I've had something similar happen to me consulting for this company... For whatever reason they don't like the way Oracle formats their checks, because Oracle expects the checks (checques?) to all have their routing & account numbers already on the check in the glorious MICR font. However this company wants to be able to select accounts at will, along with numbers...

Anyways, I'm a consultant looking at their firewalls, and they have WAY too many, and I'm pushing this "new" thing of virtualization. So using Virtual PC, on XP I show how I can take their 5 firewalls and consolidate them to a single box.. And they are loving it as it's freed up almost an entire rack of crap hardware. Then they ask me to look at other 'easy targets' for virtualization, and I look at this desktop running Windows 95.. And yes, this old programmers desktop gets all the check orders from Oracle, and ... PRINTS ALL THE CHECKS.

You name it, payroll, orders, everything.

And it's on this Pentium 75 with a 200MB hard disk that when I listen to it, I can hear a definite "ping" noise from within the case.

So I power down the machine, pop the disk out, and xcopy the filesystem into another machine to then copy into Virtual PC. Once done, I put the original machine back in the data center to then finish up the Virtual PC to demo to the CIO on how great this thing is, and how I can preserve existing application configurations (bla bla bla bla).

I'm not kidding, as soon as I have the VM booting on my laptop, the DBA busts down the doors screaming that the check printing is down! (and naturally that since I touched it last, I must have killed it). So I had my VM going, I just put it on the network live, and suddenly I'm printing all the checks from my laptop... Everyone is happy as I have just saved them from an impending hardware disaster..!

Things go so well I ended up taking a full time job with them!

Anyways the moral is that 'living' systems have a hope of being restored as the media is new.... And don't let things 'run forever' in the corner as when they eventually fail as they always do it'll be catastrophic.

As you can tell I'm a big proponent of virtualization, I'd recommend people seriously think about it, as it provides a significantly easier route to constantly move systems to newer hardware, skipping many of these issues all together.

We've only been waiting FOREVER for a Win64 version of Flash from adobe...

Although I will say this, the state of drivers for Win64 is far better then the early days (NT 3.1/3.5) of Win32.

Open up pif editor, and click the "Background" execution button, then hit the advanced button, here you can set the priority levels of background & foreground MS-DOS sessions (set them both to the same number, like 50), and click to allow Windows 3.0 to detect idle time.

Save that PIF, and launch it with some command.com's and run them in a window and watch how they all execute at the same time.

It's been this way since Windows/386, that was it's big thing was that back in 1988 Microsoft could do multiple MS-DOS sessions in a preemptive manner while OS/2 could only do one.

this is why Microsoft wanted to abandon the 286, but IBM held them to this goal...

At least you got the OS/2 stuff in there! Many people forget that Windows 3.0 was microsoft's plan B for the whole OS/2 disaster. Another factor in the whole thing was as Balmer likes to put it was DEVELOPERS.

IBM had decided that the SDK's should be a revenue source, and they charged a fortune for them! Meanwhile Microsoft was happily giving away Windows SDK's and would allow 3rd party compilers to include the SDK components for free!

The real deciding factor for a lot of people getting started was Microsoft QuickC for windows 3.0. It was $199 new and $89 if you were 'upgrading' from something else. This is significantly cheaper then the OS/2 tools that run upwards of $3000!

The DOS Extender market was also heating up, and for the 286 crowd you couldn't beat the 'runtime' charge of Windows 3.0 vs what people like Rational were charging for their dos extenders.

At the time for the $150 for Windows 3.0 & $199 for QuickC for Windows made this the cheapest way to write programs that could run in protected mode.

Another thing is that most people forget that Windows NT started it's life in 1988 as NT/OS2 the RISC portable version of OS/2. It wasn't until Windows 3.0 was clearly a hit, did they dump the primary 32bit OS/2 2.0 'cruiser' API for an expanded 32bit Windows API based on the WLO (Windows Libraries on OS/2). The great book showstopper goes into great detail about the rise of Windows NT.

Windows/386 could preempt MS-DOS sessions, remember it's the VMM infrastructure from Windows/386 that made Windows 3.0's VXD's work.

You can fit Windows/386 & MS-DOS 3.3 on a floppy and boot it up on any modern pc and you can see for yourself on real iron, or any decent emulator can run it...

All the preemption isn't on by default, but you just configure it via a pif file... It's not that hard.

install oracle on Netware 4 & SQL on NT 3.1 then come back to me on which is 'inferior'. Sorry but Netware languished in that 1980's feeling of the more insane the install & configuration the better.

There is a reason that NT took over all that server space that was occupied by Netware & OS/2.

I hate to say it, but look around the office, and how many people will have 50 copies of word running, and with 50 versions of the same document because every time they reply to an email, they keep opening the same document over & over... Companies refuse to train their users as time has gone by, and of course the users aren't going to learn on their own. I'm just happy I haven't done end user support in years.

ALT+TAB.... it was super easy to switch between stuff. good grief.

As a matter of fact, I just found a copy of MS Excel 3.0 for OS/2... on 3 diskettes!!!!

But it's cool, back then they had printed documentation! Now if only I had a real reason to have Excel 3.0 for OS/2... Or a lead to get Word for OS/2...

Windows NT is OS/2 3.0 .... I know it's hard to forget under all of the Win32/Win64 dressing, but as you have mentioned it was written portable with its first target being the i860, then the MIPS. Along the way they even changed it's primary API set, but it still came along....

DEC wouldn't let 3rd party vendors have SMP. And what chips they did sell had to have insane markups. DEC was interesting in selling DEC systems, not competing with Intel. The industry just wanted out of the x86 market a the time, hell FX!32 was some pretty dammed great engineering for the time (ha and now look at transitive...!)

But then the whole RISC thing was just 'common sense' back then... It was so imperative that everyone port from their CISC into a RISC.. And where are they today? There is no doubt that DEC would have been better served porting Tru64 & VMS to the x64. But we all know that proliants aren't the high end market compared to the Alpha/HPPA/Itanium machines.

And out of all this, who morns for SGI & MIPS? The itanium being so late to market sealed their fate.

I still keep my Alpha running for fun, and it's great to verify 'correct' code... somethings fly on the i386/MIPS but don't on the Alpha..

Windows NT was written for the i960 RISC chip then the MIPS cpu before an i386 port was even in the works. Good grief. I guess the history of NT it too passe these days, but DEC had MS write it, because MS stole the west coast hardware/software team after killing their projects. Why on earth would I run Linux? It barely runs on the x86, let alone any other platforms. Enjoy the hell that is modules, eth0 and the constantly changing ABI.

The Compiler for Windows was the *WORST* I've ever dealt with. Even the Beta x86 32bit compilers were a dream come true, compared to the Alpha compilers. Even trival programs like gzip couldn't build with /O2 flags!!! The sad part, is by the time Visual C++ 6.0 for the Alpha shipped it was finally usable... And then they killed the platform.

Now I know you want to leap right in and defend DEC from the crappy C compiler, but their name was right in the copyrights....

Microsoft (R) & Digital (TM) AXP C/C++ Optimizing Compiler Version 8.03.JFa
Copyright (C) Microsoft Corp 1984-1994.
Copyright (C) Digitial Equipment Corporation 1992-1994.
All rights reserved.

And I still recall when the NT source was leaked, a lot of the swearing in the code was pointed towards the lackluster Alpha compiler.

But then I must be mad, as I keep my Alpha running NT 4.0.

for imported softdrinks that contain cane sugar. Because HFC SUCKS!

56k serial ports in the early 1980's? Dream on, the 8250 was buggy, and had major issues.