We’re programmers. Programmers are, in their hearts, architects, and the first thing they want to do when they get to a site is to bulldoze the place flat and build something grand. We’re not excited by incremental renovation: tinkering, improving, planting flower beds.

Have you ever heard of SEMA? It’s a fairly esoteric system for measuring how good a software team is. No, wait! Don’t follow that link! It will take you about six years just to understand that stuff. So I’ve come up with my own, highly irresponsible, sloppy test to rate the quality of a software team. The great part about it is that it takes about 3 minutes. With all the time you save, you can go to medical school.

To make people happy, you have to let them feel like they are in control of their environment. To do this, you need to correctly interpret their actions. The interface needs to behave in the way they are expecting it to behave.

Thus, the cardinal axiom of all user interface design:

A user interface is well-designed when the program behaves exactly how the user thought it would.

As Hillel said, everything else is commentary. All the other rules of good UI design are just corollaries.

There’s a subtle reason that programmers always want to throw away the code and start over. The reason is that they think the old code is a mess. And here is the interesting observation: they are probably wrong. The reason that they think the old code is a mess is because of a cardinal, fundamental law of programming:

It’s harder to read code than to write it.

TRS-80 Level-I BASIC could only store two string variables, A$ and B$. Similarly, I was born with only two bug-storing-slots in my brain. At any given time, I can only remember two bugs. If you ask me to remember three, one of them will fall on the floor and get swept under the bed with the dust bunnies, who will eat it.

A lot of software developers are seduced by the old “80/20” rule. It seems to make a lot of sense: 80% of the people use 20% of the features. So you convince yourself that you only need to implement 20% of the features, and you can still sell 80% as many copies.

Unfortunately, it’s never the same 20%. Everybody uses a different set of features. In the last 10 years I have probably heard of dozens of companies who, determined not to learn from each other, tried to release “lite” word processors that only implement 20% of the features. This story is as old as the PC. Most of the time, what happens is that they give their program to a journalist to review, and the journalist reviews it by writing their review using the new word processor, and then the journalist tries to find the “word count” feature which they need because most journalists have precise word count requirements, and it’s not there, because it’s in the “80% that nobody uses,” and the journalist ends up writing a story that attempts to claim simultaneously that lite programs are good, bloat is bad, and I can’t use this damn thing ’cause it won’t count my words. If I had a dollar for every time this has happened I would be very happy.

Whenever somebody gives you a spec for some new technology, if you can’t understand the spec, don’t worry too much. Nobody else is going to understand it, either, and it’s probably not going to be important. This is the lesson of SGML, which hardly anyone used, until Tim Berners-Lee dumbed it down dramatically and suddenly people understood it. For the same reason he simplified the file transfer protocol, creating HTTP to replace FTP.

You can see this phenomenon all over the place; even within a given technology some things are easy enough to figure out and people use them (like COM’s IUnknown), while others are so morbidly complicated (IMonikers) when they should be simple (what’s wrong with URLs?) that they languish.

When I was 6 and my dad brought home one of the world’s first pocket calculators, an HP-35, he tried to convince me that it had a computer inside it. I thought that was unlikely. All the computers on Star Trek were the size of a room and had big reel-to-reel tape recorders. I thought that there was just a clever correlation between the keys on the keypad and the individual elements of the LED display that happened to produce mathematically correct results. (Hey, I was 6).

Imagine that we had a way of sending actors from Broadway to Hollywood that involved putting them in cars and driving them across the country. Some of these cars crashed, killing the poor actors. Sometimes the actors got drunk on the way and shaved their heads or got nasal tattoos, thus becoming too ugly to work in Hollywood, and frequently the actors arrived in a different order than they had set out, because they all took different routes. Now imagine a new service called Hollywood Express, which delivered actors to Hollywood, guaranteeing that they would (a) arrive (b) in order (c) in perfect condition. The magic part is that Hollywood Express doesn’t have any method of delivering the actors, other than the unreliable method of putting them in cars and driving them across the country. Hollywood Express works by checking that each actor arrives in perfect condition, and, if he doesn’t, calling up the home office and requesting that the actor’s identical twin be sent instead. If the actors arrive in the wrong order Hollywood Express rearranges them. If a large UFO on its way to Area 51 crashes on the highway in Nevada, rendering it impassable, all the actors that went that way are rerouted via Arizona and Hollywood Express doesn’t even tell the movie directors in California what happened. To them, it just looks like the actors are arriving a little bit more slowly than usual, and they never even hear about the UFO crash.

A common misconception, I assume popularized by Hollywood, is that as you get closer to shipping software, activity becomes frenetic as everybody scrambles to finish all the things that need to be done in time for the deadline. In the typical crappy movie, there’s a mad rush of typing in a room full of cool alterna-dressed programmers with found-object earrings and jeans jackets. Somebody stands up and shouts to the room in general “I need the Jiff subroutine! Somebody give me the Jiff subroutine!” A good looking young woman in Vivienne Tam urbanwear throws a floppy disk at him. “Thanks!” As the second hand swoops towards the :00, the whole team waits breathlessly around Ryan Phillippe’s computer and watches the “copy” progress indicator as the final bits are put onto a floppy disk with less than a second to spare before the VC cuts off funding. … On good teams, the days before shipping just get quieter and quieter as programmers literally run out of things to do one at a time. (Yesterday I took the day off to explore New York City with my wee niece and nephews.)

Version 5.0 of Microsoft’s flagship spreadsheet program Excel came out in 1993. It was positively huge: it required a whole 15 megabytes of hard drive space. In those days we could still remember our first 20MB PC hard drives (around 1985) and so 15MB sure seemed like a lot. By the time Excel 2000 came out, it required a whopping 146MB… almost a tenfold increase! Dang those sloppy Microsoft programmers, right?

Wrong.

In 1993, given the cost of hard drives in those days, Microsoft Excel 5.0 took up about $36 worth of hard drive space. In 2000, given the cost of hard drives in 2000, Microsoft Excel 2000 takes up about $1.03 in hard drive space. (These figures are adjusted for inflation and based on hard drive price data from here.)

In real terms, it’s almost like Excel is actually getting smaller!

So I have an announcement to make: if you are a programmer working in 2003 and you don’t know the basics of characters, character sets, encodings, and Unicode, and I catch you, I’m going to punish you by making you peel onions for 6 months in a submarine. I swear I will.

But the idea of unifying the mess of Visual Basic and Windows API programming by creating a completely new, ground-up programming environment with not one, not two, but three languages (or are there four?) is sort of like the idea of getting two quarreling kids to stop arguing by shouting “shut up!” louder than either of them. It only works on TV. In real life, when you shout “shut up!” to two people arguing loudly, you just create a louder three-way argument.

(By the way, for those of you who follow the arcane but politically-charged world of blog syndication feed formats, you can see the same thing happening over there. RSS became fragmented with several different versions, inaccurate specs and lots of political fighting, and the attempt to clean everything up by creating yet another format called Atom has resulted in several different versions of RSS plus one version of Atom, inaccurate specs and lots of political fighting. When you try to unify two opposing forces by creating a third alternative, you just end up with three opposing forces. You haven’t unified anything and you haven’t really fixed anything.)

