news.newsreader.com...
    > In article (Fri, 07 Jan 2005 07:47:58
    > -0500), Nico Kadel-Garcia wrote:
    >
    >> "Hamilcar Barca" wrote in message
    >> news:20050107013429.844$wp@news.newsreader.com...
    >>> In article <41de29f7@quokka.wn.com.au> (Fri, 07 Jan 2005 14:27:17
    >>> +0800),
    >>> Matthew Gunn wrote:
    >>>
    >>>> Does anyone else know the answers to sort these two out.
    >>>
    >>> With the possible exception of Xenix, Microsoft operating systems have
    >>> all
    >>> required a primary partition on the first hard drive.

    >>
    >> And this is blatantly untrue. Win2K and WinXP and Win2003 all work just
    >> fine
    >> on secondary drives, I'm not sure about NT 4.0.

    >
    > This is true and always has been true, despite your protestations to the
    > contrary. If this is not clear, why not read Microsoft's installation
    > documentation? Windows 2000 requires, as did Windows NT4, a primary
    > partition on the first hard drive.


    Because I've done it. Twice. Without significant difficulty.



  2. Re: Question about Linux

    In article ,
    "Nico Kadel-Garcia" writes:
    >
    >> Windows 2000 requires, as did Windows NT4, a primary
    >> partition on the first hard drive.

    >
    > Because I've done it. Twice. Without significant difficulty.


    Precisely what did you do, and how? I've seen such claims before, but
    whenever details have emerged, it's always turned out to be a case of the
    person doing one thing and believing something else was done (such as
    installing just the WINDOWS or WINNT directory to a logical partition,
    leaving the critical boot files on a primary partition on the first disk;
    or using BIOS trickery to swap the identities of two disks, thus enabling
    installation to a primary partition on the second disk only because the
    BIOS remapped it to look like the first disk). Thus, I treat such claims
    with skepticism. If you can back this up with a detailed description or a
    link to such a description, I'd love to see it.

    --
    Rod Smith, rodsmith@rodsbooks.com
    http://www.rodsbooks.com
    Author of books on Linux, FreeBSD, and networking

  3. Re: Question about Linux

    In article (Fri, 07 Jan 2005 23:14:33
    -0500), Nico Kadel-Garcia wrote:

    > "Hamilcar Barca" wrote in message
    > news:20050107180012.930$t1@news.newsreader.com...
    >>
    >> This is true and always has been true, despite your protestations to the
    >> contrary. If this is not clear, why not read Microsoft's installation
    >> documentation? Windows 2000 requires, as did Windows NT4, a primary
    >> partition on the first hard drive.

    >
    > Because I've done it. Twice. Without significant difficulty.


    Since Windows 2000 disallows this, what insignificant difficulties did you
    encounter in fooling it?

    --
    "We're big believers in interoperability."
    -- Bill Gates. Chairman, Microsoft. USA Today interview. 13 Oct 2004.


  4. Re: Question about Linux


    "Rod Smith" wrote in message
    news:51k5b2-s0c.ln1@speaker.rodsbooks.com...
    > In article ,
    > "Nico Kadel-Garcia" writes:
    >>
    >>> Windows 2000 requires, as did Windows NT4, a primary
    >>> partition on the first hard drive.

    >>
    >> Because I've done it. Twice. Without significant difficulty.

    >
    > Precisely what did you do, and how? I've seen such claims before, but
    > whenever details have emerged, it's always turned out to be a case of the
    > person doing one thing and believing something else was done (such as
    > installing just the WINDOWS or WINNT directory to a logical partition,
    > leaving the critical boot files on a primary partition on the first disk;
    > or using BIOS trickery to swap the identities of two disks, thus enabling
    > installation to a primary partition on the second disk only because the
    > BIOS remapped it to look like the first disk). Thus, I treat such claims
    > with skepticism. If you can back this up with a detailed description or a
    > link to such a description, I'd love to see it.


    Two hard drives. I had Linux on the first hard drive, added a second hard
    drive in a hot-swap bay for testing, and installed Windows 2000 Professional
    on the second hard drive. As I remember, it over-wrote the MBR, but I was
    able to fix that with a Linux rescue CD. The hard drive showed up as drive
    E:, and did its installation on drive E, although I don't remember why.

