The second virtualization contest is now on! Similar to the previous one, the winner receives $100 via Paypal and the submission is posted on this blog! Hopefully this one will be little bit more challenging. 🙂
The progress so far: A few years ago I came in to possession of a set of floppy disks pictured here:
Thanks to Al Kossow from bitsavers.org the floppy disk content has been recovered. Michal Necasek of OS/2 museum successfully converted them in to an usable format and made some modifications to get them to boot on VirtualBox:
Couple of years later, thanks to Andrew Gong, a WYSE Unix tape has been found on eBay:
More recently Al Kossow was able to read the tape in to an image, which now I have uploaded to my web server: wyseunix321a.zip
The next step is yours! Install the whole system on to a hypervisor of your choice, document the process and supply a vanilla boot image or VM.
The winner shall be the person who will first post a comment declaring success including a screenshot and can further prove it by emailing emailing me the submission shortly after. If the comment gets blocked by spam filter, don’t worry the original submission time will of course count. Oh and almost forgot: I also need aclock binary for it, however if there is no compiler and the standard SysV binary works fine, you are exempt from the requirement.
The catch? Looks like floppy disk trouble. The boot disk is fine after it has been fixed up by Michal. The Base floppy looks like has same content as boot. Copy Tools is very small. Looks like it may be truncated. Hopefully not, but if yes I count on your creativity. Remember that Dell Unix is an exactly save release of SystemV/386 and did not have or needed copy tools to install.
Update: Looks like the contest has been won by Mihai! Congratulations!
Since the last update we got some help in a few fields that have really fleshed out this ‘experimental’ port into a full fledged port. First RayerR helped us with the fun of getting us onto the latest deployment of DJGPP, 2.05 (rc1). It’s always nice to be in the latest available release. Next in a passing comment, Ruslan Starodubov had mentioned that he had gotten a much older build of our QDOS to support the Intel HDA sound chipset via the MPXPlay sound library. I wrote to the author of MPXPlay, Pádár Attila asking for us to distribute his source in our project, and he granted permission.
So at this point things were looking good. The only ‘feature’ that modern OS’s really held over us was the ability to dynamically load and unload game modules on the fly. I had tried to use DLM, but it stripped the DPMI functionality out of the MS-DOS Extender making the port really useless. I tried to build the newer DXE3 support but had no luck. I suspect now my native tool chain was interfering with the build process. But Maraakate managed to get it to not only build, but to run!
Adding in DX3 support was relatively painless. I first looked at DJGPP’s FAQ and downloaded the example code. In the example code there was small helper functions to make unions and check the symbols. If they didn’t exist a printf was spit out to alert you about it. To resolve the issue you simply just add DXE_EXPORT to the other list of missing exports.
Compiling the game code was easy, again referring to the example I saw that basically they compiled it the same, but at link time you use DX3GEN and -U flag to ignore unresolved symbols.
The biggest head scratcher was the Sys_GetGameAPI failing to find GetGameAPI from the DX3. After some piddling around I noticed that it listed GetGameAPI as _GetGameAPI inside the DX3 itself. I added the underscore and it worked!
Other things that were relatively to easy to import was R1Q2’s HTTP downloading code. Compiling CURL was kind of tricky because of the linking order, but thankfully neozeed figured it out quickly.
All of Yamagi’s Quake 2 updated game DLLs were all diff’d by hand using BeyondCompare to make sure I didn’t clash using some newer functions that weren’t available in DJGPP. I also merged their Zaero code with their baseq2 code by comparing Zaeros code to the Quake 2 SDK, marking every thing that was changed. The result is a really stable Zaero game code. If you haven’t played Zaero check it out. I think it’s a lot better than Rogue, but Xatrix is probably my favourite (even over stock Q2).
Other cool things I’m glad to get into the code was the GameSpy Browser. It took me quite a bit of work to get it where it is, but it’s really nice to just be able to ping to a master server (a custom GameSpy emulator I wrote specifically for Q2DOS. Source is not finalized yet, but will be available soon for those curious), pick a server and go! All in DOS!
So here we are at the end of the journey. Or at least safe enough for a 1.0 release.
To recap, we have:
* SVGA (LFB modes only)
* SoundBlaster and Gravis UltraSound Family
* CD-ROM music
* OGG music
* Networking (You need a packet driver)
* Loading/unloading game DLLs in DX3 format.
* Intel HDA support -hda
* Mouse wheel support with -mwheel
Well from popular request I finally got around to loading this up. I went ahead with my favourite retro emulator, PCem for this, as it can nicely emulate an EGA display, unlike most emulators which do VGA, however when it comes to older versions of Microsoft products they really can detect the difference between EGA and VGA.
So to start off, I downloaded from the project page, this version of PCem, compiled it, and installed MS-DOS 4.01 , from April of 1989. The Windows 3.0 Debug Release 1.14 itself is dated from February 22nd, 1989. Which I figured is close enough to the time period. I’m using the 486SX2/50 because I’m too impatient for the 386 speeds, but it does work fine on 386 or higher emulators. It does NOT work with any 286 emulation. I’m also using the HIMEM.SYS from MS-DOS 4.01 vs the one with the Windows 3.0 (Alpha? Beta? Technical Preview?) since it is slightly newer.
