The Daily Parker

Weather Now 3.5 is now the official, public version of my 9½-year-old demo. I first launched the site in September 1999 as a scripted ASP application, and last deployed a major update (version 3.0) on 1 January 2007.

As threatened promised, I'll have a lot more to say about it in the next few days. But I should address the first obvious question, "Why does it look almost identical to the previous version?" Simply: because my primary goal for this release was to duplicate every feature of the existing application, without adding new features unless absolutely required. It also had to run on the existing databases. That's why this version is 3.5, not 4.0 (which I hope to finish in early 2010).

I couldn't avoid some user interface (UI) differences, mainly because I used better design techniques than in the last release. And just as a matter of course, as I re-wrote each UI feature, I corrected or obviated numerous defects along the way. That said, version 3.5 has all of the features that 3.1 had, and any URLs that worked in 3.1 will work in 3.5.

I invite everyone to play with the application, and let me know about any defects or hiccups you discover. I think you'll find that it's an improvement over the last version.

This passage from Almost Perfect, Pete Peterson's autobiography of his days at WordPerfect Corp., inspired me to get out of bed, walk to my computer, and post a blog entry:

We on the Board had no one to blame for the delays but ourselves. The project directors we had chosen were inexperienced managers, and they made the mistakes inexperienced managers make. They were prone to overly optimistic forecasts and had trouble chewing people out when they missed their deadlines. Another of our mistakes was that we waited too long to add new programmers to the project....

And here is the context of that passage, which Peterson, without irony or self-awareness, set up only two paragraphs earlier:

I was not entirely honest in making the admission [that our release date had slipped]. Rather than go with a realistic date or a vague date or no date at all, I announced a hoped-for second quarter release, which was the most optimistic date from our most optimistic developer.

Yes, the Board had no one to blame...but they blamed the managers and developers. Yes, the managers had trouble chewing people out...for missing deadlines the programmers thought impossible and never agreed to. Yes, the programmers came up with a range of estimates...which turned out to include the actual ship date. And of course, if you want to foster openness and communication, the best way to do that has to include, without exception, ignoring the people doing the work, exhorting their managers to chew them out, and setting wildly unrealistic requirements in the first place.

WordPerfect Corporation had the best word processor on the planet in 1990, but somehow could never grow beyond themselves. Almost Perfect should serve as a cautionary tale to every entrepreneur, everywhere, to get out of the way of their own creations, lest they hang on and watch their babies die.

The third way, of which I heartily approve, is to eschew growth entirely. If done honestly and with full acceptance of the consequences, an entrepreneur can live a long and happy life running a business out of his living room. But having decided to grow beyond that point, the entrepreneur must necessarily give up total control of his organization in exchange for partial control over something orders of magnitude larger. One can be king, or one can be rich, but one can almost never have both.

The new Weather Now demo is feature-complete, meaning it has all of the pieces required for release. I will push it out to production, replacing the current demo, tomorrow morning, after I make some configuration changes to the web server it's going on. But because you read this blog, you've got a sneak preview.

Over the next few days I'll be writing about the demo, why it's completely new even though it looks an awful lot like the old version, and what I'll be doing in the next few months to improve it.

I haven't finished all of Almost Perfect yet, but I think I understand now what happened to WordPerfect Corp.: they had accidental success, naïvely thought they authored the success, and never thought strategically.

Now, possibly, I'm imputing Pete Peterson's own failures to the entire company, but I have to assume the other board members condoned his approach or they wouldn't have kept him on for so long. Peterson himself seems hopelessly without self-awareness, stumbling from decision to decision without a thought to the implications of each and without any coherent plan for how they all fit together. He is, in the Myers-Briggs jargon, an off-the-charts Thinker making him almost indifferent to other people's feelings, with an added load of Sensing making him detail-driven with a disdain for abstractions. For example:

For practically every week from December of 1989 through mid-1990, I invited 16 different managers to have lunch with me for three consecutive days starting on Tuesday. After lunch each day, I spoke for about an hour and a half. ...

Near the end of the first lecture, I explained what WordPerfect Corporation was not. This set the stage for the next two days, when I would explain what WordPerfect Corporation was.

