I would like to respectfully suggest that we have had enough research into the harmful effects of tobacco smoking.
I make this suggestion for two reasons.
First, we need the Federal revenue produced by cigarette taxes to reduce the deficit. There seems to be no practical limit on cigarette taxes. Smokers will basically pay whatever they have to pay for cigarettes.
Don't believe it?
Consider: Yes, we've all heard ex-smokers claim that they just couldn't afford it any more. But have you ever met an ex-smoker for whom you truly believed that price was the decisive factor in kicking the habit? There are so many other reasons to quit smoking: social ostracism, the need to climb stairs, the health of family members, inconvenience, the spot on the X-ray, dead relatives. According to my research, there has never been a case of a smoker giving it up solely because of the cost. But what about cost as a contributing factor?
Consider: Canada has raised taxes on cigarettes with absolutely no effect on cigarette sales. Granted, it did create a thriving black market in cigarettes imported from the Lower 48. But adjustments to U.S. and Mexican tax codes and a little tweaking of the North American Free Trade Agreement should
take care of Canada's black market problem and head off any such problem for the U.S.
Consider: Nicotine is an addictive drug, and the street price of other addictive drugs suggests that addicts will always find the money to support their habit. My research suggests that a tax of $100 per pack would have zero effect on cigarette consumption, and would bring in an extra $0.3 trillion per year in Federal revenues. (Or, to put it in more meaningful terms, it would pay for 20 haircuts per second for the President for the duration of his term.)
What does seem to have some slight effect on consumption is all these discouraging studies. Eliminating them and instituting a $100/pack tax would put quite a dent in the deficit. That's my first reason. Lest it appear that I am placing economic issues above public-health issues, my second reason will make clear that I am very concerned about public health.
Reason two: We should take the money currently being spent on research into the harmful effects of tobacco smoking on physical health and spend it on research into the harmful effects of tobacco growing on mental health. This area of research has been shamefully neglected, and it's obviously a serious problem.
What other explanation could there be for Senator Jesse Helms?
All the examples of Helmsmanship look pretty much the same. Here's the latest, as of press time.
The Senator from North Carolina probably wouldn't care for what's going on in operating systems today. The key word is "open". PowerOpen, Open Systems Foundation, everything is openness. Especially at Apple, where, it would seem, a long-standing tradition of proprietariness has been trash-canned.
The most dramatic evidence in how far things have gone was a statement made by Apple's Senior Vice-President and General Manager of the Macintosh Software Architecture Division and the Advanced Technology Group, David Nagel, at Apple's Worldwide Developer's Conference this May.
Nagel said, "From now on, no major technology will be introduced from Apple that is not cross-platform. The goal is near-simultaneous release."
Now, there are at least three ways one could read this statement.
You could take it to mean that Apple will henceforth follow a model of developing system tools and making them simultaneously available on the Mac and on Windows. This seems to be a safe reading, supported by a recent announcement from Kaleida, the Apple-IBM joint venture in multimedia technology. Kaleida decided to delay the release of its ScriptX multimedia scripting technology until next year so that it could be released simultaneously on Mac and Windows.
You could also take Nagel's statement to mean that Apple is decoupling its operating system and user-interface software from its hardware. I'd say that's a safe reading, too. In a dramatic break with the past, Apple is licensing its operating system technology to others, and porting or supporting the porting of the Mac OS and/or GUI to DOS, Windows, and various UNIX platforms.
You could also read Nagel's statement to mean that Apple is getting out of the hardware business. Well, couldn't you? "From now on...no major technology...that is not cross-platform." Ever heard of a cross-platform computer? Taken literally, Nagel's statement means either that Apple is out of the hardware business as of today, or that any new machines it introduces in the future will not be major technology.
But this is nonsense. Apple will be introducing PowerPC Macs within the next year, based not only on a new chip family (the PowerPC) but on a different chip technology (RISC vs. CISC). That's surely new hardware technology. Nagel didn't mean that Apple was through pushing iron, and he didn't mean that its hardware releases wouldn't be "major technology." Anyway, Nagel is a system-software guy; he was clearly talking about his domain, not all of Apple.
