Longhorn on Virtual PC 2004

by Michael S. Kaplan, published on 2005/05/07 04:45 -04:00, original URI: http://blogs.msdn.com/b/michkap/archive/2005/05/07/415335.aspx


Well, I had a few idle moments today, so I thought I would try to do what a ton of Beta sites would be doing as soon as Beta 1 of Longhorn hit the streets. Install it on a Virtual PC image.

These days, the smart beta testers are always using products like Virtual PC for their betas -- better to keep one's own machines clean, right.

So, I got started....

No, I am not going to give everyone a ton of screen shots. :-)

I thought I would point out one problem I had though. When I tried to pick the drive to install to, Longhorn would not recognize the VPC drive. To get it to recognize the drive, I did the following:

  1. Started to install Windows XP
  2. As soon as the initial NTFS formatting of the drive was done, I removed the CD and chose to reset the VPC session
  3. I then installed Longhorn -- it was able to recognize the drive once it was formatted

It was a fairly easy workaround, but it did inspire a momentary panic when I first thought that Virtual PC might not be available for the beta. There may not be an immediate fix it it is a Virtual PC 2004 issue, after all.

I then played around a bit after it was installed, everything seemed functional. I then thought I would test out something that had been pointed out by a colleague who had been experimenting with a small setup change (translation: "was goofing off in the late afternoon"). I did the same somewhat destructive act on the registry and then rebooted my virtual Longhorn image. In prior versions this would cause the following effect:

Not the most intuitive message, huh? Well, you can file this next bit in the "truth is stranger then fiction" department.

A while back, BBspot ran an article entitled Microsoft: The Next 25 Years. One of the predictive entries read:

2016    Microsoft enhances their only innovation by introducing the "Red Screen of Death." 

Well, here on Longhorn, roughly 11 years early:

Well, that will teach me, right? Luckily the recovery was easy (just rebooted, went to "last known good", undid that destructive operation, and then rebooted. And all was right with the world.

I am not sure I would class the change as an improvement. I mean, the old message ("Windows cannot start because the following file is missing or corrupt: \WINDOWS\System32") is one I understand since I know what the destructive operation was. But if I did not have that knowledge then I would not know much about what was going on.

The new message, though, is even harder to understand (though an internet search of 0xc000000f will see it relates to Windows File Protection, which I guess is a hint.

Lucky for me that this is not the sort of thing one sees unless one is intentionally performing destructive operations, huh? :-)


# Anonymous on 7 May 2005 4:34 AM:

You might want to log a bug about the spelling of "execution" in that error message :)

# bg on 7 May 2005 7:12 AM:

i was thinking of putting the next available (public) longhorn build on a virtual pc, but i read that virtual pc doesn't do 3d acceleration, so i wasn't sure i'd get much of a user experience!

bg

# M Freitas on 7 May 2005 7:18 AM:

Arghhhh! I am too trying to install Longhorn on my VPC 2004 SP1 and it comes with the same errors. What a pain :(

# Michael S. Kaplan on 7 May 2005 8:45 AM:

Well, at least there is an eay workaround, right? :-)

# Michael S. Kaplan on 7 May 2005 9:14 AM:

>> You might want to log a bug about the spelling of "execution" in that error message :)

Well, it would be a low-priority bug. It is insanely hard to make it happen, not in terms of difficuly of operation but in terms of dificulty of knowing the operation to do....

# Heath Stewart on 7 May 2005 9:32 AM:

Something must've changed from the PDC '04 bits since it installed fine (albeit slowly) on VPC 2004 for me.

I just also wanted to mention that VPC works great for remote debugging when you don't have a ton of machines sitting around at home.

# Michael S. Kaplan on 7 May 2005 9:43 AM:

Hi Heath!

Yes, lots of Longhorn changes since that PDC build. But I have been hitting this problem with the unrecognized drive pretty consistently....

# Sushant Bhatia on 7 May 2005 4:09 PM:

EEEKK! Red Screen of Doom! Why make these colors so bold. There must be some research on visual imaging to suggest a color that would make people feel good about an error :-) How about ocean blue...with two margarita bottles and a plam tree in the right bottom corner.

