Mac Musings

Classic Is Vital to OS X

Daniel Knight - 2001.08.27

Always thoughtful, Paul Shields of The Business Mac has raised the question: Is Classic really worth all that much (going forward)? His argument in brief:

Apple has invested a lot of resources to make Classic work and integrate well with Mac OS X.
A leading complaint about OS X is the lack of performance and hardware support in the Classic environment.
There are times when booting directly in OS 9 is the only solution.
Rebooting and running OS 9 is faster than booting Classic and running OS 9 there.
New Mac users are forced to learn to very different interfaces, which violates one of the most fundamental concepts of user interface design.
The Classic environment "encourages developers to drag their feet" in producing OS X applications.
Resources invested in the development of Classic would have been better invested in OS X.
A better OS& X would lead to a higher adoption rate.
The Classic benefit is short-lived and ultimately detrimental to OS X.

As an outsider (that is, someone who has not yet moved to OS X), I respectfully disagree. Shields is not taking a realistic look at the situation, whether for the business, home, or academic Mac user.

9 and 10

Apple has committed vast resources to make the classic Mac OS work inside OS X. If Apple believed it wasn't necessary to the success of OS X and the future of the Mac platform, I'm sure it would have invested those resources elsewhere.

There are precedents in the computer industry:

The Apple III was far more sophisticated than the Apple II, yet Apple included emulation of the older platform.
Microsoft Windows has always allowed the user to drop back into DOS or run a DOS shell within Windows.
The biggest obstacle to the Mac in schools was the installed base of Apple II computers. Apple's solution was an Apple IIe card for the Macintosh LC, which let schools run their old Apple II software while adopting the Mac as their new platform.
When Apple shipped System 7, it worked with software companies to produce an extensive list of applications that were and were not compatible with the new OS.
PowerPC Macs includes emulation for software written for the 680x0 family of processors.
When Windows became a 32-bit OS, it included compatibility with older 16-bit apps.

Backwards compatibility may cost a lot in terms of resources, but time and again it has been considered a crucial component in moving forward. At least part of what doomed the Commodore Amiga and Atari ST lines was a complete lack of backward compatibility with earlier Commodore and Atari computers.

Apple is not facing a hardware problem. All the current hardware runs both OS 9 and X, all the PowerPC-based models ever built support OS 9, and all but one of the G3- and G4-based Macs will run OS X.

We have a situation similar to that in the Windows world, where the same hardware can run Windows Me, Windows 2000/XP, or Linux. One of the crucial Linux tools is WINE, a Windows compatibility layer that lets Linux users run Windows applications. Without that, users would have to reboot under Windows to run non-Linux software.

As early adopters of Mac OS X are learning, rebooting under OS 9 to use your scanner, printer, or some other device is a nuisance - because when you're done, you have to reboot into OS X to use your regular applications. Rebooting to switch operating systems is simply inefficient, but it remains the only choice for OS X users with unsupported hardware.

Classic Performance

Those who complain about Classic performance should really work with Virtual PC (VPC) for a while; that will let them experience a real emulation slowdown. That said, I hear VPC is getting better by the year - and Apple is working to improve Classic performance.

As an outsider, I find the performance of my 400 MHz TiBook under OS 9.2.1 very impressive. I've maintained for a couple years that once you pass the 300 MHz G4 mark, you've got plenty of speed for all but the most power-hungry Photoshop and digital video users. My understanding is that Mac OS X has a lot more overhead, so you really want at least a 350 MHz G4 or 400 MHz G3 for comfortable performance.

To get comparable performance in Classic mode, maybe you have to use a Mac with a CPU that's 20-50% faster - anyone tested that? After all, no matter how good Classic mode is, it's running in emulation on a much more CPU-intensive operating system. My guess is that the bulk of this is not due to the core OS itself, since Unix has a reputation for efficiency, but because of the incredibly impressive, power-hungry Aqua interface.

I wonder if Classic would run faster if you turned of Aqua, which begs the question - can you turn off Aqua while running in Classic mode?

Regardless, for all but the most intensive tasks, Classic performance should be comfortable even on today's low-end Macs. No, it won't be as fast as running OS 9 all by itself, but it should at least be adequate for word processing, email, spreadsheets, etc.

Hardware Support

Shields hits the nail on the head with hardware. As we approach the first significant upgrade to OS X (version 10.1), it is inexcusable that many hardware manufacturers have not yet released drivers. If they have already committed themselves to the Mac platform, they owe it to their customers. And if they have not yet supported the Mac user, now is an opportune time to do so.

This could also eliminate the primary reason OS& X users sometimes need to boot in OS 9.

Different User Interfaces

Yes, learning DOS and Windows at the same time would have been very frustrating. New OS X users may experience the same frustration when using Classic mode - so we should do all in our power to minimize the need to do so.

That said, most of us operate within two different user interfaces every day: the classic Mac OS, OS X, some version of Windows, or one of the supported Linux GUIs and our chosen Web browser. The browser has defined a new GUI that's neither Mac nor Windows. One click on a link activates it, not the Mac's double-click. Realizing this, Microsoft is shaping its interface to reflect our familiarity with Web browsers.

