Then
and now


The 8-bit Atari computer.


by Matthew J. W Ratcliff

    Back in 1984, Atari's 800XL was something of a wonder. Unlike the 1200XLwhich was unanimously rejected by all developers-the 800XL had quite a bit to offer over the "Old Faithful" Atari 800 model.
    First, the 800XL was far more compact. The entire computer was built on a single circuit board, as opposed to the ten boards of the Atari 800. Fewer parts generally indicate a lower price and higher reliablity for any product. The original 16K Atari 800 sold for about $800, while the newer 800XLs were going for under $200-with a full 64K of memory.
    The extra RAM of the 800XL was difficult to get at, and couldn't be used as part of BASIC memory. However, it did set a "standard," since the extra RAM and the "bank switching" technology were produced by Atari. There were extra memory boards for the 800, but none were ever produced by Atari, thus, there were no standards. The boards were expensive and had little or no software support.
    It wasn't long before OSS came out with a version of DOS XL that would remain resident in memory at all times, and still give you about 4K more RAM to play with while programming in BASIC. This neat trick was done by hiding a large portion of DOS XL "under" the operating system ROMs in an extra bank of RAM. SpartaDOS, which went unnoticed for some time on the older 800, started using similar techinques. It even gave users an 8K (62-sector) RAMdisk that was called RDBASIC.COM.

    During this same period, Atari released their 1050 disk drives with the "new and improved DOS 3." I've never understood why Atari placed a true double-density "mechanism" (drive chassis) in the 1050, but only had enough electronics to support "enhanced density." The result is a wasted 50K of storage on every disk!
    Fortunately, ICD wasn't long in developing the US (Ultra Speed) Doubler. Not only did it give your drive the extra memory to handle full double-density sectors, but it improved the speed of the drive significantly when using SpartaDOS in the US sector skew mode. More recently, Duplicating Technologies has come out with a high-speed density doubler of their own, though it looks as if it's simply a clone of the one produced by ICD.
    While ICD and OSS were being innovative, Atari gave new users a DOS that was completely incompatible with anything ever produced for the 8-bit Atari. Instead of 128-byte sectors, DOS 3 gave us 1000byte blocks. This resulted in a total of 128 blocks on an enhanced-density DOS 3 disk, or about 128K of storage.
    My speculation when it was released, in 1984, was that this DOS used a single byte sector pointer. Any sector reference less than 128 would refer to side one of the disk, and those 128 and greater to side two. "But Atari never released a double-sided version of the 1050 drive," you say. True, but they had planned a complete console system, the 1450XLD. It was even listed in their new XL product catalog. It was to have one or two double-sided drives, built-in modem and speech synthesizer, plus many other goodies. Unfortunately, this unit never made it to market.
    Bill Wilkinson of OSS published "fixes" for Atari DOS 2.0S to make it run in enhanced-density mode (giving you 128K per disk), while maintaining compatibility with the older, more universal format. Not long after that, Atari came out with DOS 2.5, which gave you single density, or optional enhanced-density support, a RAMdisk for the new 128K Atari XE machine, and a utility to convert from DOS 3 to DOS 2.5 format. DOS 3 was dropped by Atari completely.
    What ever happened to Atari DOS 3? Recently, I saw a bargain on floppy disks in a B & C Computer Visions Flier. The ad read, "Diskettes as low as 20 cents, 1000 for $200." It sounded too good to be true. (The lowest price I've ever gotten is 27 cents per disk). The fine print below revealed that these disks were "unnoticed with DOS 3"!