There is no setup program per say, rather it just xcopies all the files to a directory, and from there you run ‘d.bat’ and away you go. This version is hard coded to an EGA display, which again is the reason I went with PCem. Once you start it up, you are greeted with:
And it identifies itself as Windows Version 2.1
And first thing to notice is that on my setup with 8MB of ram, I have over 6MB of RAM free. Compare this to regular Windows 2.1 which gives me 399Kb of ram in my current setup.
And with Windows/386 Version 2.1 it provides 383Kb of real memory, along with 6.7MB of EMS memory, as the Windows/386 Hypervisor includes EMS emulation.
Of course the major limitation of Windows 2 is that it runs in real mode, or in the case of Windows/386 an 8086VM. As I mentioned a while back in a post about Windows 3.0, This was game changing.
As now with Windows running in protected mode, all the memory in my PC is available to Windows, and I am using MS-DOS, with nothing special.
Besides the limitation of being EGA only, the Debug version of 3.0 is that there is no support for MS-DOS applications, as WINOLDAP.MOD is missing.
This is clearly an interim build of Windows 3.0 as mentioned in Murray Sargent’s MSDN blog Saving Windows from the OS/2 Bulldozer. As mentioned from the article they began their work in the summer of 1988, so considering this is early 1989 it shows just how much progress they had made in getting Windows 2 to run in protected mode. Along with Larry Osterman’s MSDN blog post Farewell to one of the great ones, which details how the Windows 3.0 skunkworks project was writing the new improved 386 hypervisor, and how Windows 3.0 got the green light, and changed the direction of not only Microsoft but the entire software industry.
I’ve been able to run most of the Windows 2.1 applets, however I’ve not been able to run Excel 2, or Word 1. I suspect at this point that only small memory model stuff from Windows 1 or 2 is capable of running. Although at the same time, when 3.0 did ship, you really needed updated versions of Word 2 and Excel 3 to operate correctly.
The applets from Windows 2.1 seem to work a LOT better than the one from Windows/386 2.1 if that helps any.
This is an interesting peek at an exceptionally early build of Microsoft Windows.
2014/4/15 I has integrated source of i386 and i286 edition edition. In addition, in the i286 version, I added support for int 10h/16h. equivalent to 0.149 MAME, I was replaced with a 0.152 equivalent MAME core i386 i286 core. However, the i386 core, I have omit the TLB around.
Which is very cool, although I wasn’t sure about the MAME source code being open to other projects…? I tried to contact the i86/i386 author vlinde but he then pulled his contact page. I wanted to use i386 for something of my own, but the whole “Redistributions may not be sold, nor may they be used in a commercial product or activity.” really puts the damper on it.
I was able to get some simple XMS test program to run, but nothing of any substance. No DOS4G/W or anything like that. But if you re-build it specifying MS-DOS version 5.0, some of the MS-DOS utils and even command.com work!
The weird issue I had was running out of conventional RAM, because this program gives you nearly 1MB of conventional RAM… I was surprised, as I wasn’t expecting that much!
But sadly I had a lowly 4MB 386sx-16 with CGA, so things like this game with it’s awesome VGA graphics were an impossibility.
Even more sad at the time was that ‘primitive’ 3d games like wolfenstein 3d also required VGA.
But as we all know a few short months later, DOOM was released, and then Terminator Rampage was quickly swept off the face of the earth.
I recently came across this page, and I thought I’d give it a shot. The requirements are pretty ‘minimal’:
Minimal 386DX-25 with 4MB of Ram, VGA & 18MB of disk space
Recommended 386DX-33 or 486DX-33
So I was thinking Qemu could easily run this game. Long story short, it doesn’t work. Turns out Rampage needs EMS. And for whatever reasons, running emm386.exe on Qemu (I tried a handful of versions) just crash on Qemu after initialization. Failing with stock Microsoft EMM386, I tried JEMM, which loaded, and ran it’s built in EMS diagnostics OK, but trying to run Rampage resulted in a nice crash.
So giving up on Qemu, I tried it on DOSBox. It runs but it is incredibly jerky. So I thought I’d try PCem, and see how it runs there.
So the plus side is that PCem, is able to run MS-DOS & EMM386.EXE without issues. It only took a few minutes to install MS-DOS 5.0 and Rampage on my ‘virtual’ 486DX2/66 with 8MB, of ram, and load up Rampage to be greeted with it’s jerky motion.
Thinking its something with emulation in general I fire up Norton SI to get some PCem scores how it benches against known good samples.
Then comparing the scores to this handy (if not ancient) Norton Si benchmark spread we see:
Well now that interesting, so at a ‘raw’ CPU level, PCem is delivering on what would be classical performance. So for the heck of it, I load up DOOM, and it runs a bit choppy on the 386, but flies on the 486 & Winchip emulation. Now that is strange. And just to confirm…
They really thought this would be playable on a 486 @ 33Mhz.
So how does it choke? While going straight is ‘ok’ turning around is so utterly sluggish that there is no feeling of immersion. It feels like you are driving an incredibly slow tank. At the same time, the more realistic sprites, and textures serve to make it look even more unrealistic.
