Bahamut's Lair

[Editor’s note: The following appears in MacEdition courtesy of The Heretical Monks (RdF, Bahamut, Soup, and others), and first ran for the subscribers of Robert Morgan's RFI. It was written before Apple announced MacOS X DP4 and the pending “public beta”.]

If last week, Bahamut played Happy Oracle, outlining the possibilities for a skunkworks project at our favorite fruit that may or may not see the light of day – and the past week's alliance between AOL and Transmeta both lights a fire under Apple and threatens the viability of an AMP – today Mr. B and other members of the Heretical Monks, including his old archenemy Soup, tackle something absolutely real and absolutely threatening to the outfit that once flew the Rainbow Flag with pride. The threat we speak of today is Aqua, that lickable – or is it suckable? – interface that we’re all expected to migrate to next year. The Manifesto that follows was written a few months ago in response to our experience with the version called DP3 that was issued to developers, and some of the features of Aqua have changed since then, but we’re still far from sure that we’re going to enjoy living in a Barbie World.

The Heretical Monks’ Manifesto – “Mac OS X. More Lickable. Less Usable.”

Who are the Heretical Monks? We are Robert Morgan’s partners in analyzing the computing and technologly industry and the directions it may go. We help sift through the information (or disinformation) that comes through and make sense of it.

Why the The Heretical Monks’ Manifesto? We feel it’s an opportunity to be heard at the highest level, to express concern and give encouragement to a company that’s part of our professional career paths, that we hold stock in, that we write software and evangelize for.

There’s a concern among us that Apple may be “going off the rails” again – slipping in its attention to its customers in pursuit of a vision coming forth from but a few people. So far, indications that Apple is not “messing with the brand” (another of Robert’s phrases) are few and far between. We’re left to hope that WWDC and release to manufacturing will prove us wrong, that Apple is sincere in giving customers a good migration path to X, and a consistent Mac OS experience, rather than forcing Unix and NeXT on people with a candy-coated “lickable” gloss.

The Aqua interface’s departure from the Mac OS makes it clear that this is an entirely new operating system. Although this may be Apple’s intent (although at that point one questions if renaming the company and the OS to say, La Pomme Dangereux and Aqua 1.0 might not make some sense), it is also a very dangerous strategy. If Aqua is perceived as a whole new OS with a steep learning curve and a new way of working, Apple is guaranteed significant defections at every demographic level as users reassess their commitment to the platform. It is in Apple’s strategic interest to portray OS X as an upgrade, albeit a very significant one. Hence, an option for an utterly classic GUI (Apple menu, no goddamn Apple in the center, control strip, no dock, etc. etc.) should be provided.

If, on Day One of OS X, Steve can say that this is as gradual a shift as the one from 68k to PPC, but with far more dramatic consequences and revolutionary abilities that are under the control of the user/customer, then we’ve all (customers and Apple) won. If not, Apple’s in for some seriously rough times, the outcome of which is very uncertain and could result in nothing less than the destruction of the brand.

If the latter is the case, it wouldn’t be the first time that a company botched an OS migration and gave competitors an excuse to poach market share – just look at Solaris, for instance. Solaris was a failure as an OS migration, because arbitrary and unpopular changes caused mass defections, which lost Sun the workstation market. Huge, gaping holes opened up for competitors, especially NT, to slip in and steal away Sun’s core engineering, scientific and financial markets. Sun is only alive and kicking today because it was able to sell pricey, high-margin midrange servers, and lots of them, to stop the bleeding caused by the collapse of their workstation business. Apple has fewer core markets to lose than Sun, and no enterprise strategy to bail them out.

Our experiences with DP3 show many things in an incomplete state, and while we understand it’s a developer “preview”, it still distressed us. We aren’t surprised that Classic integration with the OS is lacking – we understand that this is some of the most challenging work available. What we were surprised by is the way in which the Mac OS X Finder seems incomplete – more interested in the glitz of Aqua and Quartz than in implementation of core Mac OS behaviors like Drag and Drop.