WordPerfect Corporation was not a platform for personal achievement, a career ladder to other opportunities, or a challenging opportunity for personal improvement. The company did not put the needs of the individual ahead of its own. The company was not concerned about an employee's personal feelings, except as they related to the company's well-being.

WordPerfect Corporation was not intended to be a social club for the unproductive. While other companies might condone many personal or social activities at the office, ours did not. Things like celebrating birthdays, throwing baby showers, collecting for gifts, selling Tupperware or Avon, managing sports tournaments, running betting pools, calling home to keep a romance alive or hand out chores to the children, gossiping or flirting with co-workers, getting a haircut, going to a medical or dental appointment, running to the cafeteria for a snack, coming in a little late or leaving a little early, taking Friday afternoon off, and griping about working conditions were all inappropriate when done on company time.

In other words, how employees did things was more important than other concerns that one might prefer to use as yardsticks for employee value: productivity, creativity, well being, cameraderie. All right, as my dad says, "it's their football," and if you work for them you play by their rules. But contrast that passage with this one, from the same chapter:

[A requirement] was that we needed to communicate freely and frequently. In many companies it was common for supervisors to keep information to themselves, conceal their mistakes whenever possible, and never allow subordinates to go over their heads. I wanted a company where information could flow freely without regard to formal lines of communication. I imagined a room filled with light, without any portion remaining in darkness. I wanted a company where no one kept secrets and where everything was kept out in the open. Advisors who expected their employees to be so loyal to them that they would not take problems to someone else were exactly the ones I wanted to kick out of the company. Any loyalty should be toward the company, its purpose and objectives, not to individual advisors. Advisors who did not want the light to shine in their domain did not deserve their positions. Lines of communication should be allowed to go in any direction. If employees made mistakes, then the mistakes needed to be known so they could be corrected and avoided in the future. I wanted a company where employees could make mistakes, admit them freely, and learn how to do better without fearing for their jobs.

That's great if every person in the company is exactly like him: heads down, ISTJ, work is for working and home is for everything else. But to most people, in my experience, generating that kind of atmosphere requires flexibility around the periphery. In other words, if people have the freedom to work in the specific ways that make sense, and if they're judged by their output rather than how they achieve the output, they're usually happier. The occasional office baby shower might "cost" two hours of "company time," but it pays back disproportionately in productivity if it helps people feel happy about coming to work.

More practically, though, Peterson demanded loyalty towards the company while at the same time telling people they have no way to advance within the company because of its flat management structure. What he fails to understand, despite getting close to seeing the patterns while remaining maddeningly oblivious to them, is that in an organization the size of WordPerfect Corp., politics happens. It's an emergent phenomenon of groups. By "groups" I mean in a sociological sense: any aggregations of two or more people.

Throughout the book Peterson shows these amazing blind spots, often on the same theme. He wants people to openly speak their minds to management, yet "[f]or years I was the person most feared in the company. If I walked down a hallway, I was used to hearing the sound of desk drawers closing as people hid their snacks from view." The mind boggles.

Not too long after the events described above, the employees of the company presented a petition to the other two board members to have Peterson fired—and they fired him. Shortly after they sold out to Novell, and WordPerfect became irrelevant.

More later.

Quick update:

• The Titanic dinner at Mint Julep Bistro was wonderful. Rich's wine pairings especially rocked—as did his beef tournedos in port reduction. Mmm. Not so much fun was Metra's return schedule (featuring a 3-hour gap between 21:25 and 0:35), nor my reading of it (I did not remember this three-hour gap). The fine for taking public transit out to the suburbs (because driving to a 10-course, 9-wine-plus-apertif dinner seemed irresponsible) was $80, paid to the All-Star Taxi Service.
• I did, in fact, buy a Kindle, and I love it. I've now read three books on it and numerous articles (converting a .pdf or text file costs no more than 10c for automatic downloads), and I hardly notice the machine. It only holds 1.5 GB of stuff, but the complete works of Shakespeare ($4) only takes up 4 MB so space is not exactly at a premium.
• I may have a new release of Weather Now out today; if not, then tomorrow morning. I'll be writing over the next few days more about what's different, and why it took nearly two years to produce something that, to some, will look almost identical.
• Tangentially about my Kindle and software releases, I'm now reading Almost Perfect (hat tip Coding Horror), Pete Peterson's account of the rise and fall of WordPerfect. It's a fascinating tale of what happens when everyone in the company is just like you, and when entrepreneurs can't let go.