    I did a similar operation later with a Windows XP installation on an
    external USB hard drive, where the internal hard drive had Gentoo Linux on
    it. If you have an external such drive in hand and a Windows installation
    CD, it's a trivial test, but I'm a bit short on cycles to re-run that right
    now. (New job, more work.)

    And please note, the BIOS trickery you refer to is not necessarily in the
    BIOS, it can be done by the boot loader itself. That may in fact be what
    Windows does, that part I can't swear to. But I've certainly used the trick
    in the Linux world to teach grub or LILO to swap the drives for making boot
    image mirrors of one drive onto a secondary drive, then loading the MBR to
    booth drives so if one drive failed the other would still boot.



  5. Re: Question about Linux


    "Hamilcar Barca" wrote in message
    news:20050108132620.258$Hn@news.newsreader.com...
    > In article (Fri, 07 Jan 2005 23:14:33
    > -0500), Nico Kadel-Garcia wrote:
    >
    >> "Hamilcar Barca" wrote in message
    >> news:20050107180012.930$t1@news.newsreader.com...
    >>>
    >>> This is true and always has been true, despite your protestations to the
    >>> contrary. If this is not clear, why not read Microsoft's installation
    >>> documentation? Windows 2000 requires, as did Windows NT4, a primary
    >>> partition on the first hard drive.

    >>
    >> Because I've done it. Twice. Without significant difficulty.

    >
    > Since Windows 2000 disallows this, what insignificant difficulties did you
    > encounter in fooling it?


    As I remember, only having to reset the MBR with my Linux rescue CD to
    reboot the Linux I had on the first drive. Setting MBR's is a bit of an
    adventure.



  6. Re: Question about Linux

    In article ,
    "Nico Kadel-Garcia" writes:
    >
    > "Rod Smith" wrote in message
    > news:51k5b2-s0c.ln1@speaker.rodsbooks.com...
    >> In article ,
    >> "Nico Kadel-Garcia" writes:
    >>>
    >>>> Windows 2000 requires, as did Windows NT4, a primary
    >>>> partition on the first hard drive.
    >>>
    >>> Because I've done it. Twice. Without significant difficulty.

    >>
    >> Precisely what did you do, and how?

    ....
    > Two hard drives. I had Linux on the first hard drive, added a second hard
    > drive in a hot-swap bay for testing, and installed Windows 2000 Professional
    > on the second hard drive. As I remember, it over-wrote the MBR, but I was
    > able to fix that with a Linux rescue CD. The hard drive showed up as drive
    > E:, and did its installation on drive E, although I don't remember why.


    I've just tried reproducing this and failed. I put a couple of spare
    drives in a 1.2GHz AMD Duron system (first disconnecting the regular
    drives). The primary master drive was a 6GB unit that had a small primary
    Linux swap partition and a larger primary FreeBSD partition. (I also
    tried removing the FreeBSD partition and replaced it with a logical FAT
    partition; that change had no effect.) The second drive was a 10GB unit
    configured as the primary slave. I put a single primary FAT partition on
    the drive to be used as the installation partition. I've tried installing
    both Windows 2000 Professional and Windows XP Home on this system. Both
    failed. After selecting the primary FAT partition (which the installer
    identified as C: in all cases) as the destination and telling the
    installer to reformat it, it put up a message to the effect that it could
    not proceed without a "Windows 2000-compatible partition" on the first
    disk.

    The fact that when you installed it on a second drive, Windows identified
    the installation partition as E: strongly suggests that it found
    partitions it identified as C: and D:. My hunch is that C: was a primary
    FAT partition on the first disk. If you're absolutely positive that you
    had no such partition, then perhaps something else was going on, like
    misidentification of non-FAT partitions as FAT (which would likely cause
    serious problems, sooner or later). The fact that I can't reproduce your
    installation on my test system, though, suggests that either you're
    forgetting an important detail or whatever caused it to work for you
    isn't a universal factor, making such installation unreliable at best.

    > And please note, the BIOS trickery you refer to is not necessarily in the
    > BIOS, it can be done by the boot loader itself.


    Yes, I'm aware of that. My reference to BIOS trickery was just an example
    of ways in which the system can be configured in a way that's deceptive,
    not the only way this can be done.

