Vista and the Software Gap

by

Ross M. Miller
Miller Risk Advisors
www.millerrisk.com
October 13, 2008

My introduction to Windows Vista was sudden. This past May, the morning after I had graded my last spring semester exam, I sat down in front of my Sony desktop computer with my morning coffee. At first, I thought that I was still dreaming. Just like in the movies, streams of multi-color letters and numbers were streaming down my main monitor. I spent the next day-and-a-half rationally going through all the possibilities, replacing things one component at a time, until I decided that it was time to give up and buy a new computer. I had been thinking that the Sony, a mere three years old, was about due for replacement anyway because it was just too slow.

After the most minimal deliberation, I settled on a Gateway desktop from the dreaded Best Buy. For the price of the most expensive Mac Mini (with neither keyboard nor mouse), I got a machine with a quad core Intel processor that could handle two monitors out of the box. It came with 4GB memory and 64-bit Vista that could use it all. It also had 32-bit Vista in the box in case the 64-bit version did not work out and I didn't miss the 0.5GB of memory that it could not access. Being intrepid, I spent the next three-and-a-half days getting the machine as close to my old configuration as humanly possible. I had some compatibility issues—a few minor programs and a printer that was due to be trashed that would simply not run under 64-bit Vista without major effort.

I was initially quite happy with my new machine and wondered why Vista had gotten so much bad press. Yes, it was not perfectly backward compatible and the control settings were randomly scattered about, but it zipped right along on a new, loaded machine. After the initial setup, I rarely saw the dreaded UAC screens, and when I did there was usually a good reason. (The machine did include Service Pack 1, which supposed reduced the number of UAC screens.) The system was vastly more secure than XP and many things worked better. The system still had sleep-related issues—put any Windows machine to sleep and it never is quite the same when it wakes up.

It would take a month until I understood why Vista is so unloved among the computer cognoscenti. It was then that I first experienced the dreaded "Display driver atikmdag stopped responding and has successfully recovered" message. On good days, this is all that happens and (after saving everything), I can continue computing as if nothing had happening. On bad days, the message repeats five times and the machine stops working. As the message indicates, my machine has an ATI graphics card, but a similar message appears on machine with Nvidia cards.

There is a vast literature of web postings on this problem with ATI and NVidia drivers under Vista. Microsoft, ATI, and Nvidia are all very much aware of this problem. As I write this, no one knows definitely what causes it and while there are a slew of home remedies that work for some people and not for others, there is as of yet no cure. The common element seems to be that pushing a Vista-running machine past its comfort zone causes this problem. One common solution is to disable the Aero interface, a major selling point of Vista. I did not go that far, I merely disabled the taskbar preview feature, and the problem went away, only to return recently when I was playing with the audio-processing features of Winamp in a way that pegged all four cores of my CPU. (Using a program called SpeedFan, which monitors the temperature of each core and hard drive separately, I had been able to eliminate overheating as a cause of the problem.)

It is clear that there is some fundamental flaw in Vista, possibility involving the timing of the communication between the CPU and the graphics board(s), lurks behind this problem. Vista has been available for nearly two years and the problem remains. An operating system that has been out this long should not continue to have such a major unresolved bug.

Microsoft has a real problem here with Vista, and no amount of commercials with Jerry Seinfeld and Bill Gates moving in with normal families is going to solve it. I switch to Linux and/or Mac OS whatever in a minute, but I am to some degree locked into Windows and am not a big fan of either operating system. Indeed, as I write this, iTunes (an Apple product) is running some background processes (Apple Mobile Device Process and Apple Mobile Device Server) on my computer that are undoubtedly part of Apple's plot for world domination.

The real problem is the "software gap." Hardware is literally tens of thousands of times more powerful that when I first started messing around with an early AT&T release of Unix back in the mid 1970s. Linux and the Mac OS are expanded versions of that early Unix and Windows is essentially a direct knock-off of Unix. Given all these operating systems a very generous benefit of the doubt, they are at most one hundred times more capable than AT&T Unix. At a purely conceptual level, there is nothing new at all—the ghost of the original Unix looms large within all three major operating systems. So hardware is zooming along, getting better by the day, while the software that sits directly on top of the hardware is relatively standing still. In turn, this affects the applications that run on these operating systems.

The problem is that developing software much more difficult than developing hardware. At the heart of this difficulty is that pushing around electrons is a lot easier that pushing around programmers. Introduce corporate bureaucracy into the process and you are virtually guaranteed to get software that sucks. Exhibit 1 is Vista itself.

I might feel more positively about software if I played computer games. These games do seem to be where the bulk of the creative software development is going on. From what little I have seen of games on my rare visits to BestBuy, the typical contemporary video game looks really hokey. Better than Ms. PacMan perhaps, but still hokey.

I do, however, own a piece of software that by the standards of the 1970s is truly wondrous. That program is Reason 4.0. This program is put out by a Swedish company fancifully named Propellerhead. Reason is a virtual electronic music studio with an amazing array of synthesizers and related electronic instrumentation, mixers, effects, etc. The combination of Reason with a USB controller keyboard, a good 24-bit/96KHz sound card, and even a modest computer creates a music development environment that is simply amazing. During what Reason does in software on specialized hardware would cost hundreds of thousands of dollars and that sucks down amp after amp of electricity. A barebones system (computer included) that runs Reason adequately under Windows can be had for under $1,000. A deluxe system under either Windows or Mac OS is available for about twice that.

Reason and other similar music creation programs have partially avoided the software gap because the underlying technology, digital signal processing (DSP), is one where advances in hardware and software have developed nicely in tandem. Also, DSP can use so much CPU power (as I demonstrated with my Winamp experiment), that hardware remains the constraint in a lot of professional work.

Nonetheless, a software gap remains for most music creation software. Reason's user interface is hamstrung by having to run on top of either windows or Mac OS. Also, despite the growing availability of 5.1 and 7.1 surround sound as a standard feature in PCs and Macs, Reason and its ilk exclusively generate plain old stereo sound. While music-generating devices with no analog in the physical world have started to show up in music creation programs (Reason 4.0 has a super-synth called Thor), everyone that I've seen is merely a direct extension or hybrid of physical world instruments. One problem is that this market is driven by professionals and until recently these professionals were working in the physical world before they moved over to PCs and Macs. Once everyone forgets why synthesizer configurations are called "patches", this is likely to change.

The software gap is most noticeable in Microsoft's Office applications. Other than a string of cosmetic differences, little has changed in Word, Excel, and PowerPoint from Office 97 through Office 2007. While Microsoft has spent billions of dollars in artificial intelligence research over those ten years, the grammar checking in Word is just as dumb as ever. PowerPoint has at least evolved from truly pitiful to somewhat on a par with the other Office applications. Excel still refuses to do symbolic algebra, something that Mathematica, MathCad, and similar programs have been doing for over a decade and even some TI calculators were offering as early as 1995. OneNote is a nice addition to Office, but it is still rather pitiful, as is the general support of tablet computing by Microsoft. (It is, however, much better than what Apple offers in tablet computing, which is absolutely nothing.)

Next month is my last column of the year before my usual two-month sabbatical from this commentary. Given the rather upset nature of the world at the moment, I will leave the title of that commentary as "Whatever."

Copyright 2008 by Miller Risk Advisors. Permission granted to forward by electronic means and to excerpt or broadcast 250 words or less provided a citation is made to www.millerrisk.com.