The best book about marketing for software developers that I know is The Macintosh Way by Guy Kawasaki, and I've compared it to NeXT-developer strategies over the last five years. I can only say that Kawasaki was exactly right. He probably would have chosen to support Lighthouse Design, RightBrain, Appsoft, Pages, etc., a long time ago, rather than spending time and money on WordPerfect.
If I remember it correctly, one of the main lessons of the book was: Every platform needs to build its own heroes. And only very few companies are successful with cross-platform applications.
The PC market created its own software heroes, which became giant companies: Borland, Microsoft, Lotus, Ashton-Tate, WordPerfect, etc. None of them has been very successful on other platforms (except Microsoft on Macintosh). Apple failed with Ashton-Tate, Borland, and Lotus, though it tried very hard to get them onto the Mac. In the end, they succeeded with newcomers like Aldus, Adobe, Quark, and Symantec.
NeXT made the same experience. (Or should I say mistake?) It will not succeed with Lotus, WordPerfect, or Frame. And not with Aldus and Adobe either. The NeXT heroes will be Lighthouse, Pages, Athena, OTI, etc.
The developer meeting in San Jose in November showed this clearly. It was a meeting of NEXTSTEP-only developers. And the message of the meeting was not that it was restricted to 40 companies, but that NeXT should have focused on these companies a long time ago.
NeXT should extend its developer marketing to the VAR-type developers now, because they have their own vertical markets. I think the Object Channel is maybe a good approach to this. The "support your own heroes" strategy is not only true for software developers. The same thing holds for distribution.
d'ART Software GmbH
I've wanted a NeXT machine since I saw a picture of the first Cube, and I spent the better part of a year trying to convince my boss to let me try a NeXT machine. Finally, when it was announced that NEXTSTEP for Intel would be available at the '93 NeXTWORLD Expo, he relented.
In April, I bought a '486 from Dell that was configured according to what NeXT and Dell had published. In May, I arrived at the Expo and was amazed by the small number of products that were actually shipping for NEXTSTEP for Intel. Several software vendors told me to call them after the Expo, and they would give me a "special" price on a beta version. This was in stark contrast to everything that I'd read about NEXTSTEP Ð it was supposed to be "almost trivial" to port from black hardware to Intel machines.
And Dell, my hardware vendor, had several salesmen running around who couldn't answer a single question Ð even simple ones. I didn't think that my computer would run NEXTSTEP, even though I did everything that Dell and NeXT said to do.
When I returned home from the Expo, NEXTSTEP refused to load. Several phone calls to NeXT and Dell proved fruitless. Finally, out of desperation, I removed my video card and have been living with 8-bit black-and-white since.
On the software side, I called Pages Software as recently as December 23 to order Pages. The person I talked to told me, "Pages is not shipping, and I have no idea when it will ship." How can I or anyone else talk our bosses into switching platforms when we can't even get a decent word-processing package for NeXT?
If I were Steve Jobs, I'd give software houses free copies of NEXTSTEP and the developer tools. I'd also send an engineer to see they get everything installed. Then I'd offer free developer training. Borland would probably be a good place to start: Kahn must hate Gates as much as Jobs does.
New Iberia, Louisiana
Being a nonprogrammer that is determined to learn NEXTSTEP programming, the Garfinkel and Mahoney book moved much too fast too soon. Sure, I could follow along and build all of the sample apps, but I didn't understand why I was doing what I was doing. About half-way through, I put the book down and picked up Alex's.
Yes, at times the reading seems slow, and, yes, I kept asking myself why I was coding something I knew I could do in InterfaceBuilder. In the midst of my lamenting, I would receive a flash of insight that would help me understand what Simson and Michael were doing or what all the fuss is over InterfaceBuilder.
I'm now working my way through NeXTSTEP Programming Step One: Object-Oriented Applications again. This time the examples make more sense, and I can more easily keep up with the fast pace. The OOA and OOD background Alex provided has made all the difference.
Michael A. Duke
Big Flats, New York
This is the purpose of NeXT's Mentorship program. We help developers who are new to NEXTSTEP explore design and implementation issues and prototype ideas in the safety of the classroom, without corporate pressure. We help developers complete the first cycle of a small part of their corporate project, so that they can learn the philosophy, described above, as well as the technology.
Redwood City, California
We know that our interface can be improved, and very soon we will introduce solidThinking Release 2 to fill this gap (iconified interface, drag and drop, etc.), but we don't believe the interface should become the first parameter to judge an application. We want to satisfy the production needs before those of the eyes.
For the record
In the January 1994 "New in Shrink-wrap," the phone number for Doberman Systems was incorrect. The correct number is 801/944-4329.
NeXTWORLD welcomes your comments. Mail them to Letters at NeXTWORLD, 501 Second St., San Francisco, CA 94107; or e-mail firstname.lastname@example.org.