    --
    Rod Smith, rodsmith@rodsbooks.com
    http://www.rodsbooks.com
    Author of books on Linux, FreeBSD, and networking

  7. Re: Question about Linux


    "Rod Smith" wrote in message
    news:n4e9b2-l4i.ln1@speaker.rodsbooks.com...
    > In article ,
    > "Nico Kadel-Garcia" writes:
    >>
    >> "Rod Smith" wrote in message
    >> news:51k5b2-s0c.ln1@speaker.rodsbooks.com...
    >>> In article ,
    >>> "Nico Kadel-Garcia" writes:
    >>>>
    >>>>> Windows 2000 requires, as did Windows NT4, a primary
    >>>>> partition on the first hard drive.
    >>>>
    >>>> Because I've done it. Twice. Without significant difficulty.
    >>>
    >>> Precisely what did you do, and how?

    > ...
    >> Two hard drives. I had Linux on the first hard drive, added a second hard
    >> drive in a hot-swap bay for testing, and installed Windows 2000
    >> Professional
    >> on the second hard drive. As I remember, it over-wrote the MBR, but I was
    >> able to fix that with a Linux rescue CD. The hard drive showed up as
    >> drive
    >> E:, and did its installation on drive E, although I don't remember why.

    >
    > I've just tried reproducing this and failed. I put a couple of spare
    > drives in a 1.2GHz AMD Duron system (first disconnecting the regular
    > drives). The primary master drive was a 6GB unit that had a small primary
    > Linux swap partition and a larger primary FreeBSD partition. (I also
    > tried removing the FreeBSD partition and replaced it with a logical FAT
    > partition; that change had no effect.) The second drive was a 10GB unit
    > configured as the primary slave. I put a single primary FAT partition on
    > the drive to be used as the installation partition. I've tried installing
    > both Windows 2000 Professional and Windows XP Home on this system. Both
    > failed. After selecting the primary FAT partition (which the installer
    > identified as C: in all cases) as the destination and telling the
    > installer to reformat it, it put up a message to the effect that it could
    > not proceed without a "Windows 2000-compatible partition" on the first
    > disk.


    OK, stop right there. You didn't try my recipe, you put garlic and
    strawberries in it, and it came out fairly weird.

    > The fact that when you installed it on a second drive, Windows identified
    > the installation partition as E: strongly suggests that it found
    > partitions it identified as C: and D:. My hunch is that C: was a primary
    > FAT partition on the first disk. If you're absolutely positive that you
    > had no such partition, then perhaps something else was going on, like
    > misidentification of non-FAT partitions as FAT (which would likely cause
    > serious problems, sooner or later). The fact that I can't reproduce your
    > installation on my test system, though, suggests that either you're
    > forgetting an important detail or whatever caused it to work for you
    > isn't a universal factor, making such installation unreliable at best.


    Conceivable, but you have a complex enough little setup there that I
    wouldn't trust it to be a good test.

    >> And please note, the BIOS trickery you refer to is not necessarily in the
    >> BIOS, it can be done by the boot loader itself.

    >
    > Yes, I'm aware of that. My reference to BIOS trickery was just an example
    > of ways in which the system can be configured in a way that's deceptive,
    > not the only way this can be done.


    OK, that seems far more reasonable.



  8. Re: Question about Linux

    Nico Kadel-Garcia wrote:
    > "Rod Smith" wrote in message
    > news:51k5b2-s0c.ln1@speaker.rodsbooks.com...
    >
    >>In article ,
    >>"Nico Kadel-Garcia" writes:
    >>
    >>>>Windows 2000 requires, as did Windows NT4, a primary
    >>>>partition on the first hard drive.
    >>>
    >>>Because I've done it. Twice. Without significant difficulty.

    >>
    >>Precisely what did you do, and how? I've seen such claims before, but
    >>whenever details have emerged, it's always turned out to be a case of the
    >>person doing one thing and believing something else was done (such as
    >>installing just the WINDOWS or WINNT directory to a logical partition,
    >>leaving the critical boot files on a primary partition on the first disk;
    >>or using BIOS trickery to swap the identities of two disks, thus enabling
    >>installation to a primary partition on the second disk only because the
    >>BIOS remapped it to look like the first disk). Thus, I treat such claims
    >>with skepticism. If you can back this up with a detailed description or a
    >>link to such a description, I'd love to see it.