# Jonathan on 8 May 2005 1:29 PM:

You can get almost any color you want in text-mode - you can only get 16 colors at once, but each color can be chosen from 256K colors (18-bit color depth - 6R 6G 6B). I remember doing that in DOS through the BIOS (INT 10h - slow!) or directly through the VGA card (OUT 3C8, was it? It was really fast!).

# yourFreindlyLibPNG() on 8 May 2005 7:30 PM:

Lovely compression artifacts on that RSOD.

> Lucky for me that this is not the sort of thing one
> sees unless one is intentionally performing
> destructive operations, huh? :-)

Is installing longhorn is a destructive operation?

# Michael S. Kaplan on 8 May 2005 8:08 PM:

On a Virtual PC? Of course not!

I am not putting it on my main machine just yet. But I have several builds intalled on other machines....

# Anonymous on 8 May 2005 9:58 PM:

Here's a thought, get an OS that actually works. There are thousands of them out there.

# Michael S. Kaplan on 8 May 2005 10:09 PM:

>>>> Here's a thought, get an OS that actually works. There are thousands of them out there.

Um, huh? You think I could not take one of those thousands, recompile critical data files, and make them unbootable?

Longhorn does quite well when I am not messing with it!

# Larry Osterman [MSFT] on 8 May 2005 10:45 PM:

Congrats Michael, you're on the front page of /. (via Joi Ito, who picked up on the new RSOD in Longhorn)

# Michael S. Kaplan on 8 May 2005 10:45 PM:

Hmmmm.... weird thing to be congratulated for, huh? Like congratulations for catching that awful disease or something.... :-)

# Larry Osterman [MSFT] on 9 May 2005 2:21 AM:

The wierd thing is that I've never ever seen this error on my Longhorn machine - and I've broken a bunch of things on my machine over the past couple of months.

I've blue screened it any number of times but never a red screen. What did you do to get it? Or was this a VPC-specific thing?

Or is this a corruption being detected in ntldr?

Feel free to reply offline, you know my email address :)

# Michael S. Kaplan on 9 May 2005 7:27 AM:

It is all about the timing. :-)

A blue screen of death is also something I see a lot of -- cf: http://blogs.msdn.com/michkap/archive/2005/04/15/408474.aspx

Just an occupational hazard of not only making data updates, but also owning part of the tools that build the data!

I'll turn the question around -- had you ever seen that message I had the Server 2003 screenshot of? It is pretty hard to get, but if you had an NT 3.1 box set up, I am antisocial enough to make it happen there, too. :-)

# Simon Gnieslaw on 9 May 2005 7:41 AM:

About the Longhorn boot screen, that is the first time that I have seen it. I made a boot screen almost EXACTLY like that as far are the White Windows Flag with the small "TM" logo goes. What I did was copy the setup.bmp from the Windows XP install CD and converted it to a 16-colour 640x480 bitmap with a teal background and used it as the background for the PowerQuest BootMagic boot-loader OS selection menu.

I also use the same bitmap as the Startup splash screen of my BIOS. I just find it a remarkable co-incidence that the image that I have been using for all my pre-boot sequences except for the Actual OS startup screen (the BIOS and Boot Loader) has as of longhorn so far, has become part of the OS startup screen.

Or maybe I am mistaken and the one of the XP CD was already 16-colour but the background was not a solid background colour, it alternated between black and blue pixels instead of a nice smooth teal/black.

# Michael S. Kaplan on 9 May 2005 7:53 AM:

My favorite comment from the /. thread, from Rabbit Hole: "Why, oh why, didn't I take the blue screen?"

Now *that* was funny!

# Andre on 9 May 2005 1:41 PM:

Its about time they used a different colour, I was getting sick of the dang blue!

# Michael S. Kaplan on 9 May 2005 3:47 PM:

Andre -- the blue would still be there, in any case where it was before.

# Anonymous on 11 May 2005 12:09 PM:

> 2016 Microsoft enhances their only innovation
> by introducing the "Red Screen of Death."

> Well, here on Longhorn, roughly 11 years early:

Hey, it isn't shipping yet... :-)

referenced by

2015/09/04 It was almost 10⅓ years ago when I blogged about a 'Red Screen of Death' aka RSoD, and...

go to newer or older post, or back to index or month or day