And here’s the clincher: I noticed (and confirmed this with a recruiter friend) that Windows API programmers here in New York City who know C++ and COM programming earn about $130,000 a year, while typical Web programmers using managed code languages (Java, PHP, Perl, even ASP.NET) earn about $80,000 a year. That’s a huge difference, and when I talked to some friends from Microsoft Consulting Services about this they admitted that Microsoft had lost a whole generation of developers. The reason it takes $130,000 to hire someone with COM experience is because nobody bothered learning COM programming in the last eight years or so, so you have to find somebody really senior, usually they’re already in management, and convince them to take a job as a grunt programmer, dealing with (God help me) marshalling and monikers and apartment threading and aggregates and tearoffs and a million other things that, basically, only Don Box ever understood, and even Don Box can’t bear to look at them any more.

One of the biggest questions you’re going to be asking now is, “How much should I charge for my software?” When you ask the experts they don’t seem to know. Pricing is a deep, dark mystery, they tell you. The biggest mistake software companies make is charging too little, so they don’t get enough income, and they have to go out of business. An even bigger mistake, yes, even bigger than the biggest mistake, is charging too much, so they don’t get enough customers, and they have to go out of business. Going out of business is not good because everybody loses their job, and you have to go work at Wal*Mart as a greeter, earning minimum wage and being forced to wear a polyester uniform all day long.

So if you like cotton uniforms you better get this right.

The answer is really complicated. I’m going to start with a little economic theory, then I’m going to tear the theory to bits, and when I’m finished, you’ll know a lot more about pricing and you still won’t know how much to charge for your software, but that’s just the nature of pricing. If you can’t be bothered to read this, just charge $0.05 for your software, unless it does bug tracking, in which case charge $30,000,000 for it.

NOW WE’RE GETTING SOMEWHERE!

This is really cool. I think we’re on the verge of solving the problem of how much to charge for software! I’M SO EXCITED!

The reason I’m so excited is it looks like if you plot price against profit, you get a nice curve with a big hump in the middle! And we all know what humps mean! Humps mean local maxima! Or camels. But here they mean local maxima!

“O frabjous day! Callooh! Callay!” I chortle. We have found the optimum price, $220, and that’s how much you should charge for your software. Thanks for your time.

Ahem.

Thank you for your time! Nothing more to see here! Move along now!

You’re not leaving.

I see.

Some of the more observant members of my audience have detected through careful analysis of the scrollbar position in their web browser that I might have something more to say other than “$220.”

Well, maybe. There’s just a tiny little loose end I left untied which I might as well tie up now if you’re all still up for it. Ok? OK!

The difference between $399 and $220, i.e., $179, is called consumer surplus. It’s the extra value that those rich consumers got from their purchase that they would have been perfectly happy to do without.

It’s sort of like if you were all set to buy that new merino wool sweater, and you thought it was going to cost $70, which is well worth it, and when you got to Banana Republic it was on sale for only $50! Now you have an extra $20 in found money that you would have been perfectly happy to give to the Banana Republicans!

Yipes!

That bothers good capitalists. Gosh darn it, if you’re willing to do without it, well, give it to me! I can put it to good use, buying a SUV or condo or Mooney or yacht one of those other things capitalists buy!

In economist jargon, capitalists want to capture the consumer surplus.

There are more subtle ways to segment. You know those grocery coupons you see in the paper? The ones that get you 25 cents off a box of Tide detergent if you clip them out and remember to bring them to the store? Well, the trouble with grocery coupons is that there’s so much manual labour involved in clipping them, and sorting them out, and remembering which ones to use, and choosing brands based on which coupons you have, and so on, and the net effect is that if you clip coupons you’re probably working for about $7.00 an hour.

Now, if you’re retired and living off of social security, $7 an hour sounds pretty good, so you do it, but if you’re a stock analyst at Merrill Lynch getting paid $12,000,000 a year to say nice things about piece-of-junk Internet companies, working for $7 an hour is a joke, and you’re not going to clip coupons. Heck, in one hour you could issue “buy” recommendations on ten piece-of-junk Internet companies! So coupons are a way for consumer products companies to charge two different prices and effectively segment their market into two. Mail-in rebates are pretty much the same as coupons, with some other twists like the fact that they reveal your address, so you can be direct marketed to in the future.

In the world of software, you can just make a version of your product called “Professional” and another version called “Home” with some inconsequential differences, and hope that the corporate purchasers (again, the people who are not spending their own money) will be too embarrassed at the thought of using “Windows XP Home Edition” at work and they’ll buy the Pro edition. Home Edition at work? Somehow that feels like coming to work in your pyjamas! Ick!

Bad Idea #1: Site Licenses.

The opposite of segmentation, really. I have certain competitors that do this: they charge small customers per-user but then there’s a “unlimited” license at a fixed price. This is nutty, because you’re giving the biggest price break precisely to the largest customers, the ones who would be willing to pay you the most money. Do you really want IBM to buy your software for their 400,000 employees and pay you $2000? Hmm?

As soon as you have an “unlimited” price, you are instantly giving a gigantic gift of consumer surplus to the least price-sensitive customers who should have been the cash cows of your business.

Notice the gap? There’s no software priced between $1000 and $75,000. I’ll tell you why. The minute you charge more than $1000 you need to get serious corporate signoffs. You need a line item in their budget. You need purchasing managers and CEO approval and competitive bids and paperwork. So you need to send a salesperson out to the customer to do PowerPoint, with his airfare, golf course memberships, and $19.95 porn movies at the Ritz Carlton. And with all this, the cost of making one successful sale is going to average about $50,000. If you’re sending salespeople out to customers and charging less than $75,000, you’re losing money.

The joke of it is, big companies protect themselves so well against the risk of buying something expensive that they actually drive up the cost of the expensive stuff, from $1000 to $75000, which mostly goes towards the cost of jumping all the hurdles that they set up to insure that no purchase can possibly go wrong.

We have lots of FogBugz customers who have high-priced Remedy, Rational, or Mercury products sitting on the shelves after investments of well over $100,000, because that software isn’t good enough to actually use. Then they buy a couple of thousand dollars worth of FogBugz and that’s the product they really use. The Rational salesperson is laughing at me, because I have $2000 in the bank and he has $100,000. But I have far more customers than he does, and they’re all using my product, and evangelizing it, and spreading it, while Rational customers either (a) don’t use it or (b) use it and can’t stand it. But he’s still laughing at me from his 40 foot yacht while I play with rubber duckies in the bathtub. Like I said, all three methods work fine. But cheaper prices is like buying advertising and as such is an investment in the future.

You can have focus groups and ask people, but they’ll lie to you. Some people will lie to show off their generosity and wealth. “Heck, yeah, I’d buy a pair of $400 jeans in a New York Minute!” Other people will lie because they really want your thing and they think you’ll decide to charge less money if they tell you a low number. “Blogging software? Hmm. I’d pay, at most, 38 cents.”

