The Macintosh Revisited
For
the last month and a half, I’ve been toting a nice 15” Apple PowerBook G4
around. The operating system is OS X Panther and it is truly impressive. In
fact, it’s so strong a winner that I’m wondering about the future of corporate
systems now. No longer is the Macintosh the weak little brother to the Intel
based PC like it was when I turned away from it the last time at the
introduction of Windows 95. When OS X first hit the streets on Apple’s PowerPC
platform, there was a strong hint that they were on to the winning idea and it
was time to take another, fresher look.
The
perpetual issues surrounding MacOS in those days had to do with the fact that
Apple just couldn’t seem to bring it forward well enough to even take proper
advantage of the original Motorola 680X0 processors the Macintosh line was
founded on (and which were, perhaps still are, so vastly superior to the Intel
processor line). The cooperatively multi-tasking OS was constantly the victim
of the applications it hosted. Stability is not something I recall ever
experiencing with a Macintosh in comparison with any other contemporary
operating system of its day.
When
Apple did make the huge leap to OS X, they did things in a big, technologically
successful way. Tying the classic Mac desktop paradigm to Unix was a huge
gamble but Apple accomplished something that still hasn’t been done well with
any other Unix variant by putting a reliable, intuitive and clean GUI on top.
CDE, KDE, Gnome, etc., don’t even begin to compare to Apple’s masterstroke.
The Experience
By
profession, I’m a Windows geek. My work involves the day-to-day administration
of corporate Windows servers, some Solaris units and a large body of Windows
and Macintosh users. For me, then, the acid test of using OS X is pretty basic
and extremely complex. For this system to do the job properly it doesn’t have
to work like Windows, but it does have to offer me tools that let me treat the
Windows Enterprise as king of the user environment in my workplace. In
addition, support for working with Unix systems must be reliable and not
require tools unique only to the OS X platform. On all these counts, the system
works exceptionally well.
When
thinking of the corporate desktop user, profound knowledge of the underlying
system must not be a requirement. All the normal activities of working with a
desktop computer must be accomplishable through application interfaces that are
consistent with each other. Application specific knowledge requirements must be
very application specific and not based on variations in the standard tools
like menus, shortcuts, etc. Knowledge of Unix must not be required.
The
traditional Apple user will be very satisfied here. While there are some things
that will challenge users of the older MacOS, the transition is light and
almost fun. If you’re a Quark user, I’m sorry to say you’ve lost your favorite
tool. As of this writing, Quark hasn’t been ‘carbonized’ (a term describing the
process of porting traditional MacOS applications to OS X). Comparing the
quality of OS X to that of Quark, I’d say if you were looking for a party to
blame for this problem, don’t look to Apple. They did their part and for good
reason.
Finally,
an optional consideration in evaluating the PowerBook is from the server
perspective. Peer-to-peer operation must be supported and, for the Enterprise,
hardware and software performance, redundancy and failover capabilities need to
exist. Not surprisingly, this is where Apple’s offerings fall short most
visibly.
How Apple May Screw This Up (History Lesson)
If
there are limits to imagination, Apple really hasn’t shown it here. Some bad
old habits are still chasing them around. See, their job is actually much
easier than nearly any other big manufacturer in the personal computer
industry. They still control the hardware AND the operating system. They still
largely control what happens in the Macintosh developer community, too.
It’s
amusing that it is precisely this level of control which has commonly provided
the steepest hurdle presented to Apple’s sales efforts. If the customers have
limited choices for software and hardware on a platform, they don’t tend to
risk a couple thousand dollars on that system. The great success story of the
Intel platform, after all, has been the ability to customize a system with
products from a wide variety of vendors for a wide variety of purposes. Even
the idea of upgrading system components is almost totally unique to Intel based
computers. Apple has never fully embraced this idea although they have taken
minor steps in this direction over the years.
The
idea has always seemed to bear some sort of threat to Apple, too; and it
continues to haunt their products as they attempt to build systems to meet
needs on an enterprise scale. For instance, you may try to connect an Emulex
host bus adaptor in your X-Serve as you try to connect it to an EMC Symmetrix
SAN and you may succeed. However, when you look for support on this move, you
won’t get it from either EMC or Apple. Since both companies are hopelessly
married to the idea of ‘owning the technology’, neither has been able to find
its way to supporting such a solution and this isn’t the only example.
Apple
has also shown a lack of understanding where the idea of security updating is
concerned. Even Microsoft will continue patch support for Windows NT 4 through
the end of 2004, an operating system in use since 1996 (8 years old, kids!!)
Recently, Apple introduced some security updates and limited the support to
their newest of OS offerings. All users of previous OS X versions were left
with only the option to upgrade their operating systems (at their own expense).
Whether this assault on the customers was Apple’s intent or not, the fact
remained that it wasn’t a savvy move and resentment from the customer base was
deserved. I’m confident that Apple will someday understand that customers are
valuable, not sources of steady revenues no matter the practices employed and
that immunity to security exploits is no more an ingredient of Unix based
systems than any other offering. The Mac is hackable. Denying it is the most
extreme ignorance.
