Building DOS gaming PCs

The ultimate DOS gaming PC is a topic that I’ve seen come up in forums frequently, and that I’ve been asked directly a number of times. I guess since I published advice on running DOS games on Windows PCs on two continents, people figured I knew something about that. I guess I fooled them!

The trouble is that no single PC can really be the “ultimate” DOS game machine. Well, not if your goal is to be able to optimally run everything from early 1980s titles designed for the original IBM PC up to the last DOS version of Quake. I learned that the hard way in 1995 or 1996, even before Quake existed. Read more

The forgotten computer that changed the world

A rather hastily written and sloppily edited piece showed up on Slashdot yesterday morning that caught my attention, because it was about the Amiga 2000. The Amiga 2000 is a dear machine to me; in 1991, our family upgraded to one from a Commodore 128. I still have both machines, and there isn’t much that I know today that I didn’t first experience on one of those two machines.

This is an Amiga 2000 that looks fairly pristine. Inside there was lots of room for hard drives, memory, CPU upgrades, and video devices.
This is an Amiga 2000 that looks fairly pristine. Inside there was lots of room for hard drives, memory, CPU upgrades, and video devices.

Although I think the piece was little more than a used computer store’s effort to unload some hard-to-move inventory, I do agree with the premise. For a machine that had a tremendous impact on the world as we know it today, the A2000 is criminally unknown. Read more

Commodore 128, top-12 dud? By what measure?

PC Magazine presented a list of 12 computer duds, and while I agree with most of them, my old friend the Commodore 128 makes an appearance. Commodore released several duds over the years, but calling the 128 one of them doesn’t seem fair.
Read more

Dinosaur hunting

Today I slipped over to Laclede Computer Trading Company for the first time in many years. I was in search of an ISA parallel card. They’re not easy to find these days, mostly because they aren’t particularly useful to most people these days, but I figured if anyone would have one, it would be them.

No dice. But man, what memories.

Laclede has been around forever–at least 20 years, and probably a whole lot longer than that. I remember taking spare 286 and 386 stuff there in the early 1990s and they actually gave me money for it. Math coprocessors, Packard Bell power supplies, other oddball stuff like that. I’d salvage stuff from upgrade projects and get a little extra money that way.

Most of the stuff in the store now is Pentium 4-level. Recent enough to be useful, old enough to be really cheap. There wasn’t a single ISA board in sight. It was a little sad, but honestly, Clinton was probably still president the last time someone came in looking for something like that. No point in keeping that kind of stuff around.

I lingered around a while though. I saw lots of old SGI and Sun workstations. I remember in 1995, when I was taking a C programming class in college, we used to have to get on waiting lists to use one of the limited number of SGI workstations. They compiled code instantly, and unless you did something incredibly stupid, you weren’t going to crash them. They were a lot nicer than the NeXT workstations we usually ended up having to use when we got tired of waiting in line.

Those systems cost more than a decent car in those days. Each. And now, depending on configuration, you can get one for $30, $60, or $80. Incredible. They’re a lot more useful than the Pentium 75 I had back then, but PCs eventually overtook those weird and wonderful and odd proprietary Unix architectures.

I left, wistfully, but as I got in the car, I spied something. I wasn’t sure that distinctive shape sitting on a distant shelf was what I thought it was, but what else could it be? So I went back in. The clerk gave me a knowing look.

Yep, it was what I thought it was. There, on a tall shelf, on top. 1977 called. They want their computer back.

There it was. The Commodore PET 2001. The early one, with the built-in cassette recorder and the calculator-style chiclet keypad that was even worse than the IBM PCjr.

I’m pretty sure it wasn’t for sale. I didn’t ask, because I couldn’t afford it, and don’t have room for it. I stood there for a minute, studying it, then looked around some more. They also had a TI-99/4A, a contender from the early 1980s that couldn’t compete with Commodore, but some of its technology ended up in the Colecovision and, if I’m not mistaken, the IBM PCjr and Tandy 1000. It wasn’t a bad system, but it was horrendously overpriced. It cost more than a Commodore 64 but its capabilities were somewhere between a C-64 and a cheap VIC-20.

They also had a Commodore PC-10-III, which was one of Commodore’s PC/XT clones. And, next to the PC-10, there was a Radio Shack TRS-80 Model 1, the other forgotten personal computer from 1977.

Neat stuff. I don’t really have the interest to collect these old machines myself, but I’ll stop to admire someone else’s every chance I get.

We\’ll have to wait longer for PCI RAMdisks

In case nobody noticed, it’s August. July came and went, and there’s no Gigabyte I-RAM on the market yet.