    >
    >
    > Two hard drives. I had Linux on the first hard drive, added a second hard
    > drive in a hot-swap bay for testing, and installed Windows 2000 Professional
    > on the second hard drive. As I remember, it over-wrote the MBR, but I was
    > able to fix that with a Linux rescue CD. The hard drive showed up as drive
    > E:, and did its installation on drive E, although I don't remember why.
    >
    > I did a similar operation later with a Windows XP installation on an
    > external USB hard drive, where the internal hard drive had Gentoo Linux on
    > it. If you have an external such drive in hand and a Windows installation
    > CD, it's a trivial test, but I'm a bit short on cycles to re-run that right
    > now. (New job, more work.)
    >
    > And please note, the BIOS trickery you refer to is not necessarily in the
    > BIOS, it can be done by the boot loader itself. That may in fact be what
    > Windows does, that part I can't swear to. But I've certainly used the trick
    > in the Linux world to teach grub or LILO to swap the drives for making boot
    > image mirrors of one drive onto a secondary drive, then loading the MBR to
    > booth drives so if one drive failed the other would still boot.


    Just a note, in most cases the BIOS will boot another drive only in case
    of a hard failure of the first drive. If the first drive just dies,
    you're fine, but if it gets a bad read with data transferred from the
    first drive I have not seen the failover work.

    Clearly this is not something easy to test :-(

    --
    bill davidsen (davidsen@darkstar.prodigy.com)
    SBC/Prodigy Yorktown Heights NY data center
    Project Leader, USENET news
    http://newsgroups.news.prodigy.com

  9. Re: Question about Linux

    In article ,
    "Nico Kadel-Garcia" writes:
    >
    > "Rod Smith" wrote in message
    > news:n4e9b2-l4i.ln1@speaker.rodsbooks.com...
    >> In article ,
    >> "Nico Kadel-Garcia" writes:
    >>>
    >>> "Rod Smith" wrote in message
    >>> news:51k5b2-s0c.ln1@speaker.rodsbooks.com...
    >>>> In article ,
    >>>> "Nico Kadel-Garcia" writes:
    >>>>>
    >>>>>> Windows 2000 requires, as did Windows NT4, a primary
    >>>>>> partition on the first hard drive.
    >>>>>
    >>>>> Because I've done it. Twice. Without significant difficulty.
    >>>>
    >>>> Precisely what did you do, and how?

    >> ...
    >>> Two hard drives. I had Linux on the first hard drive, added a second hard
    >>> drive in a hot-swap bay for testing, and installed Windows 2000
    >>> Professional
    >>> on the second hard drive. As I remember, it over-wrote the MBR, but I was
    >>> able to fix that with a Linux rescue CD. The hard drive showed up as
    >>> drive
    >>> E:, and did its installation on drive E, although I don't remember why.

    >>
    >> I've just tried reproducing this and failed. I put a couple of spare
    >> drives in a 1.2GHz AMD Duron system (first disconnecting the regular
    >> drives). The primary master drive was a 6GB unit that had a small primary
    >> Linux swap partition and a larger primary FreeBSD partition. (I also
    >> tried removing the FreeBSD partition and replaced it with a logical FAT
    >> partition; that change had no effect.) The second drive was a 10GB unit
    >> configured as the primary slave. I put a single primary FAT partition on
    >> the drive to be used as the installation partition. I've tried installing
    >> both Windows 2000 Professional and Windows XP Home on this system. Both
    >> failed. After selecting the primary FAT partition (which the installer
    >> identified as C: in all cases) as the destination and telling the
    >> installer to reformat it, it put up a message to the effect that it could
    >> not proceed without a "Windows 2000-compatible partition" on the first
    >> disk.

    >
    > OK, stop right there. You didn't try my recipe, you put garlic and
    > strawberries in it, and it came out fairly weird.


    How did my test differ from yours? I provided more details than you did,
    but as far as the information you've provided goes, the only differences
    were that your second drive was in a hot-swap bay and mine wasn't, and
    that my first drive had Linux swap and FreeBSD partitions whereas you
    only mention Linux on your first drive (but you don't explicitly say how
    it was partitioned). I wouldn't expect either difference to affect the
    results. I also performed additional tests with Windows XP and with
    logical FAT partitions on the first drive, but that was introducing
    variables in an attempt to get it to work after the initial tests failed.

    If you can provide more information that you think would enable this to
    work, I'm willing to give it a try. Short of that, my inability to
    reproduce your results makes me very skeptical of your claim.

