Previous Issue | Search TidBITS | TidBITS Home Page | Next Issue

TidBITS Logo

TidBITS#395/01-Sep-97

With the clone licensing imbroglio likely to reach a head this week, we include more comments about clone licensing and about why TidBITS thinks Apple needs clones. In addition, Adam describes his latest book, one that takes a very different approach to the Internet than the Internet Starter Kits, and we publish Part 1 of shareware author Rick Holzgrafe's personal look at what it takes to create successful shareware.

Topics:

Copyright 1997 TidBITS Electronic Publishing. All rights reserved.
Information: <info@tidbits.com> Comments: <editors@tidbits.com>


This issue of TidBITS sponsored in part by:


MailBITS/01-Sep-97

Clones Stood Up by Apple for "Date" -- Last week Apple restricted the terms of its Mac OS Up-to-Date program, offering discounted OS 8 upgrade paths only to people who purchase Apple Macintosh computers. Previously, anyone who bought a machine that wasn't running the latest version of the OS (for instance, many models currently ship with Mac OS 7.6 preinstalled) could upgrade for as little as $9.95. Now, users who bought machines from clone vendors after 01-Aug-97 must contact the manufacturer to find out if there is a discounted upgrade plan (although Apple will apparently still honor orders postmarked 01-Aug-97 and earlier). [JLC]

<http://www.macos.apple.com/macos/releases/fulfillment.html>


A Case for Clones

by Adam C. Engst <ace@tidbits.com>

Based on rumors, we believe that the clone licensing fracas will come to a head this week, possibly by the time you read this article. For news junkies and those who believe that clone licensing is a key issue, during the next week we will be posting updates on our Web site should solid, verifiable information become available.

<http://www.tidbits.com/>

Last week's article about licensing issues between Apple and the clone manufacturers brought in more messages than any article in recent history. This week I want to share some of the responses to last week's article, and then offer a stronger case for why Apple must continue to license clones.

Two Heads Better than One? Back in TidBITS-372, I proposed Apple should split into different companies, with each free to do whatever was best for that company. I suggested a hardware company, an operating system company, and an applications and utilities company. Although my suggestions haven't come to pass, Apple has spun out the Newton division into Newton, Inc. (see TidBITS-381).

At that time, clone licensing wasn't at the forefront of the Apple soap opera. But, as several readers reminded me, the fact that Apple is a systems company ensures that issues like clone licensing will always be a source of internal conflict. Mel Martinez <mem@jhu.edu> commented:

Your comment, "Cloning both solves and creates problems for Apple..." points at the crux of the matter. Apple's current business model is fundamentally in conflict with the way the modern personal computer industry has evolved. You call Apple a "systems company" (as opposed to Microsoft being a software company). However, you cannot be a systems company and also try to be an open platform company that licenses its platform for sale by other systems vendors. Doing so creates inherent conflict within the company. Cloning is good for Apple's software division. Cloning is bad for Apple's hardware division. Thus Apple's two sides are in conflict, and it is always bad business to be in conflict with yourself. As one solution, consider separating Apple's hardware division from the software division so that their interests are independently realizable. Clones have shown that to the customer base, the hardware side of the business is replaceable. Apple's software is the true coin of the company business, current revenue distributions aside.

In addition, Rob Gvozden <gvozden@hk.super.net> commented:

