Skip to Content
Skip to Table of Contents

← Previous Article Next Article →

ATPM 6.02
February 2000



How To



Download ATPM 6.02

Choose a format:

Segments: Slices From the Macintosh Life

by Paul Fatula,

Welcome to Bloat OS

When I first installed my shiny new copy of Mac OS 9, I immediately noticed two things: scrolling through large windows in the Finder was noticeably slower, and applications opened up a little faster. The former, though it sounds bad, is in many cases a good thing: with the amazing speed of G3s and G4s, it’s easy to fly past what you’re looking for in the time it takes to click and release your mouse button.

Take a moment to think about that; it’s pretty amazing. The hardware is so fast that the term “too fast” actually becomes, in this instance, not just possible, but a fact of life. And the hardware is—by all clock speed, operations-per-second, and other such measurements—far faster than the machines we were using just a few years ago. Moving from a Centris 650 to a G3/400 gave me an eightfold increase in clock speed alone, not even considering the advantages of the G3 chip over a 68040, increased bus speed, a faster hard drive, graphics acceleration, ten times as much RAM, or a level 2 cache.

OK, enough awestruck praise. Because my cynical mind is starting to how come my G3 isn’t even twice as fast as the Centris? For all the speed increases in the hardware, applications take longer to open, system startup time is still abysmal, the time it takes between hitting command-F and getting a Find window has actually increased significantly, and a machine that is supposedly capable of hundreds of thousands of calculations per second still takes over five seconds to calculate the size of my system folder.

In the case of applications, poor opening speed can’t be blamed entirely, or even mostly, on the OS. The problem is in the applications themselves: Word 98, for example, takes significantly longer to open than good ol’, does-what-you-want Word 5.1a. The way Microsoft decided to go in creating successive versions was to pack in more features, rather than optimize what they already had. And, well, I dunno about you, but there’s nothing Word 98 offers that 5.1 doesn’t that I actually want, need, or, frankly, even like. What I need is compatibility with what everyone else uses...and if everyone else upgrades to 98, I’m stuck with it too...unless I’m clever enough to grab a copy of MacLink. (An awful program that performs a vital task with some to many errors, written by a company even less responsive to user feedback than FileMaker, and marred even further by a user interface that makes me wish for the ease-of-use of Windows NT...but it has no competition, and is moderately better than opening the file with BBEdit and zapping gremlins...) So people run to the stores to upgrade, spending money for bloat: features they don’t particularly need or want, at the expense of speed, which is to say, usability. In effect, software writers are filling up the time saved by faster processors by bloating up their software, so that in spite of huge speed increases in the hardware, the software, by and large, still crawls.

Sound familiar? I’m thinking about our beloved Mac OS now...because Apple is doing the same thing Microsoft, and most other software companies, do...they keep time (startup, program opening, finding) essentially constant across processors with tenfold speed differences. They are so excited about new features, that they give no thought to optimizing old features, and it never occurs to anyone to ask whether the new features are worth their cost in speed. A computer today that does things at the same speed as a machine five years old isn’t breaking even; it’s taking a huge loss.

Apple loves Sherlock. And I have to admit, once I run all the (third-party, definitely not carrying the Apple seal of approval) patches to get rid of the non-standard UI and the flashy, annoying ad banners (you do realize you just paid $99 for the privilege of having a blinking billboard in your very own home, don’t you?), I like Sherlock too. It doesn’t offer anything particularly new: its Web searches are available from portals like, and have been since before Sherlock was a glimmer in Steve’s eye. The people-finding features go way back to the days of gopher. BBEdit (and, no doubt, some of the full-fledged word processors) has long had the ability to find by content, searching your drive for a snippet of text. CNET Shopper gives you shopping price comparisons. What Sherlock does is integrate these various functions into one convenient place. A really great, “Why-didn’t-anyone-think-of-this-before?” sort of idea. It performs all of its functions well (as well as, or even better than, the various individual programs/functions it combines), but not without cost. That cost is that it now takes significantly longer to just open up the Find dialog when you just want to look for a file on your hard drive.