Then you ask another focus group the next day, and this time, the first man to speak has a crush on a pretty woman in the group, and he wants to impress her, so he starts talking about how much his car cost and everyone is thinking Big Numbers. And the day after that, you serve Starbucks during the break, and while you’re in the john everyone unbeknownst to you gets into a side conversation about paying $4 for a cup of coffee, and they’re in a real frugal mood when you ask them about their willingness to pay.

Then you finally get the focus group to agree that your software is worth $25 a month, and then you ask them how much they would pay for a permanent license and the same people just won’t go a penny over $100. People seriously can’t count.

And, in fact, you can’t even be sure that the demand curve is downward sloping.

The only reason we assumed that the demand curve is downward sloping is that we assumed things like “if Freddy is willing to buy a pair of sneakers for $130, he is certainly willing to buy those same sneakers for $20.” Right? Ha! Not if Freddy is an American teenager! American teenagers would not be caught dead in $20 sneakers. It’s, like, um, the death penalty? if you are wearing sneakers? that only cost $20 a pair? in school?

I’m not joking around here: prices send signals. Movies in my town cost, I think, $11. Criminy. There used to be a movie theatre that had movies for $3. Did anyone go there? I DON’T THINK SO. It’s obviously just a dumping ground for lousy movies. Somebody is now at the bottom of the East River with $20.00 cement sneakers because they dared to tell the consumer which movies the industry thought were lousy.

When you’re setting a price, you’re sending a signal. If your competitor’s software ranges in price from about $100 to about $500, and you decide, heck, my product is about in the middle of the road, so I’ll sell it for $300, well, what message do you think you’re sending to your customers? You’re telling them that you think your software is “eh.” I have a better idea: charge $1350. Now your customers will think, “oh, man, that stuff has to be the cat’s whiskers since they’re charging mad coin for it!”

And then they won’t buy it because the limit on the corporate AMEX is $500.

Misery.

The more you learn about pricing, the less you seem to know.

I’ve been nattering on about this topic for well over 5000 words and I don’t really feel like we’re getting anywhere, you and I.

Some days it seems like it would be easier to be a taxi driver, with prices set by law. Or to be selling sugar. Plain ol’ sugar. Yep. That would be sweet.

Take my advice, offered about 20 pages back: charge $0.05 for your software. Unless it does bug tracking, in which case the correct price is $30,000,000. Thank you for your time, and I apologize for leaving you even less able to price software than you were when you started reading this.

In Usenet, whenever a single newsgroup got too large, it tended to fork. So from comp we got comp.sys.ibm.pc which split into smaller and smaller groups like the unloved comp.sys.ibm.pc.hardware.video, created because people were sick of talking about video drivers on the main group.

I didn’t like forks, because they make discussions less interesting. I mean, it’s bad enough there’s a comp.software.windows.nt.40.microsoft.notepad, does there have to be a comp.software.windows.nt.40.microsoft.notepad.helpfile.index? Seriously now.

I want to tell you two stories from my career which I think are classic illustrations of the difference between tech companies that are well-managed and tech companies that are disasters. It comes down to the difference between trusting employees and letting them get things done, versus treating them like burger flippers that need to be monitored and controlled every minute, lest they wander off and sabotage everything.

My first assignment at my first job was working at Microsoft, where I was told to come up with a new macro language strategy for Excel. Pretty soon, I had the first draft of the “Excel Basic” spec (which later evolved into Visual Basic for Applications, but that’s another story). Somehow, this mysterious group of people at Microsoft called the “Application Architecture” group got wind of my spec, which must have concerned them, because for some reason they thought that they were in charge of things like macro language strategies, and they asked to see my spec.

I asked around. Who’s the Application Architecture group? Nobody seemed to think they were very serious. It turns out that they were a group of just four people, recent hires with PhDs (very unusual for Microsoft). I sent them a copy of my spec and went to meet them, in case they had something interesting to say.

I sent them [= the Application Architecture group] a copy of my spec and went to meet them, in case they had something interesting to say.

“Blah blah!” said one of them. “Blah blah blah, blah blah blah!” said another. I don’t think they quite had anything interesting to say. They were very enamored of the idea of subclassing and sort of thought that people making macros in Excel wanted to subclass a lot of things. In any case, one of the fellows said, “Well, this is all very interesting. What’s next? Who has to approve your spec?”

I laughed. Even though I had only been at Microsoft for a few months, I knew that there was no such thing as somebody approving my spec. Hell, nobody had time to read my spec, let alone approve it. The programmers were bugging me every day to get them more pages so that they could write more code. My boss (and his boss) made it very clear to me that nobody else understood macros or had time to work on macros, so whatever I did, it better be right. And here this PhD working in a strange research group at Microsoft assumed that things were a bit more formal than that.

I pretty rapidly realized that the App Architecture group knew even less than I did about macros. At least, I had talked to a handful of macro developers and some Excel old-timers to get a grip on what people actually did with Excel macros: things like recalculating a spreadsheet every day, or rearranging some data according to a certain pattern. But the App Architecture group had merely thought about macros as an academic exercise, and they couldn’t actually come up with any examples of the kind of macros people would want to write. Pressured, one of them came up with the idea that since Excel already had underlining and double-underlining, perhaps someone would want to write a macro to triple underline. Yep. REAL common. So I proceeded to ignore them as diplomatically as possible.

I would have been perfectly happy to leave it at that. If the Apps Architecture team needed care and feeding and wanted to argue about stuff, that was OK, I would argue with them as much as they wanted as long as they left the programmers alone to do their work. But then something even more interesting happened that blew my mind. I was sitting at lunch with some coworkers, in the Redmond sun, when Pete Higgins came up to me. At that time Pete was the general manager for Office -- I knew who he was, of course, but didn’t expect that he knew me very well.

- “How’s it going, Joel?” he asked.

“I hear you’ve been having some issues with the App Architecture group.”

- “Oh no!” I said. “Nothing I can’t handle.”

- “Say no more,” he said, “I understand.” He left.

By the next day the rumor had gotten back to me: the App Architecture group was disbanded. Not only that, but each member of the group was sent to a different department at Microsoft, as far apart as possible. I never heard from them again. I was blown away, of course. At Microsoft, if you’re the Program Manager working on the Excel macro strategy, even if you’ve been at the company for less than six months, it doesn’t matter - you are the GOD of the Excel macro strategy, and nobody, not even employee number 6, is allowed to get in your way. Period.

Let’s look at a small example. The Unix programming culture holds in high esteem programs which can be called from the command line, which take arguments that control every aspect of their behavior, and the output of which can be captured as regularly-formatted, machine readable plain text. Such programs are valued because they can easily be incorporated into other programs or larger software systems by programmers. To take one miniscule example, there is a core value in the Unix culture, which Raymond calls “Silence is Golden,” that a program that has done exactly what you told it to do successfully should provide no output whatsoever. It doesn’t matter if you’ve just typed a 300 character command line to create a file system, or built and installed a complicated piece of software, or sent a manned rocket to the moon. If it succeeds, the accepted thing to do is simply output nothing. The user will infer from the next command prompt that everything must be OK.