So what am I talking about?
Well here is a screenshot of Wolfenstein 3d on the 386DX-33 (and more than playable).
As you can see, there is no ceiling, and no floor textures. The walls are all uniform height, and the textures were clearly drawn by hand, giving it a very fake and ’16bit’ feeling. I should also add on a capable 286, this game is playable.
Now at first it looks like it has a lot in common with the soon to be release DOOM, with textured ceilings and floors.
Now as you can see the difference in DOOM is the 2.5D effect of there being lower areas so you can go up and down stairs (while you cannot go under them). Also Doom introduces dynamic lighting, and better sound rendering.
While I do like Rampage’s upfront map, as you can see thought, it is very square. In a small effort to ‘speed’ up Rampage you can turn off the ceilings and floors revealing a very Wolf3d like environment. Unfortunately the more they tried to give the world detail, the more it well just looks flat.
Which kind of kills the whole thing. Maybe they should have left out things like water fountains.
Then you get things like this computer setup (one of the programmers? The accountant’s lamp is a nice touch) but it’s a sprite, so as you rotate around it, you always see the same face.
Ironically it’s these high resolution background sprites that make the environment feel less real, as they make the rooms feel too open, and too sparse.
It is the real paradox that in a good shooter you have lots of room, and things to duck behind, but the rooms feel too large, and look bizarre with the massive open spaces. But it is more so a limitation of the time, with the engine being more of an improved Wolf3d engine, than taking a larger leap into being something more 2.5d or 3d like Doom (or the distant Quake).
Another thing that really bugged me was the doors.
In doom, the doors felt more ‘natural’ in that they weren’t super wide.
But in rampage they are stretched wide giving the impression of why you can’t turn is you are incredibly wide..
Even the ‘exit’ door texture still feels too wide.
I could probably get by the empty spaces, but it takes so long to turn around, and the controls feel so unnatural (they don’t even try to be a Wolf3d control-a-like) that it really feels klunky. No matter what speed you play it at.
As always testing is very minimal, all I’ve done is installed MS-DOS 6.22 & Doom 1.1, and tested the SoundBlaster 16 emulation. And as with the pre-release versions, the adlib code is still broken. And Ive done the ‘better’ fix in this code regarding that.
I haven’t run anything else, including fun things like the PowerPC & OS X emulation, MIPS with Windows NT, or even trying anything x64 based as I’m sure it is still broken from back in the Qemu 0.90 days.
This is probably the most significant Xenix post I’ve made since the old days when I managed to get Xenix running in Qemu all those years ago.
What has long been a frustration with the beleaguered Xenix community is that although there was a TCP/IP package for Xenix (and a much required streams package…) it only worked with a handful of ethernet cards. And all of them were early 3com’s. While the world was using NE2000’s on just about everything, the most common ethernet board Xenix would talk to was the 3c503, which is getting harder and harder to find as the years go on by.
Now I ended up having to download the source to mame 0.151 (mame0151s.zip) and building it on OS X. Of course remembering to alter the makefile to include the ‘USE_NETWORK=1’ statement, and build for Mess. And just as it looks like something out of SIMH, Mess makes use of libpcap which means that you are *unable* to send/receive on the host computer. (OS X & Win32 binaries). And of course you’ll need a ROM & Xenix diskettes.
Installing Xenix is pretty straightforward as long as you know your system key, and how to navigate the mess UI without rebooting mess or exiting by mistake (scrolllock on the PC, function/Delete on OS X).
First create a hard disk, and as always it should be ~500MB max.
Naturally you’ll need to setup the CMOS, for your memory size, and the hard disk. The BIOS I’m using didn’t autodetect the IDE drive, but it doesn’t matter as I know it’s characteristics as I created it.
From there Xenix was a pretty straight forward deal. Mess has good floppy drive emulation so it just worked. Adding TCP-IP was just as involved, and all went well. When it came time to install TCP & the network driver, remember to use thinnet, as the thicknet transceiver isn’t connected (as it would seem). The 3c503 is softset, so I went with IRQ 5, port 0x300, and thinnet, and it works fine for me!
Remember you will not be able to attach to it from your computer. Instead you must attach from another computer.
Also MESS tries to emulate true to hardware so it’ll be just as slow on MESS as it was on the real hardware. I suppose you could go with the at386 driver, but yeah it’ll be slow. The current at586 driver has issues booting from the hard disk, and I didn’t mess with it too much as Xenix is known to have issues with some Pentium systems.
Although I think the next place for adventure is the emulated Adaptec 1542CF.
Jason’s note on hybrid bootable ISO reminded me of a recent discovery. I have a bootable USB pen drive that I wanted to boot in VMware Workstation. Normally impossible, but there always is a work around! Turns out the problem is with the VMware built-in BIOS and more specifically lack of USB boot support. All you have to do is get a bootable media, floppy or CDROM with a boot loader that can redirect you to the USB device. I’m using Plop. Important thing to remember is to connect the USB pen drive to the virtual machine in a pass through mode. Also it’s very very slow.