    --
    Rod Smith, rodsmith@rodsbooks.com
    http://www.rodsbooks.com
    Author of books on Linux, FreeBSD, and networking

  10. Re: Question about Linux


    "Rod Smith" wrote in message
    news:vk9db2-08r.ln1@speaker.rodsbooks.com...
    > In article ,
    > "Nico Kadel-Garcia" writes:
    >>
    >> "Rod Smith" wrote in message
    >> news:n4e9b2-l4i.ln1@speaker.rodsbooks.com...
    >>> In article ,
    >>> "Nico Kadel-Garcia" writes:
    >>>>
    >>>> "Rod Smith" wrote in message
    >>>> news:51k5b2-s0c.ln1@speaker.rodsbooks.com...
    >>>>> In article ,
    >>>>> "Nico Kadel-Garcia" writes:
    >>>>>>
    >>>>>>> Windows 2000 requires, as did Windows NT4, a primary
    >>>>>>> partition on the first hard drive.
    >>>>>>
    >>>>>> Because I've done it. Twice. Without significant difficulty.
    >>>>>
    >>>>> Precisely what did you do, and how?
    >>> ...
    >>>> Two hard drives. I had Linux on the first hard drive, added a second
    >>>> hard
    >>>> drive in a hot-swap bay for testing, and installed Windows 2000
    >>>> Professional
    >>>> on the second hard drive. As I remember, it over-wrote the MBR, but I
    >>>> was
    >>>> able to fix that with a Linux rescue CD. The hard drive showed up as
    >>>> drive
    >>>> E:, and did its installation on drive E, although I don't remember why.
    >>>
    >>> I've just tried reproducing this and failed. I put a couple of spare
    >>> drives in a 1.2GHz AMD Duron system (first disconnecting the regular
    >>> drives). The primary master drive was a 6GB unit that had a small
    >>> primary
    >>> Linux swap partition and a larger primary FreeBSD partition. (I also
    >>> tried removing the FreeBSD partition and replaced it with a logical FAT
    >>> partition; that change had no effect.) The second drive was a 10GB unit
    >>> configured as the primary slave. I put a single primary FAT partition on
    >>> the drive to be used as the installation partition. I've tried
    >>> installing
    >>> both Windows 2000 Professional and Windows XP Home on this system. Both
    >>> failed. After selecting the primary FAT partition (which the installer
    >>> identified as C: in all cases) as the destination and telling the
    >>> installer to reformat it, it put up a message to the effect that it
    >>> could
    >>> not proceed without a "Windows 2000-compatible partition" on the first
    >>> disk.

    >>
    >> OK, stop right there. You didn't try my recipe, you put garlic and
    >> strawberries in it, and it came out fairly weird.

    >
    > How did my test differ from yours? I provided more details than you did,


    Drive one is a dual first drive of Linux and FreeBSD, and I have no idea
    what other oddnesses you've done to it., and you're trying to use a
    pre-built FAT32 partitions which you insisted on a somewhat older AMD
    motherboard. With all those rather unusual variables in place, your test
    becomes useful but hardly definitive.

    > but as far as the information you've provided goes, the only differences
    > were that your second drive was in a hot-swap bay and mine wasn't, and
    > that my first drive had Linux swap and FreeBSD partitions whereas you
    > only mention Linux on your first drive (but you don't explicitly say how
    > it was partitioned). I wouldn't expect either difference to affect the
    > results. I also performed additional tests with Windows XP and with
    > logical FAT partitions on the first drive, but that was introducing
    > variables in an attempt to get it to work after the initial tests failed.


    See above. You've added way, way, way too many "but it should still work!"
    variables here. Stop it, please.

    > If you can provide more information that you think would enable this to
    > work, I'm willing to give it a try. Short of that, my inability to
    > reproduce your results makes me very skeptical of your claim.


    I do understand your skepticism. I'd suggest entirely unpartitioning the
    second drive (dd if=/dev/zero of=/dev/hdb count=100), using a simple Linux
    install rather than whatever complexities you've done to the first drive,
    and using NTFS instead of FAT32 for the new install.



  11. Re: Question about Linux

    In article ,
    "Nico Kadel-Garcia" writes:
    >
    > "Rod Smith" wrote in message
    > news:vk9db2-08r.ln1@speaker.rodsbooks.com...
    >> In article ,
    >> "Nico Kadel-Garcia" writes:
    >>>
    >>> OK, stop right there. You didn't try my recipe, you put garlic and
    >>> strawberries in it, and it came out fairly weird.