But there are a few benchmarks out there, and Anandtech has an article that, once you get past the usual rambling and over-the-top introduction, has some useful insights.I was going to say the first problem is the somewhat disappointing speed, but actually, there are two bigger problems:

Availability. Now they’re saying it’ll be out sometime in August. And they’re initially only going to make 1,000 of them.

Price. The original $50 MSRP is out the window; now this thing is going to cost $150.

Can anything else be wrong? Unfortunately, yes. The speed is a bit disappointing. The SATA interface is the bottleneck. The very newest hard drives can come close to saturating the SATA interface for short periods of time, so the RAMdisk doesn’t outperform it by much. If this drive were using an interface with more bandwidth, there wouldn’t be as much problem, but squeezing more bandwidth out of the 33 MHz PCI bus is tough. We’re at the point now where the PCI bus is a much bigger bottleneck than the ISA bus was in 1994. The theoretical limit of the PCI bus is 132 megabytes per second, which isn’t much higher than the sustained throughput of 100 megabytes per second that the I-RAM delivers.

The combination of PCI Express and a faster disk protocol has the potential to resolve this issue, but at the expense of limiting the device’s market even further.

I’m disappointed by the review in a couple of regards, though. First, they compare the I-RAM to the fastest SATA drive available at the time of the review. That’s not necessarily what every would-be purchaser would be using. I believe that an I-RAM used to replace (or in conjunction with) a drive that’s a couple of years old would be a mind-blowing upgrade.

Second, they don’t take fragmentation into account. Enthusiasts are more likely to defragment their hard drives twice a day than everyone else, so fragmentation may not be an issue for them. But my wife, mother, and mother-in-law don’t know what fragmentation is. Well, maybe my wife does because she’s probably overheard me talk about it. The thing about the I-RAM is that it makes seek times irrelevant, so it’s never going to slow down due to fragmentation. Translation: For people who have lives, this thing could be phenomenal.

The review complained constantly about the drive’s capacity. So I’m disappointed that they didn’t test the drive with NTFS compression enabled. While data compression is still taboo, and it increases CPU usage, when you’re out of room it’s your only choice. While its effectiveness is unpredictable, it’s fairly safe to bet compression will get you another gigabyte or two of usable space on a 4-gig model. But just as importantly, under some circumstances, compression can actually increase performance. I want to know if increasing the amount of data you’re flowing over the saturated bus makes up for the increased CPU usage.

So there is a benefit to running Windows Server 2003 and XP

One of the reasons Windows Server 2003 and XP haven’t caught on in corporate network environments is that Microsoft has yet to demonstrate any real benefit to either one of them over Windows 2000.

Believe it or not, there actually is one benefit. It may or may not be worth the cost of upgrading, but if you’re buying licenses now and installing 2000, this information might convince you it’s worth it to install the current versions instead.The benefit: NTFS compression.

Hang on there Dave, I hear you saying. NTFS compression has been around since 1994, and hard drives are bigger and cheaper now than ever before. So why do I want to mess around with risky data compression?

Well, data compression isn’t fundamentally risky–this site uses data compression, and I’ve got the server logs that prove it works just fine–it just got a bad rap in the early 90s when Microsoft released the disastrous Doublespace with DOS 6.0. And when your I/O bus is slow and your CPU is really fast, data compression actually speeds things up, as people who installed DR DOS on their 386DX-40s with a pokey 8 MHz ISA bus found out in 1991.

So, here’s the rub with NTFS compression when it’s used on Windows Server 2003 with XP clients: the data is transferred from the server to the clients in compressed form.

If budget cuts still have you saddled with a 100 Mb or, worse yet, a 10 Mb network, that data compression will speed things up mightily. It won’t help you move jpegs around your network any faster, but Word and Excel documents sure will zoom around a lot quicker, because those types of documents pack down mightily.

The faster the computers are on both ends, the better this works. But if the server has one or more multi-GHz CPUs, you won’t slow down disk writes a lot. And you can use this strategically. Don’t compress the shares belonging to your graphic artists and web developers, for instance. Their stuff tends not to compress, and if any of them are using Macintoshes, the server will have to decompress it to send it to the Macs anyway.

But for shares that are primarily made up of files created by MS Office, compress away and enjoy your newfound network speed.

The first PC I ever built

I’ve noticed a disturbing trend lately: Everyone who built his own PC knows everything. Just ask him.
Now, don’t get me wrong: It’s admirable to build your own PC rather than just buying Dell’s special of the week (although some people would be better off just doing exactly that), and it does require at least skill with handling a screwdriver. But it’s not what it used to be. Today, building a PC makes you know something. It no longer makes you an expert.

The first PC I built was in the early 1990s. It was a 386. I did it as a personal favor; an organization I worked with had a 286 that blew its power supply, and I couldn’t locate a compatible replacement because of the machine’s semi-proprietary nature (it was almost a standard AT but not quite). So we bought a used AT case and a 386 motherboard, I disassembled the 286, and I started building.

