Computing at Chaos Manor

Mac Reconsidered

Jerry Pournelle
BYTE Magazine

August 1984

I have a problem with the Macintosh: I don't recommend that you get one now—but now is May 1984, and by the time you read this, things will have changed drastically. I don't know what I recommend "now," meaning as you read this column. It all depends on what's available.

As I write this, the Macintosh is a bargain only if you can get it at the heavily discounted price offered to faculty and students of the favored 24 universities in the Macintosh "consortium:" (Incidentally, I'm told that when you see a list of the universities within the consortium, you'll see only 23; one university made anonymity a condition of joining, and Apple can never use that university's name in any press releases or advertising.)

There are several reasons why you wouldn't want the Mac at full price just now. Some are inherent in the machine: negative features that must be overcome with something positive. The other reasons may well be temporary: they're fixable, but only if the Macintosh survives, and despite all the Machype, I'm not as convinced of that as everyone else seems to be.

First, the "permanent" problems. The chief one is that the Macintosh is a rather ill-designed machine. The Caps-Lock key is where the Control key ought to be; there is no Escape key; the Control key is badly placed. There are no cursor arrow keys, meaning that touch-typists must remove their hands from the keyboard and use the mouse even for small cursor movements.

The mouse itself has only one button. This, we are told, is to prevent user confusion; one button is enough. However, to use the mouse you must often do Shift-click (hold down the Shift key as you press the single mouse button) or Command-click (ditto for the Command button and mouse). This makes a two-handed operation out of what should be a one-handed affair. It would have been easy enough to put copies of the Shift and Command keys on the mouse itself. Of course, that would have cost a bit more.

The single disk drive is a real pain. For the past six weeks, I've heard stories of a new Copy routine that would use more of the Macintosh's memory and thus require fewer disk swaps for copying a disk; but I have not seen the new routine, and as of May 15 there had been no firm date set for delivery. Similarly, every week someone announces that the second disk drive will be available next week, but so far none have been delivered.

The operating system has bugs. There are a lot of ways you can trash a disk and lose all your text files. This is inexcusable, and sufficient to prevent me from using the Macintosh as a word processor. My time and energy are too valuable to gamble with. I am told there are fixes to most of the disk trashings, but since none of them are documented, they are not available to "the rest of us.'

The Macintosh documents are nearly worthless. Every now and then the Macintosh will deliver messages about "Fatal Error #2" and the like, but that's interpreted nowhere. The Macdocuments are pretty, with lots of visual appeal, but they don't tell you enough about the machine. Neither, incidentally, does Doug Clapp's book Macintosh! Complete (Soft Talk). This $19.95 book is unreserved in its praise of the Mac but tells you little that isn't in the Macdocuments. It's often deliberately cryptic. A better book is Cary Lu's The Apple Macintosh Book (Microsoft), which costs less ($18.95). Though it is incomplete (any book published simultaneously with the announcement of the Mac would be), it does have a lot of useful information. However, neither book, nor the Macdocuments, will interpret the Mac error codes.

The Macintosh disk drives are painfully slow. It takes the better part of forever to load software. It's so difficult to make backup copies that one is tempted not to make them: this can result in unrelieved grief if you're then caught by one of the disk-trashing bugs.

There's nowhere near enough memory. As I write this, only two languages are available for the Macintosh: FORTH and Microsoft BASIC. I know little about the FORTH; but when you've loaded the Microsoft BASIC, you have only 14K bytes of work area left. That means you have to write cryptic BASIC programs or none at all. Shades of the old days with 4K-byte memory boards!

I now find you can use the CLEAR command to get up to 29K bytes of workspace. It's an undocumented feature.

The "desk calculator" feature within the Mac is useless. This is a pity, because I must keep a TI calculator next to my desk in order to do the myriad computations required in daily life, and I'd hoped to replace the TI with the Macintosh. No way, though: the Mac's calculator has no memory, no change-sign key, and, of course, no scientific functions or notation. Moreover, it's very hard to use. Mice are not the best way to input numbers. I presume that one could write a calculator routine in BASIC, but what's the point? Both it and BASIC would have to be loaded-at painfully slow speed. You'd be better off with a lap computer like the NEC PC-8201A, which has BASIC in ROM.

As I write this, there is no application software for the Macintosh. Microsoft Multiplan was recalled, as it should have been—the version we bought crashes at the slightest mistake, losing everything you've calculated. Each week I hear announcements of an incredible pile of new software to be delivered next week, but when the next week comes, none of it is available. I've given up believing the breathless new announcements. More: the spate of announcements, followed by nondelivery, makes me wonder whether interfacing software into the Mac's operating system is more difficult than Apple wants us all to believe. Perhaps I'm an excessive worrier. Perhaps not.