    >>
    >> How did my test differ from yours? I provided more details than you did,

    >
    > Drive one is a dual first drive of Linux and FreeBSD, and I have no idea
    > what other oddnesses you've done to it., and you're trying to use a
    > pre-built FAT32 partitions which you insisted on a somewhat older AMD
    > motherboard. With all those rather unusual variables in place, your test
    > becomes useful but hardly definitive.


    You're grasping at straws. Why would Windows react any differently to a
    disk with both Linux and FreeBSD partitions than to one with Linux
    partitions only? As to "other oddnesses," that's extraordinarily vague
    and speculative. If the pre-built FAT-32 partition caused any problems,
    the error message (about the OTHER DRIVE) was extremely misleading
    (granted, that wouldn't be too surprising, but still....). The "somewhat
    older AMD motherboard" is newer than Windows 2000 (the BIOS date is
    3/2002). If the motherboard needs to be newer than this to install
    Windows to a second drive with no primary FAT partition on the first
    drive, then at the very least, your claim that Windows can be installed
    in this way needs a big caveat about motherboard or BIOS support. All of
    these factors, if they had any effect, would mean that the conditions for
    installing Windows in this way are extremely fragile and untrustworthy.

    Nonetheless, I tried it as you suggest later in your message: I wiped
    everything from the first drive and set up Linux partitions only on it (a
    primary ext2 partition and an extended partition holding logical ext2 and
    swap partitions); and I wiped everything from the second drive and let
    the Windows 2000 installer create its own partition there. This had no
    effect on the results; Windows still failed to install. (As you suggested
    using NTFS, I'll point out that the installation didn't even get far
    enough to ask whether I wanted to format the target partition as FAT or
    NTFS.)

    >> but as far as the information you've provided goes, the only differences
    >> were that your second drive was in a hot-swap bay and mine wasn't, and
    >> that my first drive had Linux swap and FreeBSD partitions whereas you
    >> only mention Linux on your first drive (but you don't explicitly say how
    >> it was partitioned). I wouldn't expect either difference to affect the
    >> results. I also performed additional tests with Windows XP and with
    >> logical FAT partitions on the first drive, but that was introducing
    >> variables in an attempt to get it to work after the initial tests failed.

    >
    > See above. You've added way, way, way too many "but it should still work!"
    > variables here. Stop it, please.


    I've provided more information than you did. I can't match your
    configuration exactly because I lack the precise hardware you have. If
    your installation method relies on precise hardware, though, it's hardly
    a general solution. As I said, based on the limited information you gave,
    there were only two differences in my initial test -- not what I'd call
    "way, way, way too many" differences. What's more, just one (the hot-swap
    drive) remains in my latest test. As to the variables I deliberately
    introduced (trying Windows XP and replacing the FreeBSD partition with a
    logical FAT partition), I did those things *AFTER* my initial attempt
    failed. Thus, they could not influence the results of that test; they
    were intended as changes that would make the test more likely to succeed,
    but they didn't have that effect.

    Let me ask you this: In your installation, you say Windows has identified
    itself as residing on drive E:. What's drive C:? If you open a DOS prompt
    window and type "DIR C:", what happens? Does the Windows disk
    partitioning tool show a drive C: (on either disk)? In Linux, if you type
    "fdisk -l /dev/hda", is there a primary FAT partition on the drive?

    --
    Rod Smith, rodsmith@rodsbooks.com
    http://www.rodsbooks.com
    Author of books on Linux, FreeBSD, and networking

  12. Re: Question about Linux

    In article , rodsmith@rodsbooks.com wrote:
    >In article ,
    > "Nico Kadel-Garcia" writes:
    >>
    >> "Rod Smith" wrote in message
    >> news:vk9db2-08r.ln1@speaker.rodsbooks.com...
    >>> In article ,
    >>> "Nico Kadel-Garcia" writes:
    >>>>
    >>>> OK, stop right there. You didn't try my recipe, you put garlic and
    >>>> strawberries in it, and it came out fairly weird.
    >>>
    >>> How did my test differ from yours? I provided more details than you did,

    >>
    >> Drive one is a dual first drive of Linux and FreeBSD, and I have no idea
    >> what other oddnesses you've done to it., and you're trying to use a
    >> pre-built FAT32 partitions which you insisted on a somewhat older AMD
    >> motherboard. With all those rather unusual variables in place, your test
    >> becomes useful but hardly definitive.

