Make your own free website on Tripod.com

Managing Software Engineers, by Philip Greenspun (philg@mit.edu)


anteriorpróximo

I think what _every_ respondent failed to see is that Phil's 70+ hour week does not mean 52 weeks at 70+ hours a week each year for 40 years, rather that when a project needs it, the programmer does not ask how many hours they've worked at all -- solve the problem now, let the time work itself out. Another aspect that most fail to acknowledge is that programming is not an 'at will' activity such as flipping a hamburger.

A good coder refuses to write bad code; if their mind is not conditioned at a moment to write good code, they'll not write code. IMHO, Phil's approach respects this _feature_ of quality coders/engineers that traditional management fails to accomodate.

-- brent verner, November 6, 2000

There is a lot of good sense and a lot to learn from in this article. However, there's also a lot to dispute. Comment #1:

Suppose we managers at Ars Digita, presumably an example of the point of view in this article, were given an opportunity to hire a guy named Phil Greenspun. He's certainly very smart, experienced in our kind of project, productive, maybe fantastically so, and a terrific communicator, but is he going to give us the hours? He wants to teach a couple of classes at MIT -- that'll probably cost some time. Maybe he can hold them in the comfortable environment of the office -- a good place to bring friends is probably a good place to bring students -- and maybe he can provoke some intense discussions among them so he can write some code while the students learn productively amongst themselves. This guy also likes to travel and take photos, and write extensively about them. Do you suppose he could take pictures of the visiting dogs fighting over pizza scraps in the playroom, or may snap a few shots from the windows of the ocean-view retreat when it's his group's turn to go there?

Not only Greenspun, but other highly qualified and productive people like and need to have pursuits outside the office, benefiting the world at large, their families, and/or themselves. Often, it would be a mistake to reject those people because they're not willing to give up those pursuits to work in a dotcom environment. If having a database-backed website is so urgent all of a sudden, maybe there's been a slipup in the planning stage. If it has to be designed by no more than two people maybe those two people should be given a reasonable deadline, or should agree to crunch hard for awhile in exchange for time off for the photography trip to Italy later, or should receive more than their normal salary so they can make more of their vacations.

Comment #2: Management guides of all sorts are filled with bogus calculations, and this one is no exception. The maxim that 55 hours is twice as productive as 40 depends upon the assumptions that 25 hours is 'wasted' in coordination of effort, and the remaining 15 hours of the first 40 are all useful work. Why should 25 hours of coordination be necessary in a 40-hour week? Maybe 10 would be enough, and worth striving for. But if only 10 hours were 'wasted' in coordination, it would be necessary to work 70 hours to be twice as productive as you would in 40 (but of course the 40 hours would be twice as productive as Ars Digita's 40 hours!) Or maybe 4 hours out of the non-coordination hours are devoted to surfing the web and slashdot (less than 1 hour per day). If you're trying to do twice as much productive work, you don't necessarily need to double the surfing activities, so simply subtracting off the coordination time and doubling what remains isn't right either.

Comment #3: It makes perfect sense that using time at the office for activities marginal or irrelevant to the current projects should not 'count' as productive activity. If I spend an hour every day on the climbing wall, I ought not to go home for the weekend with only 40 hours spent in the office. But if I'd rather spend that hour a day somewhere else, like walking on the beach, and spend my time at the office in more concentrated work, why should this be a black mark. What counts is the productive hours and the results. If the company wants to make it convenient to have leisure and exercise and relaxation in the office environment, fine, but the guy who wants to go home to read his kids a bedtime story, go to a chamber music concert, or even develop a new algorithm for predicting protein folding, might be getting his project done just as well as the guy who's dangling from that nubbin on the climbing wall.

Comment #4: When I referred my wife to this article, she had several comments: -- Vis-a-vis adding dogs, pianos and climbing walls to the programming environment, how about conjugal visits? -- Is this material absolutely serious? ... There are plenty of reaons for adding amenities to a sleep- and exercise-deprived workstyle in which the company has much to gain from long hours and enhanced loyalties. But, I mean, is this Machiavelli? (And was Machiavelli even totally serious?)

-- Jeff Greif, November 6, 2000

There's nothing unhealthy about longer hours if the programmer knows that 40 hours is acceptable and the extra hours are a free choice. Programming is my hobby as well as my career. If work is fun and the workplace is pleasant, I will choose to work more and it will be no different than going home to concentrate on some other hobby. I also have an S.O. and a life, so I will never work for a company which expects me to work more than 40 hours a week.

-- Chris Newman, November 6, 2000
I can see that I struck a chord with the "70-hour" thing... I don't think I got my intent across clearly. Certainly my personal goal is not to see the programmers of the world work 70-hour weeks for 45 years each and then get a gold watch upon retirement. At the same time, if one of my students came back to me four years after graduating from MIT and said "my career pays adequately but it doesn't inspire me and I watch the clock to make sure that I am out the door at 5:01", I'd be a little bit sad. I'd be happy if my student became an inspired watercolor painter, an inspird environmental activist, an inspired fiction author, an inspired cabinetmaker, maybe even an inspired campaigner for George W. Bush (how's that for extreme). As long as he or she were inspired enough by something to give every ounce of energy and effort. Would I expect that level of effort to be kept up for 45 years? For some exceptionally creative people, sure. A lot of Nobel Prize-winning faculty at MIT still work pretty darn hard though they have their tenure and their fame. For others I'd expect them to go through cycles where relaxation, travel, and family become more important. There would be no shame in working for a few years to build something great (be it a software product, reputation as an artist, non-profit org) and then doing something a bit different and less intense for a few years.

(Oh yes, and for the folks who've mentioned RSI... Microsoft Natural Keyboard. It is the only answer!)

-- Philip Greenspun, November 6, 2000


anteriorpróximo