Brett also snuck in a feature he’s been itching for: lots and lots and lots of keyboard shortcuts. There’s only one keyboard shortcut you have to memorize, though: Ctrl+; switches FogBugz into keyboard mode and little letters light up reminding you what the shortcuts are for various commands around the screen. It’s really pretty cool to be able to work through a bunch of cases, assigning, editing, and reprioritizing, without ever reaching for the mouse. Combined with the speed and responsiveness from Ajax, FogBugz has almost reached the level of speed and fluidity of my dry cleaner’s DOS 2.0 character mode database application. And that’s pretty darn responsive for a web app.

The recruiters-who-use-grep, by the way, are ridiculed here, and for good reason. I have never met anyone who can do Scheme, Haskell, and C pointers who can’t pick up Java in two days, and create better Java code than people with five years of experience in Java, but try explaining that to the average HR drone.

The most sympathetic interpretation of why CS departments are so enthusiastic to dumb down their classes is that it leaves them more time to teach actual CS concepts, if they don’t need to spend two whole lectures unconfusing students about the difference between, say, a Java int and an Integer. Well, if that’s the case, 6.001 has the perfect answer for you: Scheme, a teaching language so simple that the entire language can be taught to bright students in about ten minutes; then you can spend the rest of the semester on fixed points. Feh.

I’m going back to ones and zeros.

(You had ones? Lucky bastard! All we got were zeros.)

Of course, these are not the only factors that determine success or failure: in particular, if you have a great software team working on a product that nobody wants, well, people aren’t going to want it. And it’s possible to imagine a team of “gunslingers” that doesn’t do any of this stuff that still manages to produce incredible software that changes the world. But, all else being equal, if you get these 12 things right, you’ll have a disciplined team that can consistently deliver.

The very first version of Microsoft Word for Windows was considered a “death march” project. It took forever. It kept slipping. The whole team was working ridiculous hours, the project was delayed again, and again, and again, and the stress was incredible. When the dang thing finally shipped, years late, Microsoft sent the whole team off to Cancun for a vacation, then sat down for some serious soul-searching.

What they realized was that the project managers had been so insistent on keeping to the “schedule” that programmers simply rushed through the coding process, writing extremely bad code, because the bug fixing phase was not a part of the formal schedule. There was no attempt to keep the bug-count down. Quite the opposite. The story goes that one programmer, who had to write the code to calculate the height of a line of text, simply wrote “return 12;” and waited for the bug report to come in about how his function is not always correct. The schedule was merely a checklist of features waiting to be turned into bugs. In the post-mortem, this was referred to as “infinite defects methodology”.

To correct the problem, Microsoft universally adopted something called a “zero defects methodology”. Many of the programmers in the company giggled, since it sounded like management thought they could reduce the bug count by executive fiat. Actually, “zero defects” meant that at any given time, the highest priority is to eliminate bugs before writing any new code.

Writing code in a compiled language is one of the last things that still can’t be done instantly on a garden variety home computer. If your compilation process takes more than a few seconds, getting the latest and greatest computer is going to save you time. If compiling takes even 15 seconds, programmers will get bored while the compiler runs and switch over to reading The Onion, which will suck them in and kill hours of productivity.

Debugging GUI code with a single monitor system is painful if not impossible. If you’re writing GUI code, two monitors will make things much easier.

Most programmers eventually have to manipulate bitmaps for icons or toolbars, and most programmers don’t have a good bitmap editor available. Trying to use Microsoft Paint to manipulate bitmaps is a joke, but that’s what most programmers have to do.

At my last job, the system administrator kept sending me automated spam complaining that I was using more than… get this… 220 megabytes of hard drive space on the server. I pointed out that given the price of hard drives these days, the cost of this space was significantly less than the cost of the toilet paper I used. Spending even 10 minutes cleaning up my directory would be a fabulous waste of productivity.

This has two implications.

One: it’s crucial that tech support have access to the development team. This means that you can’t outsource tech support: they have to be right there at the same street address as the developers, with a way to get things fixed. Many software companies still think that it’s “economical” to run tech support in Bangalore or the Philippines, or to outsource it to another company altogether. Yes, the cost of a single incident might be $10 instead of $50, but you’re going to have to pay $10 again and again.

When we handle a tech support incident with a well-qualified person here in New York, chances are that’s the last time we’re ever going to see that particular incident. So with one $50 incident we’ve eliminated an entire class of problems.

Somehow, the phone companies and the cable companies and the ISPs just don’t understand this equation. They outsource their tech support to the cheapest possible provider and end up paying $10 again and again and again fixing the same problem again and again and again instead of fixing it once and for all in the source code. The cheap call centers have no mechanism for getting problems fixed; indeed, they have no incentive to get problems fixed because their income depends on repeat business, and there’s nothing they like better than being able to give the same answer to the same question again and again.

Microsoft’s Raymond Chen tells the story of a customer who complains that the keyboard isn’t working. Of course, it’s unplugged. If you try asking them if it’s plugged in, “they will get all insulted and say indignantly, ‘Of course it is! Do I look like an idiot?’ without actually checking.”

“Instead,” Chen suggests, “say ‘Okay, sometimes the connection gets a little dusty and the connection gets weak. Could you unplug the connector, blow into it to get the dust out, then plug it back in?’

“They will then crawl under the desk, find that they forgot to plug it in (or plugged it into the wrong port), blow out the dust, plug it in, and reply, ‘Um, yeah, that fixed it, thanks.’”

Many requests for a customer to check something can be phrased this way. Instead of telling them to check a setting, tell them to change the setting and then change it back “just to make sure that the software writes out its settings.”

When the Macintosh first came out, there was no software available for it. So obviously, Apple created a giant glossy catalog listing all the great software that was “available”. Half of the items listed said, in fine print, “under development,” and the other half couldn’t be had for love or money. Some were such lame products nobody would buy them. But even having a thick glossy catalog with one software “product” per page described in glowing prose couldn’t disguise the fact that you just could not buy a word processor or spreadsheet to run on your 128KB Macintosh. There were similar “software product guides” for NeXT and BeOS. (Attention, NeXT and BeOS bigots: I don’t need any flak about your poxy operating systems, OK? Write your own column.) The only thing a software product guide tells you is that there is no software available for the system. When you see one of these beasts, run fleeing in the opposite direction.