Examples of this inattention are things like the change between drag behaviors between the Desktop and file system (a drag to the desktop is an alias by default, but a move in other places in the file system), problems with getting applications to recognize filetypes dropped on them, problems getting applications to open files without extensions like .pict or .txt, problems with Dock usability and understanding Dock behaviors. DP3 doesn’t give the feel of a beta or an alpha of 8.5 or 9 – it feels very incomplete, with major operating system functions broken or replaced with inferior metaphors slapped on by someone with more experience with NeXTStep or X Windows than Mac OS. There are plenty of “demo features” that look cool at a keynote, but not core behaviors that make a Mac operating system Mac-like.

Mac OS 9 has an unprecedented level of flexibility for the professional user who encounters their Mac not just for an hour of surfing a day, but rather thinks of it as the primary environment they, quite literally, inhabit. With the control strip, the tear-off Applications menu (especially when customized with Prestissimo), the Apple menu with Apple Menu Options and contextual menus, there is a series of layers, all of which can be switched on or off to fit a pro user’s needs. All of these extensions to the core OS help the user to:

  1. manage the environment (CS)
  2. switch quickly and easily between applications
  3. navigate both frequently used items and extensive hierarchies from within any program, and
  4. perform operations on text and graphic objects within programs. Many professionals include QuicKeys and FinderPop for added layers of flexibility.

OS X reduces all of these to the dock. Moreover, it destroys the desktop metaphor in the Finder and again, undermines the user’s ability to choose how best to organize their data. Hint: a programmer, a casual user, and a graphics professional may take very different approaches toward organizing their data. If there has been criticism of the desktop in Mac OS 9 as being a “kitchen sink,” why then is the only means of quick access to both open and frequently used applications the Dock? This is nonsensical. Just as we dress in layers to keep warm, we need to dress the OS in layers to keep moving fast. To argue that third-party apps will take care of this is to introduce a nightmarish support scenario for the IT professional.

The loss of the control strip is doubly frustrating because of OS X’s lack of an on-the-fly location manager and its lack of a way of setting resolutions without rebooting. Hopefully these are issues that will be dealt with by the time of the consumer release. That they have not been dealt with so far is of great concern.

The loss of easy user access to the desktop is disconcerting for longtime Mac users. There is no excuse for an OS that, on boot up, can present you with a blank screen, the Dock and a menu. This is frightening and all too empty.

The primarily visual metaphor of the Dock is flawed as well, for similar reasons. Although an on-the-fly preview of graphics files is very attractive for graphics professionals, frequently they (and others) also have to work with textual data. How can one distinguish between two files written in Microsoft Word that are identical except that one begins “Dear Mr. Jobs” and the other begins “Dear Mr. Potatohead?” There is no visual cue possible. Now if both files were in the Apple menu hierarchy, it would be easy enough: the filenames “jobs” and “potatohead” would be different. Including the trash in a dock that constantly resizes makes it all too easy to dump a file into it by accident.

The centered Apple logo is gratuitous and wastes space. Most of us already have Apple badges staring at us from our machine’s hardware. Why add another? Moreover, it breaks the menu metaphor by introducing an illogical, nonfunctional element.

Moreover, too much emphasis has been placed on Quartz to do on-the-fly scaling at the expense of hinting. As it currently stands, the tear-off Applications menu and the Apple menu both have clearly legible icons at 16 x 16 size because there is a small icon resource developed for most Mac files. This is not the case in OS X, where a larger icon will be interpolated down. This, in our opinion, is a mistake. The icon, at small sizes, often becomes illegible. Having to pass the cursor over the icons to magnify them is an added step that slows down the user and makes the project into a high-stakes game of Hüsker Dü. Hinting at 16 x 16 and 32 x 32 sizes is a must.

So long as Apple continues making iBooks and Powerbooks, it should also realize that screen real estate is at a premium on these machines. A dock that makes a significant impact on portable screen real estate is a major problem for mobile users.