    >
    >You're grasping at straws. Why would Windows react any differently to a
    >disk with both Linux and FreeBSD partitions than to one with Linux
    >partitions only? As to "other oddnesses," that's extraordinarily vague
    >and speculative. If the pre-built FAT-32 partition caused any problems,
    >the error message (about the OTHER DRIVE) was extremely misleading
    >(granted, that wouldn't be too surprising, but still....). The "somewhat
    >older AMD motherboard" is newer than Windows 2000 (the BIOS date is
    >3/2002). If the motherboard needs to be newer than this to install
    >Windows to a second drive with no primary FAT partition on the first
    >drive, then at the very least, your claim that Windows can be installed
    >in this way needs a big caveat about motherboard or BIOS support. All of
    >these factors, if they had any effect, would mean that the conditions for
    >installing Windows in this way are extremely fragile and untrustworthy.
    >
    >Nonetheless, I tried it as you suggest later in your message: I wiped
    >everything from the first drive and set up Linux partitions only on it (a
    >primary ext2 partition and an extended partition holding logical ext2 and
    >swap partitions); and I wiped everything from the second drive and let
    >the Windows 2000 installer create its own partition there. This had no
    >effect on the results; Windows still failed to install. (As you suggested
    >using NTFS, I'll point out that the installation didn't even get far
    >enough to ask whether I wanted to format the target partition as FAT or
    >NTFS.)


    Suggestion;
    Boot with a Grub floppy, go to its command line and hide the Linux drive:
    hide (hd0)

    You should be able to install W2k then.

    To bring back the Linux drive use the Grub floppy again and unhide (hd0). To
    run windows you will probably have to use the Grub map command to swap the
    drives.

    Once you figure out what works make a menu for Grub.

    Dr. G.


    >
    >>> but as far as the information you've provided goes, the only differences
    >>> were that your second drive was in a hot-swap bay and mine wasn't, and
    >>> that my first drive had Linux swap and FreeBSD partitions whereas you
    >>> only mention Linux on your first drive (but you don't explicitly say how
    >>> it was partitioned). I wouldn't expect either difference to affect the
    >>> results. I also performed additional tests with Windows XP and with
    >>> logical FAT partitions on the first drive, but that was introducing
    >>> variables in an attempt to get it to work after the initial tests failed.

    >>
    >> See above. You've added way, way, way too many "but it should still work!"
    >> variables here. Stop it, please.

    >
    >I've provided more information than you did. I can't match your
    >configuration exactly because I lack the precise hardware you have. If
    >your installation method relies on precise hardware, though, it's hardly
    >a general solution. As I said, based on the limited information you gave,
    >there were only two differences in my initial test -- not what I'd call
    >"way, way, way too many" differences. What's more, just one (the hot-swap
    >drive) remains in my latest test. As to the variables I deliberately
    >introduced (trying Windows XP and replacing the FreeBSD partition with a
    >logical FAT partition), I did those things *AFTER* my initial attempt
    >failed. Thus, they could not influence the results of that test; they
    >were intended as changes that would make the test more likely to succeed,
    >but they didn't have that effect.
    >
    >Let me ask you this: In your installation, you say Windows has identified
    >itself as residing on drive E:. What's drive C:? If you open a DOS prompt
    >window and type "DIR C:", what happens? Does the Windows disk
    >partitioning tool show a drive C: (on either disk)? In Linux, if you type
    >"fdisk -l /dev/hda", is there a primary FAT partition on the drive?
    >


  13. Re: Question about Linux

    In article ,
    DRGrok@munged.net (Dr. Grok) writes:
    >
    > In article , rodsmith@rodsbooks.com wrote:
    >
    >>Windows still failed to install.

    >
    > Suggestion;
    > Boot with a Grub floppy, go to its command line and hide the Linux drive:
    > hide (hd0)
    >
    > You should be able to install W2k then.


    This particular subthread isn't really about practical suggestions for
    installing Windows; it's about whether Windows itself supports
    installation entirely to a second physical disk, without using extra
    tools like GRUB or BIOS options to hide the first physical disk. Nico
    Kadel-Gar has been saying Windows can do this, but so far he's failed to
    convince me. As a practical matter, this isn't a big deal; as you say,
    using boot loader trickery can work around the problem.