    One of the nicest features of the 800XL over its predecessors was that it had BASIC built in, though it could be disabled by pressing the OPTION key at boot time, making 8K more RAM available to your assembly language programs.
    I'm sure most of you "old timers" remember the infamous keyboard lockup bug with the original Atari BASIC, now commonly referred to as Revision A (or simply REV A). The bug was associated with editing. Whenever you deleted program lines, or parts of lines, resulting in a code size reduction that was a multiple of 256 bytes, the computer locked up. Bill Wilkinson diagnosed the problem early on and provided Atari with a fix several times. Atari never made another "mask" for the BASIC ROMs, however, until the 800XL was developed and REV B Atari BASIC came about.
    Apparently, several other fixes were implemented, which just caused more problems.
    Whoever programmed the original BASIC fix, must have looked at the code used for expanding the program when inserting lines of BASIC code. Since the code looked the same as the buggy delete code that had just been fixed, it was applied there as well -a big mistake. The INSERT code always worked fine-the DELETE code had the problem. So the keyboard lockup didrit go away; it just moved. Since you delete lines of BASIC code more often than you insert them, this particular bug doesn't bite as often in REV B.
    The result of these extra fixes was a major headache for over 100,000 800XL buyers. The most common problem was that every time you saved a BASIC program from this REV B BASIC, your file expanded by 16 bytes. This "growth" was cumulative. If you loaded and saved the same file enough times, you'd run out of memory.
    There is a simple way around this problem: just LIST then ENTER your files. But sometimes, after entering a large BASIC file, you get an ERROR 9 (string not dimensioned), on the very line where the DIM statement occurs! Usually you can recover by "editing" the program (anything to change its length), or by doing a SAVE and LOAD.
    These bugs were documented early in the production cycle of the 800XL computers; Atari was well aware of them. However, the machine was apparently manufactured for a year or more with the defective REV B BASIC, even though the REV C was completed shortly after the REV B bugs were revealed. Apparently, Atari had a large stock of the REV B ROMs and didn't feel the bugs were significant enough to discard them in favor of REV C. It was about this time that Atari was in dire financial straits-the year it lost nearly one-half billion dollars.
    Atari began selling the REV C cartridge to consumers for only $15. It was helpful, but didn't seem fair. I did some PEEKing around one day and found that Atari BASIC REV B and REV C are virtually identical, except for 12 bytes!
    Do you have an 800XL? Do you have revision B BASIC? To find out, try this:
The current REV C returns a 234, while REV B has a 96 at this location. If your machine returns the 96, then by all means order the REV C cartridge from Atari.

    The only computer I could afford when I started in this business was the Atari 400. Its membrane keyboard drove me nuts, so I soon purchased a "replacement" keyboard for the Atari 800 and interfaced it with the 400. The keyboards for the Atari 800 are the best Atari has produced to date. I even prefer them over the "best" keyboards made for the IBM PC and compatibles. The B-KEY 400 membrane keyboard replacement for the Atari 400 was popular for a year or two, but had its own problems. It didn't take long for the keytops to develop a lot of friction from accumulated dirt, so they required regular cleaning.
    When the 800XL came out, there were apparently two different commonly used keyboards. One had a nice, smooth feel; the other had very wobbly keytops and a great deal of friction in the key mechanisms. It could slow down a good typist by 10 to 20 words per minute.
    During the height of Atari's game production years, I started getting calls from friends about 800 and 800XL keyboard problems, especially with the space bar. The first question I always asked was, "Have you been playing Defender?" Most of the time I'd get a surprised reply: "Yes! How did you know?" In Defender, you launch your "smart bombs" with the space bar. Normally, this is a last resort strategy. In the frenzy of a hot shootout, it's easy to bang the poor space bar to death.
    Atari will be rereleasing Defender later this year for the new XE Game System. The new version will accept either the space bar or one of the console keys to launch a smart bomb. If you get this version, use the console key-it's difficult to pound it as hard as the space bar.
    With the release of the XE computer line came a new keyboard of dubious quality. The tactile feedback was less than desirable, with a very mushy feel. The bases of the keys were made of a conductive coated rubber, and seemed to have a pretty high failure rate with heavy use. Keyboards are not repaired at service centers; they're replaced with new ones -when they're available. Not all XE keyboards are constructed exactly like this, however. It appears that several versions (all with about the same tactile feel) for the XE series are in use.
    Early on, I noticed the 800XL power supply ran awfully hot. It was in a solid black case with no ventilation, and was quite hefty. Mine burned up after just two months of use. I promptly built a new one, and tore the old one apart. That was no small task, since the entire transformer and circuitry were encased in a solid plastic resin, a good electrical insulator and an excellent thermal insulator.
    The efficiency with which a power supply can shed heat generated from regulating the output voltage largely determines the longevity of the supply. My local dealer always keeps spare 800XL power supplies in stock. Atari has gone through two or three versions of XL/XE power supplies to date. The latest is in a larger case, well ventilated, capable of providing 5 volts at 1.5 amps, and 50 percent more power capacity than the original. This was done to support the extra RAM of the 130XE, and is adequate for all the RAM upgrades available for the XL/XE machines.