Using multiple user interfaces is not an insurmountable problem, and there are ways for Apple to make it easier. The best thing Apple could do is provide a default Aqua appearance for the version of Mac OS 9 that ships with OS X 10.1 in September. If not the default, Apple should make an Aqua option available or at the very least not attempt to quash development of Aqua-like appearances for the classic Mac OS.

At the same time, Apple should devise an appearance for OS X that is more classic-like - or at least not discourage efforts like the Sosumi Theme. By making the Classic environment and Aqua more similar, Apple can make it easier for new users to become proficient in both.

Dragging Feet or Driving the Market?

For every Power Mac and clone owner trying to hack Mac OS X onto unsupported hardware, there are probably dozens of Mac users who own supported software and are not moving X-ward. A very limited market is a great obstacle to developers. Until OS X reaches critical mass and grows a large enough market for their application, there's no strong need to create a Carbon or Cocoa version of a program that works just fine on the classic Mac OS that the vast majority of Mac users are still using.

You could accuse me of dragging my feet, since I have no intention switching to Mac OS X until version 10.1 is available and there's are least one OS X application that makes it worth switching. I may buy OS X 10.1 to become familiar with it, but I won't switch until I have a reason to do so.

I'm sure I'm not alone.

There are three obstacles to consumer upgrades:

The classic Mac OS works just fine.
There is no X-only application to drive consumers.
It costs a fair bit of money to switch.

Apple could address some of these issues and force developers to move forward. For instance, if Apple developed a new version of AppleWorks that was X-only, directly supported Microsoft Word and Excel formats, and came free with every Mac and every copy of the Mac OS, you can bet Microsoft would put development of Office:mac for OS X in high gear. Likewise, a free X-only iPicture application could light a fire under Adobe to get Photoshop and Photo Deluxe running on OS X.

An AppleWorks that was fully compatible with Word and Excel could also provide a compelling reason for individuals, businesses, and schools to invest in the OS X/AppleWorks package, helping to create the larger market that smaller software vendors need.

We Still Need Classic

But we're still dealing with a lot of outsiders - potential OS X users who still have no reason to switch. In addition to a compelling software solution to entice us to move forward, Apple needs to support all our old software until good replacements are available. And the precise meaning of "good" in this context is highly individual.

For instance, I love Claris Emailer, which has worked well under System 7.5.5, 8.1, 8.6, and now 9.2.1. I like the fact that it doesn't display styled text and only sends 100% compatible standard text email. That means I know the recipient can read what I send; with most newer email clients, I'd have to override default settings to use what is an industry standard - plain text email. And I'd have to worry about reading HTML email that will attempt to load graphics over the Net and let the sender know their email (too often spam) has been received.

Not only that, but I'd have to reconstruct the filing system I use, recreate the address book and groups, rewrite all the scripts that file incoming messages, set up all my email accounts. Changing to any email program, X-native or not, would be a tedious process and break work habits I've developed over many years.

As long as I stick with the classic Mac OS, whether freestanding or under Classic mode in OS X, I don't have to give up Emailer.

Likewise, I can stick with Photoshop 4, BBEdit Lite v4.6, Mizer 1.2, Microsoft Word 5.1a, Claris HomePage 3.0, and several other vintage applications that work just fine for me and will never run as native OS X applications.

If Apple hadn't developed Classic, I would not be able to consider buying OS X 10.1 in September. If I had to reboot into OS 9 every time I ran Claris HomePage, checked my email, etc., I wouldn't even buy 10.1 to experiment with it.

We need Classic mode so current Mac users won't lose all their software and have to learn new work habits to go with new programs. Classic lets up migrate to OS X on an application-by-application basis.

Return on Investment

Apple realizes that, and that's why it has invested so much into making Classic work and work well. Steve Jobs already lost one visionary, bleeding edge company (NeXT) because it was not backward compatible with anything.

Without a solid Classic environment, Apple would be creating more obstacles to our adopting OS X. By investing resources into Classic, Apple facilitates the migration from the classic Mac OS to the Mac OS of the 21st century.

As Shields notes, a better OS X will lead to a higher adoption rate, but I believe that a better OS X with a better Classic environment will lead to an even higher adoption rate. Just as Apple benefited from 680x0 emulation when it introduced the Power Mac, it will benefit from a strong Classic environment inside Mac OS X.

Sure, in time most users will be running mostly X-native software, but what of Claris HomePage, SimCity 2000, Mizer, AutoCAD, Shanghai II, and any number of other programs that will never be ported to OS X? Without strong backward compatibility, why choose OS X over Linux or the even more popular Windows?

Without Classic, OS X would be facing the same kind of uphill battle the Mac did in 1984. It wasn't a DOS PC, but neither could it leverage the Apple II user base.

For us to move forward, we need backward compatibility. For Apple to survive the transition to OS X, it needs the best Classic possible. Whether Classic needs further development or not is a decision I leave in Apple's capable hands, but I believe it is wrong for anyone outside of Apple to tell it that Classic is good enough and to invest their resources elsewhere.

As Shields notes, though, Classic is only half of the equation. In addition to supporting classic Mac software, Apple needs to encourage developers to create the X-only applications that will compel us to adopt OS X.

Still, not a penny spent nor a minute invested in creating a solid Classic mode has been wasted, since Classic is as essential to our forward migration and Apple's continued existence as good X-native software.