As I see it, you have come close to the heart of the issue in your contrast of Apple and Microsoft in terms of their respective business activities. Publishing computer software and mass producing computer hardware are very different activities in terms of market structure. Publishing an operating system, thanks to some protection from patent laws, is pretty much a monopoly business insofar as competitors should only be able to offer less-than-perfect substitutes. Building a box with a CPU in it is more like "monopolistic competition," where Compaq and Dell are identifiably different computers but both can run the same software with only minor variances in performance due to the different system designs.
The implication in terms of pricing is clear: the monopolistic competitor is forced - by the competition - to reduce prices to a level that just covers the average cost of production. The monopolist, however, can look forward to juicy margins, because as long as he gets a handle on the size of the market for his product, he will be able to sell his product above its average cost due to the lack of competition.
When Apple started, it had a monopoly in both hardware and software: you couldn't run one without the other. Opening up the hardware side to clones means coming to terms with a shift in the company's business model, because the hardware business by definition will see a contraction in margins due to competition.
The implication of this is that Apple would do well to split itself into separate hardware and software businesses if it wants to survive. If Apple stays a single company, the antics of its hardware business put the entire operation in jeopardy. As two separate companies, the software side could survive unmolested and profitable, regardless of who made the hardware it ran on. The hardware side would have to learn how to compete under slimmer margins or else pass the way of the dodo.
I hope Apple resolves its identity crisis, because I think that this, and not squabbles over how best to grow the market for the Mac OS platform, is what will ultimately bring down the unified Apple we know today.

Those Who Do Not Learn... Speaking of Rhapsody for Intel machines, Karen Nakamura <karen@gpsy.com> raised an interesting point that hasn't received much attention. She noted that when NeXT ported NeXTSTEP to Intel chips, the market for NeXT's proprietary Motorola 680x0-based hardware dried up. Although NeXT made generally well-received and innovative hardware, the fact that it was proprietary and came from a single source ensured most people purchased cheap PC clones to run NeXTSTEP. Karen wrote:

It seems Apple is heading down this road as well - but deliberately shooting itself in the foot in the process. If Apple is serious about Rhapsody, then now is not the time to shut the door on clones and revert to a not-invented-here attitude. Clones are coming, but Apple's hardware division should be most worried about the Intel clones, not the Mac clones. I'm not optimistic about the future of Apple's hardware division, but the only way it can survive past Rhapsody is if it has sufficient market share (as a percentage of total workstations; not versus Mac clones) to ensure that developers will still support it.

The parallels are striking, but the primary difference is that Apple has a much larger installed base of Macs than NeXT ever enjoyed, and many of those users won't switch to the industrial-strength Rhapsody (in fact, many won't be able to, because Rhapsody will minimally require recent PowerPC-based machines). But Karen's point stands: will Macintosh hardware be sufficiently more attractive than lower PC clone prices to keep Rhapsody users on the Mac? And, to bring the discussion back to clone licensing, would a thriving clone market help the position of the Macintosh in the Rhapsody world?

The Real Deal -- I missed a couple of facts surrounding the entire situation in my previous article - most notably that terms for continued clone licensing had been agreed upon (but not on signed contracts) between Apple and the clone manufacturers just before Gil Amelio stepped down. Amelio has commented on this in several interviews.

<http://www.macaddict.com/exclusive/giltalks.html>
<http://www.mercurycenter.com/opinion/docs/009108.htm>

So what's the holdup? Most fingers are pointing at Steve Jobs, who appears to believe that the clones are significantly hurting Apple. That may be true, and it would imply that even the renegotiated terms weren't satisfactory to the post-Amelio Apple. The question remains: does Jobs want to eliminate clone licensing or is this a power play to squeeze yet more dough from the clone manufacturers? Is the overall reason that, as Gil Amelio says in those interviews, that Apple is afraid of competition? I don't know, but it's hard to discount Gil's opinions.

I said last week and I'll say again here that the license agreements must be acceptable to both sides. (Obviously, they were a short time ago.) Apple needs Macintosh clones, and either eliminating clone licensing altogether or requiring such usurious license fees as to make clone pricing non-competitive is self-destructive.

In short, I still believe the situation is complex and in many ways incomprehensible because of the limited quantity (and quality) of information that's been made public. However, I believe even more strongly that Apple must find a way to continue to support clone manufacturers. The elimination of clone licensing could be tantamount to corporate suicide. Consider these negatives, should Apple eliminate clone licensing altogether:

On the positive side, all Apple would seem to gain by eliminating clone licensing would be those sales that would have been cannibalized by clones. That's real money, but I can't imagine that it's enough to solve Apple's overall financial problems. I suppose a case could be made for Apple wanting to regain complete control over Macintosh hardware development, but that sounds utterly childish. No matter what, the negatives of eliminating clone licensing as I see them so far outweigh the positives that one wonders what Jobs or others at Apple could possibly think tips the scales in the another direction.


