Brian's Brain

Tuesday's post on emulation has an underling personal story with even bigger-picture implications than the specifics of the particular interaction I'm about to tell you about. Late last week, I received a fairly critical email that began with an eye-catching and eyebrow-raising 'Bait and Switch?' subject line. The reader (who shall remain nameless) was seemingly quite disappointed because my recent cover story talked about virtualization instead of covering hardware (specifically CPU instruction set) emulation, which he apparently was personally interested in. He was frustrated in spite of the fact that my article's title began with the word virtualization, not emulation.

To be fair, the first draft of my article included a sidebar in which I'd explicitly directed readers to the Brian's Brain blog where they'd find Tuesday's addendum post on emulation (among others); it got cut at the last minute due to limited available pagecount. Perhaps, had it survived the edit process and had he seen it, his disappointment would have been lessened. Then again, though, he admitted that he didn't make it past the first paragraph of my article, so...

As I shook my head with bemusement (and other emotions) at this particular reader's narrowly focused agenda, I thought back to one of the top sessions I attended two weeks ago at the Audio Engineering Society Convention. Entitled 'Engineering Mistakes We Have Made In Audio', it was a panel session staffed by several well-known hardware and software technical experts (including widely regarded 'father of lossy audio compression' James 'JJ' Johnson, who lit up the room with his #1 mistake, an apology for "what the music industry's done with my baby").

Many of the mistakes, some made by the panelists themselves and others made by co-workers and those advised by the panelists on a contract basis) were predictable, such as:

• Power supplies connected to systems with cables' polarities reversed
• Data buses pin-ordered in incrementing order on one board, and in decrementing order on another board (with a third board, also plugged into the backplane, quickly designed at the last minute with the sole function of swapping data lines around and thereby working around the defect), and
• 'Spaghetti' code created in an unstructured, uncommented and incomplete fashion by a single person, who subsequently and abruptly left the project...eventually resulting in the code needing to be re-created from scratch since nobody else could figure out the code that already existed (thereby also exemplifying the bigger-picture problem with putting too much non-overlapping responsibility in any one team member's hands)

But I was gratified to see that many of the topics discussed by the panelists were also more psychological in nature. Issues raised both in the presentations and subsequent Q&A included:

• Entering into the definition and design processes with predetermined and rigid ideas of the end product's characteristics
• Similar rigidity when debugging problems, leading to...
• Picking and choosing experiments (subconsciously or not) that support the predetermined believed root cause, and
• Similarly sorting through the experiments' results, elevating data that supports the predetermined root cause and suppressing data that doesn't

There was, for example, the situation where the computer hard drive failed and the presenter sequentially popped in (and ruined) all of the backup tapes before he discerned that the tape drive, not the tapes, had also gone bad. Similarly, another presenter fried an entire set of add-in cards by not realizing (until it was too late) that the backplane connector hook-up, not the add-in card design, was at fault.

I encounter examples of narrow, rigid perspectives and pre-determined conclusions all the time, in the private and personal feedback I receive to my online and print writeups. VUDU owners, desperate to justify their hardware investments, ignore the hard data that indicates that the company is making inappropriate use of their LAN and WAN bandwidth for its own benefit. Back in the HD DVD vs Blu-ray format war days, equipment and content owners in each camp similarly fought with me and with each other, angry that I wouldn't agree that their particular wallet-anointed preference was all-good (and that the opposing camp's option was all-bad). Apple and Linux backers think I'm a Microsoft mouthpiece. AMD and NVIDIA fans respond the same way whenever I say anything remotely positive about Intel. Nobody, it seems, roots for the market leader.

For grins, as additional evidence of the point I'm striving to make here, I scanned through the last week's worth of comments left on the Brian's Brain blog:

• Ho-Hum claims that I've said Intel integrated graphics are adequate for any games, when my writeup clearly says that I feel they're sufficient for entry-level games.
• springmonkey (same link as above) accuses me of widespread Intel bias because he found integrated graphics insufficient for his particular needs. He doesn't, of course, say what those needs are...nor what his hardware budget is.
• Tyler and VOD download are loyal (translation: biased) VUDU backers. Fortunately, Ray Von had my back.
• plugh called me 'lazy' and accused me of trying out for the NY Times (whatever that means) because according to him the Nike+ RF protocol isn't Wibree. But I never said the Nike+ RF protocol was Wibree..

I could keep going, but I think you get my drift. However, regarding the 'point I'm striving to make here', I need to remain diligent regarding the possibility of narrow, rigid perspectives and pre-determined conclusions creeping into my psyche, too. Unlike some "journalists", I don't believe that being 'balanced' means presenting an exactly equivalent number of pros and cons for each side of an issue. I do the research, in as undistorted a fashion as I can muster, and then I present the data and my conclusion. Because I'm a human being, not a computer (or, for that matter, a Vulcan), I realize that my approach can't be completely distortion-free (which is why I'm an advocate of network neutrality, enabling access to a diversity of perspectives by a thereby informed populace). But what I can commit to is flexibility; as new data emerges, I strive to revisit (and if necessary alter) my prior conclusions.

I'll have more to say on this topic next week. Until then, I welcome your thoughts, either tech-specific or more general. Happy weekend, everyone!