    Shortly after Atari came out with the 800XL, they released the translator disk. Several other translators have been making the rounds over the past two years as well. We should never have needed them, however. Here's why:
    In the early days of the Atari, some hacker types discovered two "illegal" entry points in the system. Called EOUTCH and EGETCH in Mapping the Atari, these illegal entry points would print a single character to the screen and get a single character from the keyboard. These locations never changed throughout the life of the 800/400 systems, or during two revisions of the operating system. Because the illegal entry points were documented in the magazine, many people assumed this was a "safe place" to do a quick and dirty screen write and keyboard read.
    Unfortunately, when the 800XL came out, these locations moved. Everyone pointed the finger at Atari, and blamed them for coming out with a new, "incompatible" computer. But, as anyone who's read Atari's Technical Reference guide for the 8-bit knows, these illegal entry points were never documented by Atari. Bill Wilkinson's "Insight Atari," column in Compute! showed users how to perform these functions "legally" with calls to the CIO (Central Input/Output) utility in the operating system. With the CIO and the proper setup code, you can legally perform I/O with any device on the Atari, including screen, keyboard, cassette, printer and disk.
    It wasn't Atari's fault that a lot of public domain and some good copyright covered software (written by irresponsible programmers) wouldn't run on the new XL and XE machines. Atari helped bridge the gap with the translator disk (at a $10 charge), until the old code was replaced with newer and better programs.
    Setting up a CIO call to do the exact equivalent of the EOUTCH and EGETCH routines only takes six lines of assembly code each. It's unfortunate that, around the time the XLs came out, Compute! brought out "Assembly Language for Beginners" using the illegal calls!

    One of the "improvements" in the 800XL was a little card edge connector called the parallel bus. Actually, there already was a parallel bus connector in all the 400/800 computers produced. It was used to interface the mother boards to a mainframe computer system for "burn in" testing during the manufacturing process. Once the 400/800 computers were assembled, this connector was hidden underneath the ¼-inch aluminum RFI shield.
    The connector on the 800XL was special, however. It provided some additional control lines for hooking parallel devices to the computer. They could be DMA (Direct Memory Access) for transferring up to 100K of information per second. That's mind boggling, when you consider that the Atari 800XL has only 64K to begin with.
    I had hoped Atari would come out with a hard drive interface or 80-column adapter card to make use of this connector. Many thought it would never happen, and discarded their new XL OS ROMs in favor of some "upgrades" that made the 800XL compatible with the old 800 system. They ignored the fact that eliminating the XL OS was throwing out all support for the Parallel Bus Interface (PBI).
    Well, a couple of guys at ICD were making a little money off the ever-growing popularity of SpartaDOS. But they wanted to do something different, exciting - something that could be called the "ultimate" for the 8-bit. After a lot of talk and design work, they created the MIO (Multiple Input/Output) board. For starters, this box uses the Atari parallel bus to connect your 800XL or 130XE to 1 meg of memory. This 1 meg of RAM can be configured as printer buffer and/or RAMdisks -extremely fast RAMdisks. The box also gives you an RS232 interface that's completely compatible with all software written for the R1: port of the Atari 850 interface.
    While they were at it, they stuck a printer interface in there, too. Then, as if that wasn't spectacular enough, they added a hard drive interface connector with enough room to add an 80-column board later.
    What innovations has Atari brought us? Well, the XEP80, an 80-column display board, is now beginning to make its way to reviewers' hands. It uses the Atari joystick port, a "bit-banging" serial interface. This configuration is slow, and probably not as reliable as the parallel bus.
    Why didn't Atari use the parallel bus on its 800XL and 130XE computers for this product? Was it to give us continued support of the old 400/800 computers, which are now outnumbered by XL/XEs by approximately 8 to 1? Not at all. I think Atari really wanted to use the parallel bus on the XL/XEs, and begin to phase out support for the 400/800 machines, but they couldn't. You see, Atari "forgot" to put the parallel bus connector on the 65XE. Since this is a "current production" model, Atari was obligated to support it.