Some Love It

Theres a good side to the Macintosh; good enough that if Apple fixes the fixable defects, it may truly be "the computer for the rest of us."

First off, I don't know anyone—including me—who doesn't think it's fun to play with. MacPaint, with its spray can and paintbrush; MacWrite, with the ability to change type fonts and sizes; and the cut-and-paste utilities combine to let you create some remarkable graphics. The days of the illuminated manuscript have returned. It takes a long time to print a letter written with MacWrite, but it looks spectacular.

Second, for those who must write short illustrated documents, Mac is superb.

Third, the Mac attracts people who hate computers. A typical case: businessman gets Mac; finds he can't do much with it; takes it home; wants to experiment with it, but finds he can't get near it because his wife and children, who previously hated computers, have taken it over and will not let go. There is, apparently, something about mice and pull-down menus and icons that appeal to people previously intimidated by A > and the like. Experienced users may know that the Mac won't do very much, but beginners can make it do something, and that's important.

Experienced hackers tend to divide into two groups: a sizable number think the Macintosh is a ridiculously priced Etch-A-Sketch, but just about as many think it's wonderful. They don't do anything useful with it, mind you; but they cease not to explore, and they're in love with all the exciting new concepts and features. I think they're more in love with the idea of the Mac than with the actual machine, but that's irrelevant. The important fact is that knowledgeable computer wizards think the Macintosh is wonderful; they believe that it will soon have lots of excellent application software; and they passionately detest anyone who has doubts or misgivings about the Mac. This is, they say, the computer for "the rest of us'—even though they've long ceased to think of themselves as one of "the rest:'

The Macdilemma

For all the enthusiasm, the Macintosh is not really all that advanced. The Modula Computer Company offers an American version of Lilith with black-and-white graphics capabilities better than Macintosh; and it has Niklaus Wirth's Modula-2 operating system, complete with mouse, pull-down menus, scroll bars, and most of the other Macgoodies. At least two firms are developing new Modula-2-based bitmapped screen systems; one, the Sage, will use the 68000 chip, have a megabyte of memory plus two disk drives, and sell for about twice what Apple wants for the Macintosh.

Apple has decided to maximize profit rather than market share. That may not be a wise decision during this critical period. Part of the excitement of the Mac was that it truly was the "computer for the rest of us": which implied that "the rest of us" could afford it. Alas, that didn't happen. At $2500 plus tax for the basic machine, Macintosh is no bargain; and when you've bought it, you're not done.

The Mac needs an external disk-hard or a second microfloppy—and more memory. It needs, but can survive without, a better keyboard. You'll spend at least $750 for a second drive and memory update, and if Apple prices these items the way the Mac itself was priced, you'll spend twice that. Moreover, not all these items are available; and some hardware improvements will be developed only if a lot of Macs are sold.

No machine is useful without software. Apple has chosen to make "software developer" status a lot harder to get than was implied in presentations at the West Coast Computer Faire and other hobbyist/hacker meetings. To be eligible for the "developer" discount and support, you must already be in the software business—which means automatic exclusion of the innovative hobbyists who weren't interested in more traditional machines but were turned on by the Mac.

Then, too, the machine has a closed architecture, and the Macdocuments are incomplete. You can buy, for $150, Inside Mac, a pile of xeroxed loose-leaf fact sheets (including interpretation of the error codes), but even these aren't complete. At present there is no assembler, and when the assembler becomes available, you will need two Macs in order to run it with the debugger. All in all, Apple has told the hobbyists to drop dead. This may not be wise. TI tried it. Long live the TI-99.

So What Should We Do?

If the software does come flooding forth—after all, there are all these announcements, and there must be some substance behind them—then Macsales may soar. Hardware developers will come up with fixes for the major problems. This will generate even more sales and more software.

As that base market expands, other developers will get into the field. Last fall at COMDEX we saw nothing but IBM PC clones; if the Mac proves popular, we'll see Maccompatible machines everywhere; and since there's a considerable profit margin in the Macintosh price, competitors will be able to offer more features for the same money or, more likely, much the same features for a lot less.

So: at the moment there's no compelling reason to get a Macintosh. If you're in the mood to enthrall a recalcitrant spouse or child who now hates computers, perhaps it's worth the price; but you can get a lot more utility for what the Macintosh costs.

Things may be different later, but by then the clones will be appearing. Alexander Pope said, "Be not the first by whom the new are tried, Nor yet the last to lay the old aside." It's still good advice. Pournelle's Rule is: "If you're in the mainstream, the best business system is last year's state-of-the-art development system." Whether or not Macintosh is in the mainstream, it's still a development system. I'd wait.

Copyright 1984