Finally, in a tiny piece of good news, it looks like we'll have tolerable weather Friday for my first Cubs home game this season.

American Airlines, my carrier of choice, will finally replace its fleet of awful MD-80s, many of which it inherited from TWA:

The acquisition of 76 Boeing 737-800s through early 2011 represents a doubling of that airplane model flown by Ft Worth-based American.

All the new planes will be based at O'Hare International Airport.

The move also will lead to the eventual retirement of American's McDonnell Douglas MD-80s--a reliable but noisy aircraft that gulps 35 percent more fuel than the 737-800.

American plans to phase out its 280 MD-80s over about 10 years, said Dan Garton, the carrier's executive vice president of marketing.

"All of the new planes will be based at O'Hare..." Sigh. That makes me so happy.

Via Calculated Risk, a really scary graph from the Oregon State government:

Oregon’s seasonally adjusted unemployment rate rose to 12.1 percent in March from 10.7 percent (as revised) in February. The state’s unemployment rate has risen rapidly and substantially over the past nine months, from a rate of 5.9 percent in June 2008. The U.S. seasonally adjusted unemployment rate rose to 8.5 percent in March, from 8.1 percent in February.

In March, Oregon’s seasonally adjusted nonfarm payroll employment declined by 14,000 jobs, following a drop of 22,800 (as revised) in February.

Industry Payroll Employment (Establishment Survey Data)

In March, five of Oregon’s seven largest private-sector industries recorded substantial seasonally adjusted job declines. The losses were widespread with trade, transportation and utilities down 3,600 jobs and four other major industries each down approximately 2,400 jobs. None of Oregon’s major industries gained a substantial number of Jobs in March.

(Low whistle.)

Can't wait for Illinois' report.

The Chicago Tribune has a photo history of the disaster that shut down Chicago's Loop 17 years ago today. Worth a look.

The plane -- carrying up to 215,000 liters of highly flammable aviation fuel -- was less than 70 cm off the ground when it crashed through lights almost 200 m from the end of the runway.

...The fully-laden Airbus A340-500 was believed to have been travelling about 280 km/h when it reached the end of the runway without becoming airborne.

At the last minute, the two pilots "rotated" the plane [too steeply] causing its tail to crash into the end of the runway.

Damage to the $220 million plane is so severe that the airline is considering writing it off rather than repairing it.

... Aviation officials said ... Emirates' pilot training and competency standards are almost identical to those in Australia....

I corrected the penultimate paragraph because the reporter seems to believe that "rotating" is an emergency procedure that involves yanking the yoke back hard enough to bounce the tail off the runway. Actually "rotating" just means rotating the airplane on its horizontal axis so the nose points up. Rotating a Cessna 172, for example, is such a subtle maneuver at takeoff speed that non-pilot passengers often wonder how the plane got airborne. What the Emirates pilots did was to pull back so hard that they caused a tail strike. In other words, they panicked, which compounded the problem because the plane's angle of attack seems to have been too steep to generate sufficient lift to take off.

As you pull back on the yoke, the tail of the aircraft is pushed down, which pushes the nose up. This translates speed into lift. It also increases drag, which means pulling back too far slows the plane too much which causes lift to drop. In an airplane the size of a Cessna 172, this can cause a takeoff stall; with an Airbus 340, the plane is so long that the tail scrapes the runway long before the plane stalls. That causes a different kind of drag, of course, and now you're "in the world of physics" as pilots say and no longer flying.

I hope to read follow-up about this when it's available; in particular, I'm wondering what went on in the cockpit at takeoff.

Note that the photo is not of the accident airplane.

After some experimentation, I may have found a setup that works:

It has the added advantage of showing the Inner Drive Technology Worldwide Data Center in the background.