The Official AT&T WorldNet Web Discovery Guide

by Adam C. Engst <ace@tidbits.com>

My latest book, The Official AT&T WorldNet Web Discovery Guide (Osborne/McGraw-Hill, ISBN 0-07-882336-6, $24.99), is now available. I finished the book a while back, but a printing error and the UPS strike conspired to keep it off the shelves for several weeks. Although the title implies that the book is only of use to those using AT&T WorldNet and the CD does contain software for use with AT&T WorldNet (for both Mac and Windows, with a free month for readers of the book), the book is in fact a general introductory Internet book aimed at novice to intermediate users. It's very different from my Internet Starter Kit books, which attempt a comprehensive look at the Internet and Internet software. Instead, this new book assumes people will primarily use a Web browser and explains how to work with the necessary tools on the Internet, including search engines, Web catalogs, business and individual directories, and so on.

<http://www.tidbits.com/adam/book-frame.html>

In my view, the most interesting part of the book is the final section, which contains eight chapters about how to integrate the Internet into daily life. In my view, computers in general and the Internet in specific are overly isolated from our real lives. Just as we've allowed technologies like the telephone and the automobile into our lives in a wide variety of ways, so too must we integrate computers and the Internet they make accessible into the rest of what we do. Anyway, aside from being an example of that belief, those chapters are also rather autobiographical, so no snickering if you meet me at a Macworld Expo and I launch into one of the stories I tell in the book. We all have a limited number of anecdotes, and many of mine are now public.

You can look for the book at your favorite bookstore, and if they don't carry it yet, I encourage you to ask for it. There are also numerous Web-based bookstores that carry it, and we've joined the Amazon Associates program, which means that if you order a book by following a link from our Web site (also listed below), we receive a small additional payment from Amazon. Consider it another way to help support TidBITS.

<http://www.amazon.com/exec/obidos/ISBN=0078823366/tidbitselectro00A/>

If you're wondering what I've been up to since finishing this book (other than answering hundreds of email messages about clone licensing), watch this space for some major announcements surrounding TidBITS and another book (entirely about Eudora, one of my favorite programs).


Successful Shareware, Part 1

by Rick Holzgrafe <rick@kagi.com>

[Editor's note: This is the first part an article written by Rick that has appeared elsewhere on the Internet. We found it fascinating not only for the cogent advice to shareware authors but also as a personal explanation of how shareware works from the eyes of a long-time Macintosh developer. We asked Rick if we could edit and reprint his work in TidBITS, and he graciously agreed.]

Hello! I'm Rick Holzgrafe of Semicolon Software, and I'm a shareware author. My friends all know better than to bring up the subject of marketing shareware with me: once started, I won't stop talking about it. When I meet other shareware authors, I find that they're just like me: we're all seeking ways to be more successful. I'm not the world's greatest shareware success story, but I have been at it for over a decade, and my shareware income is now a substantial fraction of my day-job income.

Much of the advice in this article was contributed by others: most notably Peter N Lewis and Jeremy Nelson of Stairways Shareware, Kee Nethery of Kagi, Tonya Engst of TidBITS, and the authors on the Kagi Authors mailing list. Thanks to them all.

What is success? I considered myself fairly successful several years ago, when I sold 500 copies of my first shareware product (an adventure game named Scarab of Ra). There weren't as many shareware authors then as there are now, and few had written anything ambitious. At $10 per copy I made a few thousand dollars over the course of about five years: enough to buy some good software and even a little hardware. I also acquired a small reputation and a few nice reviews.