    The Atari 130XE is a serious competitor for the ever-popular Commodore 128. In this machine, Atari has expanded memory to 128K. Since it's an "official" configuration (e.g., it comes from the Atari factory with 128K), it's supported by PaperClip 2, with an integral spelling checker. The Syn series has all been updated to give you more workspace for databases and spreadsheets, and-for general purpose use-all the popular DOSS for the Atari have been updated with additional RAMdisk support.
    There could be even more, however. There are several popular RAM upgrades for the 800XL which take it up to a full 256K of memory. This upgrade is so popular, in fact, that PaperClip and a few other commercial applications support it-even though it's not an official version of Atari's 8-bit line. These upgrades are generally compatible with the 130XE bank switching technology, but provide more
    Now that the Atari 130XE is so popular, why aren't games like Trinity (which uses Infocom's "interactive fiction plus" parsing technology) ported over to the 130XE? They run on the Commodore 128, so you'd expect to see them for the 130XE as well.
    The problem is the "standard" disk technology. True double-density disk drives have been produced by Commodore for some time, and the newer game software requires all the disk storage it can get. Until Atari releases its own true double-density disk drive, additional support for the extra RAM of the 130XE will be limited by the disk drive itself.

    This brings me to the latest "new technology" for the 8-bits: the 3½-inch drives we've been hearing about. Atari is one of the largest purchasers of 3½-inch drive mechanisms in the world, in support of the ST product line. In an admirable effort to bring you more "power without the price," Atari announced the development of a new 3½-inch disk drive for the 8-bit line. We 8-bit owners get more storage on a smaller disk (340K), and ST buyers get a better price because Atari is buying more 3½-inch disks.
    The latest word from some of the Atari Fests is that this product has been dropped, before it ever got beyond the "vaporware" stage. It's never been unveiled at the computer shows, or photographed for the magazines, so where is it? It does seem to have been dropped.
    Why would Atari drop it? Maybe it's because developers would have to ship both 5¼- and 3½-inch format disks to reach the greatest market share. I've heard that developers are not pleased with the prospect of the extra cost of producing software on "double disks," considering the low profit margin and high piracy factor they face at the outset. Many developers would completely drop their support of the Atari 8-bit before going that route.