More attention needs to be given in following the Mac OS Human Interface Guidelines, or, if they are to be modified for X, explaining in clear and convincing detail how they are to be modified (as the Mac OS 8 HIG documentation did). Just providing layout guidelines doesn’t cut it. Apple has done no work in explaining how the core metaphors for the Mac OS will be moved forward or changed – we suppose as part of the secrecy behind Aqua. Hopefully, Apple will use WWDC to begin this process.

As a further word, we’ve found it much easier to understand where the Mac OS was going in Mac OS 8, 8.5 and 9.0 – their development teams have been open in communicating product direction and interface evolution, and they’ve performed admirably in delivering the most important product feature – a high-quality shipping product in a timely fashion (that is, when they said they would ship), without massive reorganization in development teams. From what we’ve heard, Mac OS X has not had this level of success. We’d encourage Apple to use the talents on the Mac OS development and SQA teams more extensively and listen to their contributions carefully. While Mac OS X has Unix underpinnings, it is a Macintosh operating system that’s being made here – not a fancy X Windows scheme, not a NeXT operating system with a Mac gloss that doesn’t get significant review beyond a few select people. While some may not consider widescale usability testing and design review beyond iCEO’s important, we do.

And going to the “gloss” that is Aqua – we sincerely hope that the UI is as customizable as it need to be. While “liquigels” look nice to that mythical “Joe Six-Pack”, they are a distraction to “Jane Graphics Professional”, who might feel much better editing with Photoshop in a less distracting environment. If Apple isn’t willing to give that as an option, there’s another reason Mac OS X might not be their cup of tea.

“Jane Graphics Professional” and “Jim Video Professional” are really Apple’s “enterprise” customers, Fortune interviews to the contrary or not – they are the only two toeholds Apple has left in many enterprises (even if those enterprises are small print shops). Ignore their workflow issues, change the operating system to make it “look cool” and be less usable in pursuit of market share, and watch those last bastions disintegrate. If Mac OS users have to retrain, they might well retrain on Windows 2000 Professional, no?

In MacWEEK, Stephen Somogyi recently wrote “It’s the drivers, stupid,” and he doesn’t know how right he is. It’s a phrase we coined months ago, while we were wondering why there was no IOKit. We’re in a position to be even more informed, and we’re already worried.

Apple recently made its IOKit, the counterpart to the Windows 2000 DDK, publicly available to developers. We point this out because Microsoft made their first version of the DDK available in 1998 – for a product released at about the same time as Mac OS X DP3. And what’s one of the major reasons for not having Windows 2000 as Microsoft’s main operating system? A lack of drivers, you say?

While we welcome the facts that USB printing is working in the OS and that Apple has gotten some “hired gun” software development labs to assist in driver development, and look forward to future progress, we’re concerned that 6-9 months is not enough lead time. Especially if the prototypical iMac customer Robert refers to as “Joe Six-Pack” brings his iMac DV 2001 with Mac OS X home and finds that his Agfa scanner and Sony iLink DV camera don’t work with it because the drivers are still under development. We won’t even begin to mention customers with investments in SCSI devices.

We hope that Apple takes seriously the critical feedback from developers and professional users, and uses it constructively to do whatever it takes to ship a Mac OS X that feels like an enhancement of the user experience people receive now, rather than replacing it. This would conclusively remedy a perceived lack of “listening” capabilities by Apple towards one of their most valuable assets: Mac users. Any company worth its salt has policies in place to take into account customer feedback and suggestions. Not only getting the feedback, but acknowledging concerns and providing solutions back to the customer.

Whether Apple’s responses involve making Mac OS 9.x teams more involved in the design and production cycle of X, or even delaying shipment for as long as is needed, similar to the decision to ship Carbon (a great example of listening to critical feedback and responding to it in a constructive manner), Apple needs to respond, or face the specter of defections. If people are going to have to be retrained for a new OS’s metaphors because the Mac OS is being made more “lickable” and less usable, it might as well be one with a much bigger market share.

E-mail this story to a friend