<http://www2.semicolon.com/Rick/Scarab.html>

Right there you have several criteria for success: fame, praise, and, money. In addition, by choosing your own projects you can gain experience you wouldn't get elsewhere. You can also make friends - some friends I've met through shareware are among the best I have; some are not only good friends but useful people to know.

I recommend shareware authors decide what their goals are. Some people are more interested in reputation, experience, and friends than anything else: they are the ones giving away good-quality freeware. However, most will be interested in "all of the above," with a heavy emphasis on money. (You need not be shallow and greedy to want to make money. No one knows better than a developer how much all this hardware and software and time costs. Expensive hobbies are easier to support if they bring in a good income!) The rest of this article is therefore mostly a discussion of how to increase shareware sales. If you can do that, you'll not only make more money, but the rest of the rewards you might want will come along with the dollars.

After some thought, I've realized that there are seven keys to shareware success. I call them the "Seven P's" instead of the "Seven Keys."

1. Product
2. Patience
3. Polish
4. Pay Up
5. Propagation
6. Promotion
7. Politics

You'll see as we go that there's more to shareware than just writing good code. For most of us code is the easy part!

The First P: Product -- You must choose a product to create and sell. There's a lot of software available already, and coming up with something new can be a stumper. I think most shareware authors build products they want to build, rather than software they think will sell. There's nothing wrong with that. However, if you want your product to sell well, it must have an audience, frequent use, an elevator presentation, robustness, commercial-class features, and an attractive price.

Audience -- Let's say you think paleobotany is a totally fascinating subject. You've just spent two years creating a killer application, using your programming skills and your deep understanding of paleobotany. Your program is so powerful and useful that's it's an absolutely essential tool for every person in the world... who is a paleobotanist specializing in the early Cretaceous period who uses a PowerBook in the field. How many copies will you sell?

That's right, not many. A successful product must be attractive to a wide audience. There must be a large number of people who might want to own your product; otherwise your sales will be few and far between. That might sound obvious, but apparently it isn't. I've heard more than one aspiring author describe an application almost as limited in audience as that paleobotany tool and complain bitterly that no one was buying it.

Here are a few examples of successful products with a wide audience:

<http://www.stairways.com/anarchie/>

<http://www.kaleidoscope.net/>

<http://www2.semicolon.com/Rick/STD.html>

You can also have excellent success with products that interest a large enough category of users, even if they aren't in the majority. Image-processing programs come to mind, for example: not everyone uses them, but their audience is still large.

Frequent Use -- Next, your product must be used often: daily if possible, and constantly is better than daily. I remember a shareware author complaining about poor sales. Why? His product was used only once by any user: it made some tweaks to the system software on disk; once made, the tweaks were permanent. This made the program easy to use and forget.

The author was right to complain that people used his product without paying for it. That's piracy, and it shouldn't happen. But the sad truth is that it does happen, and complaining about it won't change a thing. We'll talk more about getting people to pay later [next week -Tonya]. For now, the point to remember is that most users won't pay for a product unless they're constantly reminded about how useful or fun it is.

The three products above are good examples of programs used regularly. Anarchie is probably used at least weekly by anybody who likes to download software and updates from the Internet, and many people use it many times a day. Games like Solitaire Till Dawn are addictive, and people play them frequently. And Kaleidoscope, of course, is in constant use: its effects show in every window, button, and menu.

The Elevator Presentation -- Experienced salespeople know the value of a good "elevator presentation," a product description that quickly explains to a stranger what your product is and what its benefits are. In essence, an elevator statement must be quick and clear enough to be successfully delivered during an elevator ride.

There's a lot of shareware available, with more coming out every day. Drawing attention to your product in the middle of that stampede is tough. Nobody can try all the shareware available, so they download only the items that sound useful and interesting. You have about a paragraph (on a Web site, in a short review, in the Info-Mac Digest - wherever) to catch your audience's attention and convince it to try your product.

Robustness -- Here's one I'll bet you all understand, but it bears repeating: your product must be robust. No crashes, no bugs, no misbehavior - no excuses!