But the assertion wasn't qualified in any way. So isn't it just a little suggestive that it didn't raise any eyebrows within Apple? This major announcement surely passed under many eyes before being delivered at the conference. Yet nobody thought to mention that Apple was still, of course, going to produce machines, for a while at least. Which shows, perhaps, where their minds are. Apple really is starting to think of itself as something other than a computer manufacturer. It's comforting, I guess, to know that some things don't change, and that the Apple corporate mind remains farther out on the curve than the Apple market reality.
In any case, the uncoupling of the Macintosh operating system and user interface from the Mac hardware is radical. What does it mean?
First, there's the Apple-Novell plan to put the Mac operating system and UI on Intel-based machines. This sounds like it would have been a stunning coup for Apple two years ago, but today it seems a little late to challenge Windows on its home court. What's the deal?
Well, it's supposedly a plan to offer an alternative to DOS users who are ready to bite the bullet and move up to a graphical UI, or Windows users considering the move to NT. It's due out in the first half of '94, and here's what it consists of:
The Mac Finder and Toolbox, displaying a System 7 interface, running on top of Novell's DR DOS, with the Mac file system mapped onto the DOS file system, the whole thing capable of running on 386, 486, and Pentium boxes, and supporting multitasking of Mac and DOS applications. Not Windows apps, at least not yet.
And that's not just existing DOS applications, but ported Mac applications. So what the deal is for developers is this: You port your Macintosh apps to this Intel platform and you'll have an expanded market for them. Since the Toolbox and some other Mac components are already ported, the port won't be too daunting a job, and Apple and Novell will explain to you what steps you need to take to do the port.
I'm sorry, but I just don't see this as a terribly compelling deal for application vendors.
If they are considering porting their apps to Windows, they now get to compare the currently nonexistent market for applications on this MacDOS platform with the existing huge market for Windows apps, weighed against the relative porting costs. Unless the port is very easy or Apple and Novell push this platform very strenuously, it's hard to see the comparison working out in Apple's and Novell's favor. Vendors will port to Windows first.
And if the vendors have already ported to Windows, then any success that this platform has in cutting into the Windows market will arguably be at the cost of the application vendors' sales on Windows. If they do find it necessary to port to what I'm calling "MacDOS," it will be grudgingly, to win back market share that they were already counting as theirs. Not an inviting scenario.
There are scenarios for success with this thing. If porting is extremely easy and if enough vendors port their apps to it soon enough that Apple is able to say convincingly, "You can now have the virtues of the splendiferous Macintosh interface rather than that inferior copy of haphazard design and inconsistent execution, and run true Macintosh applications on those inelegant dirt-cheap clone boxes from the Orient that you're so fond of," then maybe there's a decent-sized market. And that market would be even more decent-sized if Apple and/or Novell could figure out how to give an affirmative answer to what you just know will be the first question that anybody who owns an Intel machine will ask about this platform; to wit: "Will it run [fill in your favorite Windows app]?"
But how likely is any of that?
Then there are the plans, very plural, to port the Mac system and/or interface to UNIX.
First of all, there's Apple Services for Open Systems. This approach lets Mac applications run without modification on UNIX platforms in an X window. The pitch for the user is, you keep the benefits of the UNIX operating system, and the existing pool of Mac applications is open to you as well.
Second, there's the plan to port the Toolbox and APIs to various UNIX platforms. This would mean, if I've got it right, that in addition to porting existing apps to UNIX, you could write something that is obviously a Mac app, but that's written specifically for the UNIX platform you're working on. A weird trip. The pitch to the developer is, you want to work in UNIX but you like QuickDraw or AppleShare or Apple Open Collaborative Environment? We'll sell you the technology you want, ported to your particular hardware platform.
Then there are the Quorum approaches. Quorum Software Systems (Menlo Park, California) has two products that let Mac software run on UNIX machines. One is a $6,000 porting tool named Latitude; it lets you port Mac apps to UNIX without touching source code and reportedly delivers good performance. The other is a $695 end-user product that lets several specific popular Mac applications run on UNIX systems, using the host system's user interface; Motif, for example.
If you're one of those who thinks that the worst thing about the Mac is the operating system, then some of these UNIX moves may sound a lot more interesting than the Apple-Novell plan of grafting the Mac UI onto DOS.