Amiga, Atari ST, Gem, IBM TopView, NeXT, BeOS, Windows CE, General Magic, the list of failed “new platforms” goes on and on. Because they are platforms, they are, by definition, not very interesting in and of themselves without juicy software to run on them. But, with very few exceptions (and I’m sure I’ll get a whole host of email from tedious supporters of arcane and unloved platforms like the Amiga or RSTS-11), no software developer with the least bit of common sense would intentionally write software for a platform with 100,000 users on a good day, like BeOS, when they could do the same amount of work and create software for a platform with 100,000,000 users, like Windows. The fact that anybody writes software for those oddball systems at all proves that the profit motive isn’t everything: religious fervor is still alive and well. Good for you, darling. You wrote a nice microEmacs clone for the Timex Sinclair 1000. Bravo. Here’s a quarter, buy yourself a treat.

Software developers don’t really like to make schedules. Usually, they try to get away without one. “It’ll be done when it’s done!” they say, expecting that such a brave, funny zinger will reduce their boss to a fit of giggles, and in the ensuing joviality, the schedule will be forgotten.

Most of the schedules you do see are halfhearted attempts. They’re stored on a file share somewhere and completely forgotten. When these teams ship, two years late, that weird guy with the file cabinet in his office brings the old printout to the post mortem, and everyone has a good laugh. “Hey look! We allowed two weeks for rewriting from scratch in Ruby!”

Hilarious! If you’re still in business.

Most estimators get the scale wrong but the relative estimates right. Everything takes longer than expected, because the estimate didn’t account for bug fixing, committee meetings, coffee breaks, and that crazy boss who interrupts all the time. This common estimator has very consistent velocities, but they’re below 1.0. For example, {0.6, 0.5, 0.6, 0.6, 0.5, 0.6, 0.7, 0.6}

Let me walk you through a quick example. To make this example as simple as possible, I’m going to imagine a very predictable programmer, John, whose whole job is writing those one-line getter and setter functions that inferior programming languages require. All day long this is all he does:

private int width; public int getWidth () { return width; } public void setWidth (int _width} { width = _width; }

I know, I know… it’s a deliberately dumb example, but you know you’ve met someone like this.

Anyway. Each getter or setter takes him 2 hours. So his task estimates look like this:

{2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, … }

Now, this poor guy has a boss who interrupts him every once in a while with a two-hour conversation about marlin fishing. Now, of course, John could have a task on his schedule called “Painful conversations about marlin,” and put that on his timesheet, but this might not be politically prudent. Instead, John just keeps the clock running. So his actual times look like this:

{2, 2, 2, 2, 4, 2, 2, 2, 2, 4, 2, … }

And his velocities are:

{1, 1, 1, 1, 0.5, 1, 1, 1, 1, 0.5, 1, … }

Assuming you had everything planned down to the last detail when you started work, EBS works great. To be honest, though, you may do some features that you hadn’t planned. You get new ideas, your salespeople sell features you don’t have, and somebody on the board of directors comes up with a cool new idea to make your golf cart GPS application monitor EKGs while golfers are buzzing around the golf course. All this leads to delays that could not have been predicted when you did the original schedule.

Way back when I was working on Excel 5, our initial feature list was huge and would have gone way over schedule. “Oh my!” we thought. “Those are all super important features! How can we live without a macro editing wizard?”

As it turns out, we had no choice, and we cut what we thought was “to the bone” to make the schedule. Everybody felt unhappy about the cuts. To make people feel better, we told ourselves that we weren’t cutting the features, we were simply deferring them to Excel 6.

As Excel 5 was nearing completion, I started working on the Excel 6 spec with a colleague, Eric Michelman. We sat down to go through the list of “Excel 6” features that had been punted from the Excel 5 schedule. Guess what? It was the shoddiest list of features you could imagine. Not one of those features was worth doing. I don’t think a single one of them ever was. The process of culling features to fit a schedule was the best thing we could have done. If we hadn’t done this, Excel 5 would have taken twice as long and included 50% useless crap features that would have had to be supported, for backwards compatibility, until the end of time.

Last week, Microsoft published the binary file formats for Office. These formats appear to be almost completely insane. The Excel 97-2003 file format is a 349 page PDF file. But wait, that’s not all there is to it! This document includes the following interesting comment:

Each Excel workbook is stored in a compound file.

You see, Excel 97-2003 files are OLE compound documents, which are, essentially, file systems inside a single file. These are sufficiently complicated that you have to read another 9 page spec to figure that out. And these “specs” look more like C data structures than what we traditionally think of as a spec. It’s a whole hierarchical file system.

If you started reading these documents with the hope of spending a weekend writing some spiffy code that imports Word documents into your blog system, or creates Excel-formatted spreadsheets with your personal finance data, the complexity and length of the spec probably cured you of that desire pretty darn quickly. A normal programmer would conclude that Office’s binary file formats:

• are deliberately obfuscated
• are the product of a demented Borg mind
• were created by insanely bad programmers
• and are impossible to read or create correctly.

You know, as much as I hate to see Joel backhand ole Perl, I’ve seen it done before (http://www.cabochon.com/~stevey/blog-rants/blog-ancient-perl.html) and I’m over it. Whatever. It’s not a real object oriented language, too many idiosyncrasies, yadda yadda yadda I woke up next to python.

But, I mean, PHP?? Uh, what?

Weekly root exploits, a thousand ways to escape a DB insert, object oriented is even further behind in adoption than Perl (there’s actually tons of pretty clean oo Perl code, on CPAN), doesn’t play well with apache2.

Oh but it’s popular. Yes, it’s very easy to find PHP programmers. You know, it’s also pretty easy to find JavaScript programmers! COBOL was apparently pretty hot at one point.

It’s not that I have anything against PHP. I mean, people seem to get sh*t done with it. And they’re not all friggin frigtards. Even 37 signals built their website in PHP and those guys are supposed to be the bees knees.

But… PHP is industrial strength, Python is halfway there, and Perl is ass?

Yes. Also: Toyota can be deeded to your grandkids, but Honda will explode before you drive it home; Heath tastes totally incredible and Skor will make you vomit.

A few months ago when we released CityDesk, I got an email from a customer complaining that he was used to doing Alt+F, Alt+S to save files. Unfortunately due to a tiny, unnoticed bug, that keyboard shortcut saved the file and then closed it, irritatingly. I had never noticed because I’m in the habit of doing Alt+F,S to save files, not Alt+F,Alt+S -- a tiny difference -- and Alt+F,S worked fine.

Once you get into the habit of doing Alt+F,Alt+S to save, it becomes so automatic you don’t think of it as Alt+F,Alt+S. You think of it as save. And when you push the “save” button in your brain and the file you were working on goes away, it makes you feel like you’re not in control of your environment. It’s a small thing, but about the fourth time that it happens, you’re going to be seriously unhappy. That’s why I spent several hours tracking down this bug and fixing it. In a bizarre application of Murphy’s Law, this fix led to a cascade of events that caused us to waste something like a week, but that’s neither here nor there. It was worth the time spent. This is what it means to be concerned about usability. If you still think that something as small as how long you hold down the Alt key when you active a menu command doesn’t matter, well, your software is going to make people unhappy. These tiny inconsistencies are what makes Swing applications so unbearably annoying to use, and in my opinion it’s why there are virtually no commercially successful Java GUI applications.