Perhaps it’s worth the price to you. But I, ever cynical, say Apple should have done better. In any of several possible ways:

The third of those suggestions is the one you, dear reader(s, I hope!), will probably like best. because it would require no sacrifice of anything you currently know and love about Sherlock. All it took for me to come up with it was a little common sense and a desire to optimize a piece of software rather than merely expand it. Back in the day of 5 1/4" floppy drives in computers with no hard drives and less RAM than my Casio digital wristwatch (really), programmers had to make every tiny bit of every program they wrote as resource-efficient as possible. The programming game wasn’t about adding features; it was about getting the task the program was written for done, as quickly and efficiently and painlessly as possible.

(And yes, ladies and gents, I know that’s the origin of the Y2K bugs the media was screaming about pretty much constantly in the latter half of last year...programmers decided it wasn’t worth the resources (like storage space) to hold all those 19’s before the years. Back in ’60 and ’70 it was a good call. But at some point people stopped thinking about whether it was a good idea and just kept doing what they had always done. A classic example of tradition trumping common sense.)

Apple, like Microsoft and most (though not all, God bless Nisus) other software companies, have long since stopped asking “archaic” questions like “is it worth it?” because as far as they are concerned, the answer is, “Computers are Fast. RAM is Cheap. Hard drive space is Cheap,” perhaps followed by “Features sell upgrades.” But long lost are words of wisdom passed down from grandfather to grandson, “If it’s worth doing, it’s worth doing right.”

Apple and friends will say they are doing it right. And they’ll point to two things: features and stability. If they talk about speed at all, they’ll be touting the potential of the G4, and hiding the fact that their OS is keeping the hardware from living up to its potential by unnecessarily overburdening it with nonessentials (or even niceties, like Sherlock, QuickTime, etc.). (You can do a “minimum install” of the OS, and it’ll come in at around 13 MB. That’s a lot more than the days of emergency 3.5" startup disks, but it’s a sharp improvement over 160 MB of the recommended install. What’s in that extra 140 MB or so, and did anyone at Apple ever ask if it’s worth it, or if the same job could be done in 120 MB?) To me, doing it right means taking the existing product and making it more sleek, more usable, faster...before adding new features. That’s how you avoid what I call bloat...asking “How do we make what we have better?” rather than just “What can we add to what we already have?”

Apple has added a lot to the Mac OS, but there’s been pitifully little improvement to some of the most basic functions of the OS that have been around since System 7...calculating folder sizes, finding a file on your hard drive, starting up your computer, emptying the trash, starting up programs...none of these has improved much in the last several major releases, and several have actually spite of significantly faster hardware. (If you question...How much improvement is possible here?...check out BeOS. On a Power Mac 9500, it is literally jaw-droppingly fast, beating a G3 with 8.6 or 9.0 head-to-head, hands down, at tasks like emptying trash, calculating folder sizes, and performing finds.) Apple has simply stopped concerning itself with mundane things like performance, favouring the flashier and more advertisement-friendly New Amazing Features.

All of the recent additions to Mac OS have proven that Apple can and does Think Different. What I’d really like to see, however, is an Apple that can Think Better.

Copyright © 2000 Paul Fatula, Send your Segments submissions to

Also in This Series

Reader Comments (2)

Albert Rogers · April 13, 2000 - 01:01 EST #1
Couldn't agree more! But perhaps I can offer a long perspective, to show how right you are! Back in the days when a mainframe didn't expect to have as much as 16 Megabytes, IBM (God bless 'em) hadn't yet hired the fool that designed the "PC", and even then the OS had memory protection and used it, and if you tried to read or write in RAM (we called it "core") that didn't belong to you, your program learned about it immediately. And you fixed it or else. Perhaps we need "open source" so that programmers will be ashamed to write such sloppy code as these bloated apps must be. Then again, even CodeWarrior doesn't do much to encourage you not to bundle everything but the kitchen sink in with your final linkage. One final gripe: what does the Appearance Manager do that's worth all that complication?
Michael Tsai (ATPM Staff) · April 13, 2000 - 01:01 EST #2
The Appearance Manager provides a standard Apple-supported way for programmers to incorporate user interface widgets that are now standard, but were not part of the original Mac toolbox. Apps that use the Appearance Manager will automatically inherit the Aqua look when run on Mac OS X. They should also be less bloated because there's no need for the app to include its own code for drawing buttons, lists, etc. As you say, this is a common problem for CodeWarrior programmers using PowerPlant.

Add A Comment

 E-mail me new comments on this article