One almost gets the feeling that these ports of major pieces of the Macintosh to existing platforms are just practice for the main event, which is the port to the PowerPC. In fact, Apple is telling vendors that making their apps portable to the Intel platform will help in porting them to the PowerPC. This IBM/Apple/Motorola-engendered RISC chip is Apple's future (and not so distant future, at that) hardware platform. At the developer's conference, Apple executives were gloating that PowerPC was delivering better performance than expected, and drawing a lot of comparisons with the bigger, doubtless more expensive, more heat-producing Pentium. More than double the Specmarks per watt was one comparison. Apple expects to sell a million PowerPC Macs in '94; this is the first mass-market delivery of RISC technology, and Apple really plans to mass-market it.
Unported Mac software can run on Apple's PowerPC machines via emulation, an approach that has rarely in the past produced acceptable performance on any platform. But at the developer's conference, Apple demonstrated Mac software running with impressive speed in 680x0 emulation mode on a PowerPC machine. Because the Toolbox has been ported to the PowerPC, and because of the advantages of the PowerPC over existing Motorola CPUs, the emulated software can actually run as fast as or faster than it would run on a fast 68040 Mac. If ported rather than emulated, it could run several times faster.
There are currently two ways to port Mac apps to the PowerPC. EchoLogic's FlashPort is a direct binary-to-binary translator from 680x0 code to PowerPC code. That doesn't sound ideal, but Apple is using it to port elements of the Mac OS to PowerPC, with apparently excellent results. And Apple and Symantec should have a C compiler for PowerPC out about the time you read this. Pascal programmers will have to use FlashPort or wait longer, I guess. And then Bedrock is the next step. It'll be possible to get your hands on Bedrock late this year; the real final release is scheduled for mid-94.
Bedrock is the replacement for Apple's unloved MPW, which is going away unmourned. Bedrock is Symantec's and Apple's C++ application framework for writing portable applications. The idea is that you'll have a single source for your entire program, including all resources, and it'll be written with Bedrock, and that source will compile to a Macintosh application for a 680x0 Mac, a Windows application for an Intel machine, or a PowerPC Mac application for a PowerPC Mac. It's important to note that: 1. Apple intends this to be used not only by commercial developers but also by corporate developers; and 2. Mac and PowerPC Mac and Windows platforms are just the first platforms that Bedrock is intended to support.
Bedrock isn't Apple's (or Symantec's) last word on multiplatform development. Bedrock has been designed to let you use the parts of it you like and bypass it where there is some advantage in doing so. Apple and Symantec are also working on various native PowerPC development tools, including, I suspect, MacApp. Some of these will be out this year.
Why all this porting madness all of a sudden? Is it just a prelude to PowerPC, or is there another reason? Maybe, mmm, oh I don't know, could it be--SATAN?!?
Yep, Apple's responding to the latest campaign by the Evil Empire to take over the world.
And one of the most dramatically open technologies is a direct response to Microsoft's OLE. (Well, some might call it a response to NeXTstep.)
I'm talking about Amber.
Amber is a technology for the development of a new form of software product for a new model of computer use. The model is document-centered computing, and the form of software is the component tool.
Document-centered computing means that, rather than working with application programs that support specific kinds of documents, the user works with a document and calls up particular tools or suites of tools to operate on different kinds of data in that document. The user never leaves the document to get something done. Any document can support any kind of data. All tools are editing tools of some sort, and all editing tools are usable in all documents.
This model implies a different form of software product, the small, focused tool or suite of tools, as opposed to the monster app. Since the document-centered model takes away from the third-party developer most of the interface and all of the integration of capabilities, it pretty much destroys the whole concept of the application as we know it today. Writing document-centered software tools won't be hard, but it will take some guidance and specialized tools from Apple, and that's what Amber is about.
Amber is like Microsoft's OLE and will compete with it, but there are a couple of interesting twists. First, Amber will fully support OLE. Second, Amber source code will be public domain. Apple is going to release all source for any use without restriction. Apple is setting up a nonprofit Amber association to promote Amber, discuss standards, port the code, and the like. Amber is intended to be available for Windows and UNIX as well as Mac. The Mac source code will be seeded to developers by the end of this year.
The final twist is that Amber itself, Apple admits, is just a stepping stone to programming for Taligent, Apple's next operating system.
Next month: Satan's side of the story.
Copyright © 1993, Dr. Dobb's Journal