You’re about to see the mother of all flamewars on internet groups where web developers hang out. It’ll make the Battle of Stalingrad look like that time your sister-in-law stormed out of afternoon tea at your grandmother’s and wrapped the Mustang around a tree.

This upcoming battle will be presided over by Dean Hachamovitch, the Microsoft veteran currently running the team that’s going to bring you the next version of Internet Explorer, 8.0. The IE 8 team is in the process of making a decision that lies perfectly, exactly, precisely on the fault line smack in the middle of two different ways of looking at the world. It’s the difference between conservatives and liberals, it’s the difference between “idealists” and “realists,” it’s a huge global jihad dividing members of the same family, engineers against computer scientists, and Lexuses vs. olive trees.

And there’s no solution. But it will be really, really entertaining to watch, because 99% of the participants in the flame wars are not going to understand what they’re talking about. It’s not just entertainment: it’s required reading for every developer who needs to design interoperable systems.

And the whole problem hinges on the little tiny decision of what IE8 should do when it encounters a page that claims to support “standards”, but has probably only been tested against IE7.

What the hell is a standard?

Don’t they have standards in all kinds of engineering endeavors? (Yes.)

Don’t they usually work? (Mmmm…..)

Why are “web standards” so frigging messed up? (It’s not just Microsoft’s fault. It’s your fault too. And Jon Postel’s (1943-1998). I’ll explain that later.)

There is no solution. Each solution is terribly wrong. Eric Bangeman at ars technica writes, “The IE team has to walk a fine line between tight support for W3C standards and making sure sites coded for earlier versions of IE still display correctly.” This is incorrect. It’s not a fine line. It’s a line of negative width. There is no place to walk. They are damned if they do and damned if they don’t.

That’s why I can’t take sides on this issue and I’m not going to. But every working software developer should understand, at least, how standards work, how standards should work, how we got into this mess, so I want to try to explain a little bit about the problem here, and you’ll see that it’s the same reason Microsoft Vista is selling so poorly, and it’s the same issue I wrote about when I referred to the Raymond Chen camp (pragmatists) at Microsoft vs. the MSDN camp (idealists), the MSDN camp having won, and now nobody can figure out where their favorite menu commands went in Microsoft Office 2007, and nobody wants Vista, and it’s all the same debate: whether you are an Idealist (”red”) or a Pragmatist (”blue”).

Imagine that you went to Mars, where you discovered that the beings who live there don’t have the portable music player. They’re still using boom boxes.

You realize this is a huge business opportunity and start selling portable MP3 players (except on Mars they’re called Qxyzrhjjjjukltks) and compatible headphones. To connect the MP3 player to the headphones, you invent a neat kind of metal jack that looks like this:

Because you control the player and the headphone, you can ensure that your player works with your headphones. This is a ONE TO ONE market. One player, one headphone.

So far, all is well. We have a de-facto standard for headphone jacks here. The written spec is not complete and not adequate, but anybody who wants to make a compatible headphone just has to plug it into your personal stereo device and test it, and if it works, all is well, they can sell it, and it will work.

Until you decide to make a new version, the Qxyzrhjjjjukltk 2.0.

The Qxyzrhjjjjukltk 2.0 is going to include a telephone (turns out Marslings didn’t figure out cell phones on their own, either) and the headphone is going to have to have a built-in microphone, which requires one more conductor, so you rework the connector into something totally incompatible and kind of ugly, with all kinds of room for expansion:

Completely different 25-conductor connector. And the Qxyzrhjjjjukltk 2.0 is a complete and utter failure in the market. Yes, it has a nice telephone thing, but nobody cared about that. They cared about their large collections of headphones. It turns out that when I said Marslings are very particular about the color of things that they stick in their ears, I meant it. Most trendy Marslings at this point have a whole closet full of nice headphones. They all look the same to you (red), but Marslings are very, very finicky about shades of red in a way that you never imagined. The newest high-end apartments on Mars are being marketed with a headphone closet. I kid you not.

A few years pass; you’re still selling Qxyzrhjjjjukltks like crazy; but now there are lots of Qxyzrhjjjjukltk clones on the market, like the open source FireQx, and lots of headphones, and you all keep inventing new features that require changes to the headphone jack and it’s driving the headphone makers crazy because they have to test their new designs out against every Qxyzrhjjjjukltk clone which is costly and time consuming and frankly most of them don’t have time and just get it to work on the most popular Qxyzrhjjjjukltk 5.0, and if that works, they’re happy, but of course when you plug the headphones into FireQx 3.0 lo and behold they explode in your hands because of a slight misunderstanding about some obscure thing in the spec which nobody really understands called hasLayout, and everybody understands that when it’s raining the hasLayout property is true and the voltage is supposed to increase to support the windshield-wiper feature, but there seems to be some debate over whether hail and snow are rain for the purpose of hasLayout, because the spec just doesn’t say. FireQx 3.0 treats snow as rain, because you need windshield wipers in the snow, Qxyzrhjjjjukltk 5.0 does not, because the programmer who worked on that feature lives in a warm part of Mars without snow and doesn’t have a driver’s license anyway. Yes, they have driver’s licenses on Mars.

And eventually some tedious bore writes a lengthy article on her blog explaining a trick you can use to make Qxyzrhjjjjukltk 5.0 behave just like FireQx 3.0 through taking advantage of a bug in Qxyzrhjjjjukltk 5.0 in which you trick Qxyzrhjjjjukltk into deciding that it’s raining when it’s snowing by melting a little bit of the snow, and it’s ridiculous, but everyone does it, because they have to solve the hasLayout incompatibility. Then the Qxyzrhjjjjukltk team fixes that bug in 6.0, and you’re screwed again, and you have to go find some new bug to exploit to make your windshield-wiper-equipped headphone work with either device.

NOW. This is the MANY-MANY market. Many players on the left hand side who don’t cooperate, and SCRILLIONS of players on the right hand side. And they’re all making mistakes because To Err Is Human.

If you’ve ever visited the ultra-orthodox Jewish communities of Jerusalem, all of whom agree in complete and utter adherence to every iota of Jewish law, you will discover that despite general agreement on what constitutes kosher food, that you will not find a rabbi from one ultra-orthodox community who is willing to eat at the home of a rabbi from a different ultra-orthodox community. And the web designers are discovering what the Jews of Mea Shearim have known for decades: just because you all agree to follow one book doesn’t ensure compatibility, because the laws are so complex and complicated and convoluted that it’s almost impossible to understand them all well enough to avoid traps and landmines, and you’re safer just asking for the fruit plate.

Last week Kent Beck made a claim that you don’t really need bug tracking databases when you’re doing Extreme Programming, because the combination of pair programming (with persistent code review) and test driven development (guaranteeing 100% code coverage of the automated tests) means you hardly ever have bugs.That didn’t sound right to me. I looked in our own bug tracking database here at Fog Creek to see what kinds of bugs were keeping it busy.