Back then, a motherboard was a motherboard. It had a CPU, it had slots, and that was all that was guaranteed. Some motherboards didn’t even have their memory slots onboard. This one did, but that was only true about half the time. Plug and Play didn’t exist yet, unless you were working on an Amiga or a Macintosh. You had 15 IRQs to work with, several of which were taken up by the system, and no sharing was allowed.

The store mounted the motherboard in the case for me. It wasn’t a terribly difficult job, but since they didn’t charge me for it, why not let them save me some work? Besides, they didn’t want me connecting the power supply leads incorrectly and blowing up the motherboard, then returning the whole mess as “defective.” Yes, in those days if you got your power leads plugged in wrong–they weren’t keyed to prevent you–you would blow the board.

When I got the loot back home, I started off with the easy stuff. After I plugged in the memory, I pulled the serial/parallel board out of the 286 and transplanted it to the 386. Now I could hook up a mouse and a printer. Next was the video card. That, too, was easy. Those boards all had lots of jumpers and DIP switches, but they’d already been configured for me.

The disk controller was trickier. The 286’s controller was on the motherboard, which was very unusual for the time, and was what kept the 286’s motherboard from simply fitting into a new case–it was an odd size and shape, even though its keyboard port was in the right place and its slots were in the right place. So I had to set IRQs. I did that, then put the card in place, then transplanted the floppy drives over.

I fired the system up and inserted a DOS floppy. It did its memory test. Then I went into CMOS setup and told it the types of floppy drives it had. I exited and let the system boot. It booted. Excellent. I had a working computer, and I was just a step away from having a computer complete with a hard drive.

To that end, I plugged in an IDE controller. I jumpered it properly–or so I thought–and then transplanted the IDE drive over from the 286. It choked. No memory test. No nothing. I couldn’t even get into CMOS setup.

Backwards IDE cable, right? Nope. I checked that. I even tried a couple of different cables.

I struggled with the hard drive for hours and didn’t get anywhere. Finally I called the computer store. They couldn’t suggest anything I hadn’t told them I already tried, so I drove in and dropped off the mess. They struggled with it for a couple of days. Eventually they found the drive was one of the few 8-bit IDE drives made, and it was compatible with a limited number of controllers, and those controllers only worked in some 286s. They took the drive as a trade-in, and they backed up the data to tape, then restored the tape to the replacement drive for me. I took the drive home, plugged it in, and fired up the machine.

This time, it did the memory test. Excellent. I went into CMOS setup, and fed it the parameters for the drive–no IDE autodetect here, boys. I exited, and the system restarted, and booted off the drive.

It was definitely a rite of passage. I started on a Thursday afternoon, expecting to have the system running by Thursday evening. By the end, it was Tuesday evening when it was running. I did everything right, but an unanticipated problem–one that even stumped the experts–hung me up.

It humbled me. At the end I knew a lot. I knew I knew a lot. But I also knew there was a world of not-so-obvious problems and tricky solutions out there, and that I didn’t know everything. There certainly were people out there who knew a whole lot more than me.

It was a while before I tried building a PC again. But for a year or so in the late ’90s, it was how I made my living. I know I built more than 100 in 1997-98. And I saw some weird stuff then too, but never anything as weird as my first. A lot changed in those few years, and a lot more has changed in the five years since.

Today, you can buy an ATX motherboard with everything–not just the serial and parallel ports and disk controllers, but even sound, video, and Ethernet–integrated onto one board. Sometimes “building a computer” is literally nothing more than mounting a motherboard, hard drive and CD/DVD drive in a case, then running power to the motherboard and a power cable to each drive, then an IDE cable to each drive, slapping in a stick of memory, and closing the case up. Even if you go for discrete cards for video and sound, it adds less than five minutes to the process. The hard part is busting out the back cover from the card’s slot and finding the right screw to hold it in place. Any flunkie can plug the card in and expect it to work.

The people who soldered together computers from kits in the 1970s would say that what I did in the early ’90s wasn’t building a computer. What they did took a lot more time and patience and skill to accomplish. When they finished, they had a lot more understanding of what went into the computer and how it worked than I did. Who am I to disagree? Working with completed ISA boards that only need a couple of DIP switches flipped separated me a lot from the heart of the machine. But the computers of today move you yet another order of magnitude away from the internals of the machine. Just as it’s possible today to pump gas into your car without knowing anything about how a car works or how to fix it if something breaks, it’s possible today to build a PC without knowing anything about how a computer works or how to fix it if something breaks.

You just throw it away and buy a new one, right?

Right, buddy.

