While cleaning out the “closed stacks” in the garage (boxes of books), I found my unread copy of Microserfs by Douglas Coupland and put it back on the in-house “to read” pile. Three weeks later, I’ve read it.

Short version: I haven’t learned anything from this book. I used “learn” in a pretty broad sense that includes any new experience, not just facts.

Long version

This book is about the West Coast coding culture, something that I was part of a decade before it was published (1995) and continue to be a decade after that date. Any journalistic content is not new information for me, so the book’s value to me is all in that creative remainder. Perchance I resemble an upper-class Regency woman reading Jane Austen. All that period fru fru is the reality I swim in (sigh, reduced to using Google to spell-check “fru fru”, so sad to vote on spelling).

I like journalistic work and really enjoyed both The Soul of a New Machine and Blue Sky Dream, so I’m ready to learn more about things that I already know.

Surprisingly, considering the title, that culture is only Microsoft for the first bit, after which a Deus ex Silicon Valley causes the crew to decamp to a startup and house in Palo Alto a few blocks off of my former commute to HP.

I fully understand that fiction is made up (objectively false, subjectively true), but when a realistic setting is a key part of the work, getting it wrong just isn’t an option. Coupland includes carefully crafted typos in the e-mails, so I know he was paying attention. I can just imagine the mail back and forth with the copy editors trying to get those typos published properly. Yet he didn’t do his homework on the simple things.

• Why isn’t the startup in a garage? Was that already too cliché? Google did it after this was published.
• Bug testers (his term, we call them “QA”) don’t immediately switch to being major hackers at a startup. Testing and coding are different skills and most people just like doing one better than the other. Even if you want to switch, you need to build your skills and your cred.
• Can Daniel please stop using “random” as if it means “unexpected” instead of “unpredictable”?
• It isn’t the “open-hills fire”, it is the “Oakland Hills fire”. Jeez.
• “Cal-Tec”? That sounds like a gasoline additive. It’s “Caltech”.

I’ll give him a bit of slack for those East Coast editors who can’t be bothered to care about computers or any place West of the Mississippi, but his name is on the book so it is a teeny-weeny bit of slack. [Re “East Coast editors”, ask me about a couple of howlers in Infinite Jest.]

I remember a comment from the introduction to Best Short Stories of the Year Whenever that quoted some famous short story writer saying that she stops reading if she finds a factual error. She felt that the writer has a responsibility to the reader to avoid those jarring moments, and if they couldn’t be bothered to do that, she couldn’t be bothered to continue reading.

Then there is the plot, which is mostly imposed, unmotivated events that increase in frequency toward the end of the book until we end with with a big fairy tale group hug. It reminded me of that baby programmer mistake where you stick to the initial spec even though you’ve run out of time and you start gluing on poorly-integrated barely-working features as the deadline approaches. That is the time to find the essence of your product and leave out anything that is peripheral. It is when the iron goes through the fire. It is Occam’s Chainsaw.

Oh yeah, another problem. Not much sense of impending deadline — the plot skips straight from beta to already having a distribution deal. Huh? The first half of the book keeps making a Big Deal of the Microsoft “Ship It” award, then he doesn’t bother to follow his characters as they ship their 1.0? That goes beyond ignorant to stupid. Every engineer in the valley can tell you exactly what they have shipped. Shipping is the essential act in engineering. It makes your work real.

In some sense, the novel is just an expanded version of a fine short story, published in Wired and used as the first chapter of this book. A common move and a very risky one. Short stories and novels are very different beasts, in my experience. When it doesn’t work, it is glaringly obvious. Two different examples: Flowers for Algernon is devastating read in thirty minutes but numbing when expanded to novel length, and you can stop reading Starship Troopers after that stunning first chapter with the powered combat suits since the rest alternates between “my life in the military” and libertarian ranting.

The original short story really is pretty good. Obviously, it was good enough to get a book deal, but it remains good reading. You can feel the rain and the green in Redmond and the tension between being a cog in the Microsoft machine and doing something you care about. Just stop reading before it switches to Silicon Valley.

I guess I have learned one thing from Microserfs. I’m not going to read any more Douglas Coupland.