    Personally, I'm interested in this because I write about computers for a
    living, so if there's a way to do it, I want to know so I can write about
    it if the need arises. I already know about the GRUB options, and in fact
    have written about them in print; for instance:

    http://www.linux-mag.com/2003-12/guru_01.html

    --
    Rod Smith, rodsmith@rodsbooks.com
    http://www.rodsbooks.com
    Author of books on Linux, FreeBSD, and networking

  14. Re: Question about Linux


    "Rod Smith" wrote in message
    news:nc7gb2-hu3.ln1@speaker.rodsbooks.com...

    > Nonetheless, I tried it as you suggest later in your message: I wiped
    > everything from the first drive and set up Linux partitions only on it (a
    > primary ext2 partition and an extended partition holding logical ext2 and
    > swap partitions); and I wiped everything from the second drive and let
    > the Windows 2000 installer create its own partition there. This had no
    > effect on the results; Windows still failed to install. (As you suggested
    > using NTFS, I'll point out that the installation didn't even get far
    > enough to ask whether I wanted to format the target partition as FAT or
    > NTFS.)


    Huh. That's a much better experiment. You stil have some uncertainty left
    with the motherboard (it's Windows after all, and AMD has never been as well
    supported as Intel), but if I can find some cycles I'll check it again
    myself. Unfortunately, a lot of Windows PC vendors send their media as a
    "blow the entire partitions away and install it our way".

    >> See above. You've added way, way, way too many "but it should still
    >> work!"
    >> variables here. Stop it, please.

    >
    > I've provided more information than you did. I can't match your
    > configuration exactly because I lack the precise hardware you have. If


    True. And this was over the last few years, and even I don't have the
    hardware anymore.

    > your installation method relies on precise hardware, though, it's hardly
    > a general solution. As I said, based on the limited information you gave,
    > there were only two differences in my initial test -- not what I'd call
    > "way, way, way too many" differences. What's more, just one (the hot-swap
    > drive) remains in my latest test. As to the variables I deliberately
    > introduced (trying Windows XP and replacing the FreeBSD partition with a
    > logical FAT partition), I did those things *AFTER* my initial attempt
    > failed. Thus, they could not influence the results of that test; they
    > were intended as changes that would make the test more likely to succeed,
    > but they didn't have that effect.


    Good. You've been doing good experimentation, and my experience is losing
    credibility. I do remember it, but you've actually succeeded in making me
    slightly suspicious of my memory.

    > Let me ask you this: In your installation, you say Windows has identified
    > itself as residing on drive E:. What's drive C:? If you open a DOS prompt
    > window and type "DIR C:", what happens? Does the Windows disk
    > partitioning tool show a drive C: (on either disk)? In Linux, if you type
    > "fdisk -l /dev/hda", is there a primary FAT partition on the drive?


    Nope. If I put in a CD drive, it gets mounted as drive C:, which confuses
    the hell out of some installers. Unfortunately, I don't have that box
    anymore to play with (it was at my previous workplace).



  15. Re: Question about Linux

    In article ,
    "Nico Kadel-Garcia" writes:
    >
    > "Rod Smith" wrote in message
    > news:nc7gb2-hu3.ln1@speaker.rodsbooks.com...
    >
    >> Let me ask you this: In your installation, you say Windows has identified
    >> itself as residing on drive E:. What's drive C:? If you open a DOS prompt
    >> window and type "DIR C:", what happens? Does the Windows disk
    >> partitioning tool show a drive C: (on either disk)? In Linux, if you type
    >> "fdisk -l /dev/hda", is there a primary FAT partition on the drive?

    >
    > Nope. If I put in a CD drive, it gets mounted as drive C:, which confuses
    > the hell out of some installers. Unfortunately, I don't have that box
    > anymore to play with (it was at my previous workplace).


    That's odd, but not unheard-of. Windows 2000 and XP (and I believe NT) can
    remap drive letters. They don't normally do so by default, but it's
    conceivable you remapped things (possibly without even realizing it). The
    contents of what Linux calls /dev/hda would be much more diagnostic.

    (Oh, and I didn't mean to ask just about FAT on that drive; Windows 2000
    will work with an NTFS partition there, as well.)

    --
    Rod Smith, rodsmith@rodsbooks.com
    http://www.rodsbooks.com
    Author of books on Linux, FreeBSD, and networking

+ Reply to Thread
Page 3 of 3 FirstFirst 1 2 3