Rediscovering OS/2

So I picked up a surplus computer from work this week. Honestly, I bought it more because it was cheap than because I needed it. But it was a giveaway price for a good-quality system. Micron’s Client Pro line (its business-class line) is as well-built a PC as I’ve ever seen. The machine didn’t come as advertised, but it was still a good price for what I got: a 266 MHz Pentium II, 64 MB of RAM, a 4-gig Maxtor hard drive, a Lite-On CD-ROM drive of unspecified speed (it seems to be at least 24X), an Intel 10/100 PCI NIC, Nvidia Riva-based AGP video, an ISA Sound Blaster, and an ISA US Robotics 56K faxmodem.
Of course my first thought was to put Linux on it. But I have better machines already running Linux, so what’s the point, really? Then a few things sent me hurtling down the roads of my oldschool retro computing past, and a thought hit me: OS/2!

What I consider my first real job involved installing OS/2 literally a couple hundred times. That was version 3, on 50 MHz 486s. But by the time a Pentium-166 was a hot machine, I wasn’t using OS/2 much anymore. I realized I’ve never really seen OS/2 on something as hot as this P2-266 before. And I used to know how to optimize the living daylights out of OS/2, so this could turn into the best computer I’ve ever owned.

I had to patch my OS/2 v4 installation disk 1 to deal with the drive in the machine (download IDEDASD.EXE and unzip it, then follow the instructions in the README file) but once I got that going, installation was smooth. I need to track down device drivers for the NIC and video card yet. But I got a basic system up and running in about 35 minutes. That’s not bad.

I can’t wait to see Mozilla Firebird on this thing.

A day of catching up

I might finally have reliable DSL. Gatermann and I spent a good part of the day cleaning up my phone wiring. The wiring appeared to have been done by someone who couldn’t make up his mind how he wanted to do it. Seeing as I had two jacks that didn’t work anyway, and I own exactly three telephones plus an answering machine, we pulled out a number of the runs altogether (the wires are still there, just not hooked up at the box). And we cleaned up some oxidation that had shown up on some of the lines that were there.
My DSL connection does seem to be more reliable as a result. We’ll see in time how it turns out, but I know the brief storm we had tonight would normally knock me off the ‘net, and I haven’t fallen off yet since we did the work.

We also rebuilt a system. I’ve been intending to rebuild this one for some time (I pulled the case out of storage months ago) but never got around to it. Anymore, it seems like it’s a lot more fun to mess with other people’s computer projects than with my own. Anyway, we pulled out the system that served up this web site up until about a year or so ago (a Celeron on one of the last of the AT motherboards, a socket 370 job from Soyo), removed it from the old Micron case I’d put it in, and we put it in a monster server case, a former Everex 486/33. It’s a really good-looking case–battleship gray with black drives. And it’s built like a battleship too–very heavy gauge steel. It was pretty funny when we pulled out the full AT motherboard that had been in there and installed the Soyo, which is even smaller than what we used to call baby AT. We installed my CD-RW and DVD-ROM drives and a few other bits and pieces, and… an ISA video card. Yes, I’m sick. I was out of PCI slots and I loaned the AGP video card for the system (a Radeon 7000) to Steve last week and won’t be able to meet up with him to get it back until Wednesday at the earliest. I am half tempted to go ISA for either the sound or network card for the time being in order to free a PCI slot for an Nvidia Riva 128 card I have kicking around. It would be a big improvement. The screen writes remind me of BBSing; the text comes onto the screen at a rate somewhere between what I remember 300 bps and 1200 bps looking like.

But then again, what I want this system for (primarily) is to do things like burn CDs, and I don’t need superfast video for that. And I don’t know that I’m going to be burning anything between now and then.

Yes, I know, catch-up days are terribly exciting to read about.

But somewhere around here I think I have some stuff I wrote last week and never posted. I’ll have to see if I can find it to post tomorrow.

New Freesco

Freesco 0.3 is out. It now has working PPPoE support, I understand. I don’t know what else it does just yet. But I intend to find out.
Freesco is easily my favorite single-floppy Linux distro, because once you get the hardware going, it’s easy to get running. Anyone familiar with computer networking can do it, without knowing a thing about Unix. And once it’s working, you can move it to the hard drive, which is good, since tiny hard drives are common as dirt and cheap and reliable (Freesco spins the drive down after it’s done booting, so a hard drive should work pretty much indefinitely, seeing as you’ll only reboot the thing when there’s a power failure), whereas floppy disks are anything but reliable.

That’s not to say that getting the hardware going isn’t a pain sometimes, but that’s not Freesco’s fault. Resolving a bunch of IRQ and I/O conflicts to get a 486 with a pile of ISA cards in it working perfectly is a pain no matter what OS you intend to run on it.