Perfection is difficult to achieve. I don't know anyone who has never shipped a bug. But you should hate doing that, and you should fix bugs as soon as you hear about them and release an update. Testing is a tremendously important part of your development process. When the major features work reasonably well, find a few people you trust and have them use your program and give you feedback about usability. Keep working and testing on your own until you think the product is ready to ship. At that point you must resist the temptation to ship. Instead, begin beta testing.

There are two approaches to beta testing: a small, hand-picked group or a large group. A small group lets you work closely with each tester, and you can choose your group so that every tester is a valuable contributor. But with a small group it's hard to get wide coverage of different hardware and software configurations, and different use patterns. A large group provides that wide coverage, but you'll have to settle for inferior bug reports (on the average), and you won't have time to maintain a relationship with every tester.

If you post a call for testers to a relevant newsgroup, you will have plenty of volunteers. I usually work with small, carefully chosen test groups. To find good people, I post detailed requirements in my call for testers. I reject those who fail to follow all the instructions in my posting because I don't think they're sufficiently conscientious. I require them to tell me about their system hardware and software, and to describe themselves, their experience with testing, experience with products like mine, and experience with computers in general. I use the answers both to separate the good testers from the bad and also to ensure a wide range of testers.

Allow at least two months for your beta test, and be prepared for it to last longer if significant bugs keep turning up. Run your beta test until the latest test version has gone at least two weeks without any problems reported. Nobody will buy a shoddy product. Make it work!

Commercial-Class Features -- Even though you're producing shareware, it must be as good as a commercial product. Anything less looks cheesy.

It's best if you can be as good as - or better than! - your best commercial competitor. That's not always possible, of course: few shareware authors can match the efforts of the army of programmers, artists, and writers that Adobe unleashes on every release of Photoshop, for example. You can still have success with a lesser product, provided it's a good one, and provided it's far cheaper than your competition. You can then be perceived as an economical alternative for users who don't need or can't afford the high-end product. An example that comes to mind is KeyQuencer 1.0, a $10 shareware program: not as fancy as QuicKeys, but cheaper and still a fine, powerful product. [KeyQuencer is now at 2.0 and sold as commercial software along with the shareware KeyQuencer Lite 2.0; see our review in TidBITS-351. -Tonya]

<http://www.binarysoft.com/kqmac/kqmac.html>

Attractive Price -- Shareware fans usually list bargain prices as one of shareware's primary attractions. If your product is as good as its commercial competition you might get away with charging as much; but my belief is that you're better off setting a noticeably lower price. Commercial products often come with materials like floppies, CD-ROMs, and printed manuals, which represent costs you don't have. Pass along some of the savings to your customers.

The Second P: Patience -- Maybe your first release will take off and make you a million bucks. Most don't, and that leads us to patience, a rather short topic.

Your first release probably won't be a raving success. That's okay: it's true for professionals too. The secret is not to give up. You'll learn from your first release: why people didn't buy it, what they didn't like about it, what features you should have added, what user interfaces were clumsy, what functions ran too slowly. In your second release you'll fix all of those, and your sales will increase. And you'll do the same thing again: listen to your customers and your non-customers, and make release three a real killer.

Consider Microsoft Windows. Version 1.0 was a dog. It was used by few and laughed at by many. But Microsoft didn't give up; they shipped Windows 2.0. It still wasn't a success but was better regarded than the first release. Real success didn't come until Windows 3.0 - and Microsoft still continues to improve it.

[Tune in next week as Rick discusses more keys to shareware success.]

[Rick Holzgrafe has programmed for a number of well-known Silicon Valley firms when he's not crafting shareware products.]


Non-profit, non-commercial publications and Web sites may reprint or link to articles if full credit is given. Others please contact us. We do not guarantee accuracy of articles. Caveat lector. Publication, product, and company names may be registered trademarks of their companies. TidBITS ISSN 1090-7017.

Previous Issue | Search TidBITS | TidBITS Home Page | Next Issue