Lo and behold, I discovered that very few of the bugs in there would have been discovered with pair programming or test driven development. Many of our “bugs” are really what XP calls stories -- basically, just feature requests.

A lot of the other bugs were only discovered after much use in the field. The Polish keyboard thing. There’s no way pair programming was going to find that. And logical mistakes that never occurred to us in the way that different features work together. The larger and more complex a program, the more interactions between the features that you don’t think about. A particular unlikely sequence of characters ({${?, if you must know) that confuses the lexer. Some ftp servers produce an error when you delete a file that doesn’t exist (our ftp server does not complain so this never occurred to us.)

I carefully studied every bug. Out of 106 bugs we fixed for the service pack release of CityDesk, exactly 5 of them could have been prevented through pair programming or test driven design. We actually had more bugs that we knew about and thought weren’t important (only to be corrected by our customers!) than bugs that could have been caught by XP methods.

But Kent is right, for other types of development. For most corporate development applications, none of these things would be considered a bug. Program crashes on invalid input? Run it again, and this time watch your {${?’s! And we only have One Kind of FTP server and nobody in the whole company uses Polish Windows.

“Seriously, there is no reason why the command line should be required to configure Oracle just like there is no reason why the command line should be required to configure Linux. It is an indication of a company and people who are trying to raise the barrier to entry in order to hold on to revenue streams from support and training.”

You’re so right. I mean, all those wonderful GUI tools for Linux administration out there that Linux Inc. won’t let you use.

Oh wait. That’s right. There’s no single company deciding what you can and can’t use.

Okay, maybe it’s the fact that the people who *could* write tools for Linux system administration *already know* how to administer Linux systems, so they don’t need GUI tools. Yeah, that sounds a bit more likely.

Maybe with Oracle you have point. I’m sure they balance the lost support revenue from better tools against the lost sales revenue from more people wanting to buy their product because of the tools.

Think of the history of data access strategies to come out of Microsoft. ODBC, RDO, DAO, ADO, OLEDB, now ADO.NET - All New! Are these technological imperatives? The result of an incompetent design group that needs to reinvent data access every goddamn year? (That’s probably it, actually.) But the end result is just cover fire. The competition has no choice but to spend all their time porting and keeping up, time that they can’t spend writing new features. Look closely at the software landscape. The companies that do well are the ones who rely least on big companies and don’t have to spend all their cycles catching up and reimplementing and fixing bugs that crop up only on Windows XP. The companies who stumble are the ones who spend too much time reading tea leaves to figure out the future direction of Microsoft. People get worried about .NET and decide to rewrite their whole architecture for .NET because they think they have to. Microsoft is shooting at you, and it’s just cover fire so that they can move forward and you can’t, because this is how the game is played, Bubby. Are you going to support Hailstorm? SOAP? RDF? Are you supporting it because your customers need it, or because someone is firing at you and you feel like you have to respond? The sales teams of the big companies understand cover fire. They go into their customers and say, “OK, you don’t have to buy from us. Buy from the best vendor. But make sure that you get a product that supports (XML / SOAP / CDE / J2EE) because otherwise you’ll be Locked In The Trunk.” Then when the little companies try to sell into that account, all they hear is obedient CTOs parroting “Do you have J2EE?” And they have to waste all their time building in J2EE even if it doesn’t really make any sales, and gives them no opportunity to distinguish themselves. It’s a checkbox feature -- you do it because you need the checkbox saying you have it, but nobody will use it or needs it. And it’s cover fire.

In those days, I thought, golly, there are zillions of bug tracking packages out there. Every programmer has written a dinky bug tracking package. Why would anyone buy ours? I knew one thing: programmers who start businesses often have the bad habit of thinking everybody else is a programmer just like them and wants the same stuff as them, and so they have an unhealthy tendency to start businesses that sell programming tools. That’s why you see so many scrawny companies hawking source-code-generating geegaws, error catching and emailing geegaws, debugging geegaws, syntax-coloring editing tchotchkes, ftping baubles, and, ahem, bug tracking packages. All kinds of stuff that only a programmer could love. I had no intention of falling into that trap!

Of course, nothing ever works out exactly as planned. FogBUGZ was popular. Really popular. It accounts for a significant chunk of Fog Creek’s revenue and sales are growing steadily. The People won’t stop buying it.

So we did version 2.0. This was an attempt to add some of the most obviously needed features. While David worked on version 2.0 we honestly didn’t think it was worth that much effort, so he tended to do things in what you might call an “expedient” fashion rather than, say, an “elegant” fashion. Certain, ahem, design issues in the original code were allowed to fester. There were two complete sets of nearly identical code for drawing the main bug-editing page. SQL statements were scattered throughout the HTML hither and yon, to and fro, pho and ton. Our HTML was creaky and designed for those ancient browsers that were so buggy they could crash loading about:blank.

Yeah, it worked brilliantly, we’ve been at zero known bugs for a while now. But inside, the code was, to use the technical term, a “big mess.” Adding new features was a hemorrhoid. To add one field to the central bug table would probably require 50 modifications, and you’d still be finding places you forgot to modify long after you bought your first family carplane for those weekend trips to your beach house on Mars.

Something important is almost never mentioned in all the literature about programming and software development, and as a result we sometimes misunderstand each other.

You’re a software developer. Me too. But we may not have the same goals and requirements. In fact there are several different worlds of software development, and different rules apply to different worlds.

You read a book about UML modeling, and nowhere does it say that it doesn’t make sense for programming device drivers. Or you read an article saying that “the 20MB runtime [required for .NET] is a NON issue” and it doesn’t mention the obvious: if you’re trying to write code for a 32KB ROM on a pager, it very much is an issue!

I think there are five worlds here, sometimes intersecting, often not. The five are:

1. Shrinkwrap
2. Internal
3. Embedded
4. Games
5. Throwaway

When you read the latest book about Extreme Programming, or one of Steve McConnell’s excellent books, or Joel on Software, or Software Development magazine, you see a lot of claims about how to do software development, but you hardly ever see any mention of what kind of development they’re talking about, which is unfortunate, because sometimes you need to do things differently in different worlds.

When you start out as a beginning programmer, or you try to read code in a new language, it all looks equally inscrutable. Until you understand the programming language itself, you can’t even see obvious syntactic errors.

During the first phase of learning, you start to recognize the things that we usually refer to as “coding style.” So you start to notice code that doesn’t conform to indentation standards and Oddly-Capitalized variables.

It’s at this point you typically say, “Blistering Barnacles, we’ve got to get some consistent coding conventions around here!”, and you spend the next day writing up coding conventions for your team, and the next six days arguing about the One True Brace Style, and the next three weeks rewriting old code to conform to the One True Brace Style, until a manager catches you and screams at you for wasting time on something that can never make money, and you decide that it’s not really a bad thing to only reformat code when you revisit it, so you have about half of a True Brace Style, and pretty soon you forget all about that, and then you can start obsessing about something else irrelevant to making money like replacing one kind of string class with another kind of string class.