    The 5¼-inch disk standard is universal. The 5¼-inch disks are much cheaper than the 3½-inch floppies, which is important to those with the less expensive 8-bit machines.
    The extra RAM power of the 130XE is limited by its standard peripherals. It's time for a new disk drive, another 5¼-inch drive. It should be double-sided, double-density and a full 360K. These drives have been used in the IBM PCs for years now. Since the AT (which uses 1.2-meg floppies) has gained in popularity-and gotten cheaper on the clone market -the PC XT clones and associated hardware are glutting the market. This includes those 360K drives, which are commonly advertised in trade journals for $60 or less. Better yet, how about putting an AT drive mechanism on the 8-bit Atari, with a full 1.2 meg of storage per floppy disk? These mechanisms are now selling for about $100.
    I'm happy to report the latest word is that Atari has in fact dropped the 3½-inch 340K drive in favor of the XF551, a 360K 5¼-inch drive mechanism, which was shown at the June CES. This should be inexpensive for Atari to implement, in comparison to the 3½-inch format, since Atari already has 5¼-inch cases, and has developed double-sided 5¼-inch drive control technology in the past (for the 1450XLD). So we should be seeing a full 360K drive for the XE line soon.
    The new and improved, powerful ADOS under development for the 3½-inch drives will run just as well on the 5¼-inch units. The drive will run on an older 400/800 system, but I hear the DOS may be XL/XE compatible only (requiring additional computer memory banks to run).
    If you're hanging on to your old faithful Atari 800, you may need to go to SpartaDOS 1.1 to get full use out of one of the new 360K 5¼-inch drives from Atari. However, I can't be sure until the new drive and ADDS are released.
    The 800XL continues to stand above the 65XE, particularly among loyal Atarians. The 256K RAM upgrade for the 800XL makes it a more powerful machine than the 130XE. The 130XE has been upgraded to 512K by hackers, but isn't nearly as simple to modify as the 800XL. Atarians are ready for more RAM, and the prices are right. The 130XE should become the "low end machine," with a new 256K or 512K XE to top off the line. The 65XE should be dropped, so that the entire current 8-bit line of machines have the parallel bus connector. Then maybe we'd see some really innovative hardware from Atari to place on this parallel bus.
    It looks as if Atari wants to keep producing the current line, however, and leave the upgrades to ICD and talented hackers. If any more technology is added to the XEs, then a complete bundled system with drive and monitor will approach the base price of the 520STs.
    Considering all the custom chip technology they've produced over the past few years, Atari appears to be serious about hardware. I've heard that they recently purchased a small chip manufacturing company in the Silicon Valley, but haven't been able to confirm it. Apparently, Atari wants to produce their own ST blitter chips, since schedules and yields have been less than desirable with outside firms. If it goes well, Atari may be more responsive in updating ROMs in the 8-bit and ST computers. Maybe we'll even see some new game programs on cartridge once again. After seeing the new XE Game System at CES, it looks as if we'll be seeing a lot of new software in cartridge format, with the new super bank switch technology.
    Modems are getting cheaper all the time. A new 2400-baud modem chip set will bring 2400-baud modems into the realm of affordability (under $200) within a year or two-just as Atari is finally preparing to release their SX212, making this project another that may not make it to market.
    One of the most interesting ideas never brought to fruition by Atari was the integral speech synthesizer for the 1450XLD. If speech were part of the hardware of the machine, it would be a standard part of the user interface in application, educational and game software for the 8-bits, for years to come.
    The Alien Voice box was quite popular, considering its relatively high price. S.A.M. and other software-driven synthesizers were nice, but too memory hungry-and less than ideal when it came to sound reproduction quality, because of the 60Hz vertical blank interrupt. Turning off the VBI interrupt improves sound quality, but also shuts off the screen display. While little beeps, bops and boops are the norm for sound effects in IBM and Apple game programs, Atari has been way out in front for years with its four powerful sound registers.
    The hype and hoopla has centered around the ST for about two years now. We've heard enough about blitters, math coprocessors and desktop publishing. The 8-bits are serious machines. The Apple II series is alive and well because of the innovations and continued compatibility Apple has provided for the II GS. If Atari rests on its laurels now, we're going to see many more 8-bits at garage sales-while a few of us 8-bit diehards continue to support technology (software and hardware) companies like COVOX, ICD, Epyx and MicroProse. With the introduction of the XE Game System, we should see more cartridge-based games and less piracy on the 8-bit Ataris. Let's hope that this generates a resurgence of interest in the Atari 8-bit product line, so the XE computer systems will continue to be enhanced by Atari and third-party developers.