Many of you have been hearing for quite some time about Apple's not-so-secret Marklar project, a project designed to allow Mac OS X and its iApps to run natively on Intel's x86 (Wintel) platform. Marklar has been cast as a strategic Plan B for Apple, in case the PowerPC microprocessor platform cannot keep pace with Intel's various chip linestoday consisting of its market dominant 32-bit x86 platform known as IA-32 (Intel Architecture 32-bit) and its emerging Itanium 64-bit platform (IA-64).
However, without going into the merits of having such a strategic plan and without getting into a philosophical discussion about the virtues of RISC versus CISC chip architectures, does such an option make any sense at all for Apple's Macintosh developers? How would it benefit them? Would moving to Intel make any sense from their point of view? And if this happened how would it shape the Macintosh software industry? And would a Mac be a Mac without PowerPC?
To help us delve into these complex issues and get a feel for what developers' think, we asked two of the Macintosh community's most important CAD and 3D software developers. Both differ in terms of their solutions' market position and their financial strength. One is what you would call a small but brilliant "garage scale" developer. The other is a large developer with a worldwide installed base of an award-winning line of CAD and 3D software.
Both have decades of experience developing for the Mac and are truly passionate about the platform. Architosh had a good chat with both. Read on...
We start our discussion by talking to Sean Flaherty, CTO of Nemetschek North America (sponsor), a subsidiary of Nemetschek AG of Germany (and formerly Diehl Graphsoft Inc.) Some background is clearly in order. For those who do not know, Nemetschek NA. makes the award-winning line of VectorWorks CAD and 3D software. The product was previously known as MiniCAD and has been on the Mac platform since the beginning.
So we start by asking one central key question that must be asked to make the whole Marklar topic valid.
AFR: Recently there has been reports of Apple's secret 'Marklar' project, an effort to bring the Mac OS X platform to x86. As a Macintosh developer do you think there is any chance of this happening?
SF: I think since the roots of NeXT were available on x86 platforms there must be a strong debate about whether or not this should be done. For me, this asks the root question: what is a "Macintosh"? I think Apple would lose an important part of its appeal if it decides that the Macintosh is just an OS and not a platform combining an OS and hardware specifically designed to work together as a "unit" [quotes added by ed.]. From a business standpoint, successful Apple years historically have also been matched with strong hardware sales. It will not be an easy decision to make, but I would guess that it hinges on the future outlook for the G5 processor family.
Sean raises a key question that many Mac fans have raised on the Internet whenever this topic comes up. And that is, what is a Macintosh? The issue has come up prior during the Clone Years and some Mac fans had no problem accepting the clones at all because it was the OS that made the "Mac Experience". But today Apple's industrial design is raising the bar across the board and shaping the Mac experience right down to the very way in which you buy a Mac.
It should be brought up that Sean's reference to NeXT's roots in the Mac OS of today goes beyond the code base to the very politics of Apple's OS group since much of its key programmers are former NeXT employees and managers...all of them agents of influence in the halls of Cupertino.
Regardless of this NeXT contingent and their desires it will likely all come down to Steve Jobs and his desire to control and mastermind the Macintosh experience entirely. What Jobs is up against however are the realities of Apple's competitive fitness in compute performance. Can Apple and PowerPC continue to compete? But if the answer was to turn to negative, and if Marklar was to move forward, what next? How would it affect the developer? We asked.
AFR: What kind of impact would moving to x86 bring to you as a developer? How would it affect your company's strategic direction and bottom line...in general terms?
SF: Supporting two processor architectures can be quite expensive. For us, the cost would not be so high since we already support x86 hardware with our Windows version, but applications available exclusively for the Macintosh would have significant work, similar in scope to the PowerPC introduction a few years ago [ed. refers to 1994]. The work specifically involves a lot of low level assumptions such as byte order ("endian" order), basic data type sizes, efficiency of data alignment, differences in memory architecture, etc. In addition, delivery technologies, testing, support, and a host of systems outside strict software development must be expanded to include the new architecture. I know we breathed a big sigh of relief when we could stop support for 680x0 machines since it cut our Macintosh support costs significantly.
Most Macintosh CAD and 3D developers currently support Windows as well. There are a few apps--like PowerCADD--which are entirely Mac-based. Sean raises points that often are overlooked in discussions by end users online speculating the merits of Marklar...namely, such items as support cost.
Where as Nemetschek is a large Mac developer (relatively speaking), Nader Family, CEO of BOA Research Inc. (sponsor) is not. His response to this issue is quite different on many levels.
NF: Most people have heavy investment in CPUs and peripherals. As their hardware are generally robust enough for CAD, there will be little incentive for them to move to a new platform as an upgrade. In other words, the advances in CPU speed over the last two years make it unnecessary to upgrade as often. It is at the time of upgrade that people may think of switching to another platform.
NF: Second, as Autocad is truly entrenched and because of its business commitments will not move onto a platform beyond Intel, we can't expect people to switch both hardware and software. They have heavy investment in both, especially software training. The only option for small software companies like BOA Research is to make sure that their products are complementary to Autocad and fill specific voids (design and design development) on the same hardware platform as Autocad.
NF: OS X on Intel will help small developers like us, who can't afford to serve both platforms economically. Today we are faced with budget shortfalls and need to make a clear decision: OS X or Windows?
Nader's comments reflect the tougher economic realities for today's smaller developers, as well as his products' strong attempts to penetrate the architectural CAD market. Nevertheless, while Autocad may be a hinge-point for people's hardware decisions, that company may eventually break that hinge-point for Apple and Mac users if it was ever to release their core products for Mac OS X.
For the smaller developer like BOA, OS X on Intel would alleviate the expense of supporting two hardware platforms. But what does that do to a true Mac developer? A Mac developer that has been on the Mac for years?
NF: On the one hand, we have a lot of intellectual investment on the Mac side. We don't want to start over on Windows. On the other hand, if we do make the choice to move over to Windows, the pay back will be that we will dramatically increase the number of potential users. If OS X ran on Intel, it would make our decision for us.
The question Apple must be facing is: "if the smaller developer abandons PowerPC because OS X on Intel 'makes their business decision for them', then what does that do to PowerPC long term? And how does that impact Apple"? Various speculation would include that it would set into course an irreversible destiny of Intel-dependency, a place where Apple may not wish to go for various strategic reasons.
Apple currently has the ability to outperform Windows systems when Intel falls behind to the PowerPC platform. Back in the G3 days this actually happened in a noticeable way and it meant not only robust sales for Apple but it actually helped save them. This is something that Apple might wish to keep as a market advantage.
Steve Jobs remarked during a recent shareholder's meeting that Apple likes "to have options" in regards to processors but is Marklar too risky a move? And with IBM's resurgence in PowerPC processors, shouldn't Apple give PowerPC one major last go? We asked a similar question to Sean Flaherty:
| 1 | 2 |