Why Microsoft Ought to Worry
People
often confuse me…okay, we’ll accept this as truth and one of my own
shortcomings. But more to the point they also commonly confuse me with being a
diehard Microsoft fan. I can accept that except for those occasions in which I
have retreated to the most complete level of objectivity I can muster. Now is
just such a time. Call me an OS bigot if you must because this system is a
viable threat to Windows everywhere. If Apple takes a look back and learns from
its historical blunders in the market place they’ll be able to leverage this
system nicely against Microsoft’s high value market; the corporate desktop.
In
this single development decision, Apple stemmed the flow of cash into a proposition
which, year after year, failed and simultaneously delivered the best desktop
solution since the PC form factor first invaded the workspace.
Knowing
that Bill Gates and Steve Ballmer have impeccable records for success and that
both tend to make the most of their intelligence, I find myself amused by a
perverse vision of Bill in his role of Chief Architect strolling into Steve’s
office one day, filled with nervous excitement and showing only a little
through an outwardly calm demeanor.
“Hey
Steve, got a minute?”
“Sure,
Bill. Got an idea for our next video?”
“No
Steve. Something slightly weirder is on my mind.”
“Oh
yeah? What would that be?”
“Well,
you know how we are always getting hammered by the community for poor
security and bugs and how every Clearasil candidate in the world turns
our customer systems into the kind of mindless, attacking creatures seen
in Night of the Living Dead?
“Bill,
you haven’t let me forget that for at least 2 years. Yeah, I know what you’re
talking about.”
“And
you know how Jobs put a poker in our eye by willing Apple to live in defiance
ofNewton’s First Law?”
“You
mean OS X?”
“Yeah!
So, I was thinking; what if we took Explorer and all the other GUI interfaces,
ripped it away from the NT platform and put it on, say, System V? Hell, if Jobs
can do it with Darwin, you know we can, right? Suddenly, many of
the things we’ve always said would be true. I mean, our security wouldn’t be
any worse than anyone else’s and we’d be adopting a method for which we’ve
already seen measurable success at someone else’s expense, our forte! We’d
prove our point and save money at the same time! We already know we can
force the application providers to move with us. I mean, you know, what
if, Monkey Boy?”
“Bill,
I’ve told you not to call me that. But hey!! Didn’t we throw a chunk of change
at SCO last year? Maybe we could actually use that license!!”
Then
I wake up. I haven’t the imagination to take that story any further, but I bet
Bill and Steve do…and if they do it, I get to claim I thought of it before they
did.
Why the Biggest Loser is the Linux Desktop
By
now, you’d have to have a real life to have not heard of Linux. That means you
couldn’t possibly have anything to do with computers and if you’re still
reading this article and believe you have your own life, you’re only fooling
yourself.
Further,
if you think any of the Linux offerings are strong contenders for the corporate
desktop, you’ve probably forgotten that most of those folks only need to USE a
computer in performing their jobs. Throwing away their CPA licenses in favor of
joining the uber-geeken isn’t really their gig at all. Yes, installing Red Hat
Linux is almost trivial now. Learning to work with it isn’t. The application
set is dodgy, it’s still not unusual to load Linux on a system and have to
manually start the GUI because, for some obscure reason, it won’t start on it’s
own. And the command to start the GUI is not described in a query to system
help…especially if the boot process leaves you looking at a flashing cursor.
In
very few of the available GUIs adapted to Linux have the Unix filesystems been
sufficiently abstracted for use by LCD (Least Common Denominator) customers.
Users are still presented with /, /opt, /etc, etc., and the entire concept
beats on the minds of new Linux users.
In
almost none of the Linux distribution will a new user stumble into a memorable
and easy method for changing their screen resolution. If that user does stumble
across the method for changing the res, the experience, not the precise method,
will remain in memory because nothing about it would be a solution you might
arrive at using tuition or familiarity with any other system.
Inter-application
functions like the Clipboard are also very weak in these shells. Some
applications just simply can’t hit the clipboard and get that screenshot pasted
into the current document.
Linux
GUI solutions are just a really raw place to work compared to OS X. In
recognition of this truth (good for them to have the clarity of vision to spot
it) several of the bigger players in the Linux market have changed their focus
to theEnterprise server and have left consideration of the
desktop user behind with the barest of apologies. That’s too bad, really. Those
vendors wind up being another Sun or SGI without really offering any savings to
the corporate world. They do, after all, have monthly kernel security updates
to offer (read recompile) and Linux is aimed at the processor with amongst the
poorest of security architectures on the market, Intel. As a systems engineer,
I’d have to be somewhat sadistic to suggest that a Linux on Intel solution will
save the company money compared to a Sun chassis. I can’t get real OOB access
to a Linux on Intel solution, for instance, so the real savings come at the
expense of the most expensive of corporate assets – the systems engineers.
Admittedly, though, this assessment might be too general in scope because the
real point of loss in this picture is the dream of the Linux Desktop withering
due to Apple’s better solution.
The Inescapable (in My Opinion) Conclusion
Apple
has almost accidentally solved a real problem faced by so many corporations
over the years that there is relatively little they would have to do to steal
that market completely. Coupling their GUI to a mildly modified Unix operating
system has nearly eliminated the great problem of system interoperability for
enterprises basing their core services on Unix systems. They’ve done it so well
that the end user doesn’t have to invest in a complete re-education. The
hardware is almost competitively priced, the application set is fairly well
rounded out, improving and, of course, the entire system is pretty solid.
For
once, a Unix environment exists which offers something previously available
only to Windows enterprises – a common platform between users and systems
engineers - the Apple of their i(s).
|