Perfectionism

If I was as much of a perfectionist as some here would have me be, I would never get out the door in the morning, I’d be so busy scrubbing the floors of my apartment until they sparkle and shaving every ten minutes and removing lint from my clothing with masking tape, and by the time I finished that I’d have to shave again and take out the trash because there was masking tape in the trash and re-scrub the floor because when I took the trash out I might have tracked in dust. And then I’d have to shave again.

I could go insane with the web page behind the discussion board. First I could make it 110% xhtml 1.1 + CSS. Heck, why not xhtml 2.0 just to be extra addictive-personality-disordered. Then I could neatly format all the html code so it’s perfectly indented. But the html is generated by a script, and the script has to be indented correctly so that it’s perfect too, and a correctly indented ASP script does not, by definition, produce correctly indented HTML. So I could write a filter that takes the output of the ASP script and reindents it so that if anybody does a View Source they would see neatly indented HTML and think I have great attention to detail. Then I would start to obsess about all the wasted bandwidth caused by meaningless whitespace in the HTML file, and I’d go back and forth in circles between compressed HTML and nicely laid out HTML, pausing only to shave.

I could spend the rest of my life perfecting the HTML behind every page on all of our sites, or I could do something that might actually benefit someone.

Perfectionism is a very dangerous quality in business and in life, because by being perfectionist about one thing you are, by definition, neglecting another. The three days I spent insuring that all icons in CityDesk 3.0 are displayed with perfect alpha-blended effects came at the price of having a web site where the descender of the “g” is not a hyperlink. And both are at the price of working on my next book, or writing another article for Joel on Software, or making CityDesk publish really big sites faster.

If you’re noticing a recurring theme, it’s that I never like to talk about whether or not to do X. The question should never be “X, yes or no?” As long as you have limited time and resources, you always have to look at the cost and the benefit of X. Questions should be “Is X worth the time” or “Will X or Y have a greater return on investment?”

In one of Gerald Weinberg’s books, probably The Secrets of Consulting, there’s the apocryphal story of the giant multinational hamburger chain where some bright MBA figured out that eliminating just three sesame seeds from a sesame-seed bun would be completely unnoticeable by anyone yet would save the company $126,000 per year. So they do it, and time passes, and another bushy-tailed MBA comes along, and does another study, and concludes that removing another five sesame seeds wouldn’t hurt either, and would save even more money, and so on and so forth, every year or two, the new management trainee looking for ways to save money proposes removing a sesame seed or two, until eventually, they’re shipping hamburger buns with exactly three sesame seeds artfully arranged in a triangle, and nobody buys their hamburgers any more.

Would Linux have succeeded if Linus Torvalds hadn’t evangelized it? As brilliant a hacker as he is, it was Linus’s ability to convey his ideas in written English via email and mailing lists that made Linux attract a worldwide brigade of volunteers.

Have you heard of the latest fad, Extreme Programming? Well, without getting into what I think about XP, the reason you’ve heard of it is because it is being promoted by people who are very gifted writers and speakers.

Even on the small scale, when you look at any programming organization, the programmers with the most power and influence are the ones who can write and speak in English clearly, convincingly, and comfortably. Also it helps to be tall, but you can’t do anything about that.

The difference between a tolerable programmer and a great programmer is not how many programming languages they know, and it’s not whether they prefer Python or Java. It’s whether they can communicate their ideas. By persuading other people, they get leverage. By writing clear comments and technical specs, they let other programmers understand their code, which means other programmers can use and work with their code instead of rewriting it. Absent this, their code is worthless. By writing clear technical documentation for end users, they allow people to figure out what their code is supposed to do, which is the only way those users can see the value in their code. There’s a lot of wonderful, useful code buried on sourceforge somewhere that nobody uses because it was created by programmers who don’t write very well (or don’t write at all), and so nobody knows what they’ve done and their brilliant code languishes.

I won’t hire a programmer unless they can write, and write well, in English. If you can write, wherever you get hired, you’ll soon find that you’re getting asked to write the specifications and that means you’re already leveraging your influence and getting noticed by management.

Most colleges designate certain classes as “writing intensive,” meaning, you have to write an awful lot to pass them. Look for those classes and take them! Seek out classes in any field that have weekly or daily written assignments.

Start a journal or weblog. The more you write, the easier it will be, and the easier it is to write, the more you’ll write, in a virtuous circle.

Years ago, the Excel team was trying to figure out if it would be a good idea to allow users to drag and drop cells using the mouse. They had a couple of interns “whip up a prototype” suitable for usability testing, using the cutting edge Visual Basic 1.0. Building the prototype took all summer, because it had to duplicate so much of Excel’s real functionality or you couldn’t do a real usability test.

The conclusion of the usability test? Yes, it was a good feature! The programmer in charge spent maybe a week and completely implemented the drag and drop feature. The joke is, of course, that the whole point of creating a prototype is to “save time.”

A year later, another top-secret Microsoft team built a complete prototype for a new user interface using the cutting edge product Asymetrix Toolbook (good lord, it’s hard to believe that thing survived). This prototype took something like a year to build. The real product: Microsoft Bob, the PCjr of the software world. Another wasted prototype.

I’ve basically given up on software prototypes. If the prototype can do everything the product can do, it might as well be the product, and if it can’t, it’s not much use. Luckily, there’s a much better idea: paper prototypes, which neatly solve this problem and the iceberg problem in one fell swoop. Even luckier, Carolyn Snyder has just written a great new book, Paper Prototyping, on the subject. This is an essential reference for anyone designing user interfaces, and it’s well written to boot.

Almost every tech support problem has two solutions. The superficial and immediate solution is just to solve the customer’s problem. But when you think a little harder you can usually find a deeper solution: a way to prevent this particular problem from ever happening again.

Sometimes that means adding more intelligence to the software or the SETUP program; by now, our SETUP program is loaded with special case checks. Sometimes you just need to improve the wording of an error message. Sometimes the best you can come up with is a knowledge base article.

We treat each tech support call like the NTSB treats airliner crashes. Every time a plane crashes, they send out investigators, figure out what happened, and then figure out a new policy to prevent that particular problem from ever happening again. It’s worked so well for aviation safety that the very, very rare airliner crashes we still get in the US are always very unusual, one-off situations.

One reason people are tempted to rewrite their entire code base from scratch is that the original code base wasn’t designed for what it’s doing. It was designed as a prototype, an experiment, a learning exercise, a way to go from zero to IPO in nine months, or a one-off demo. And now it has grown into a big mess that’s fragile and impossible to add code to, and everybody’s whiny, and the old programmers quit in despair and the new ones that are brought in can’t make head or tail of the code so they somehow convince management to give up and start over while Microsoft takes over their business. Today let me tell you a story about what they could have done instead.

FogBUGZ started out six years ago as an experiment to teach myself ASP programming. Pretty soon it became an in-house bug tracking system. It got embellished almost daily with features that people needed until it was good enough that it no longer justified any more work.

Various friends asked me if they could use FogBUGZ at their companies. The trouble was, it had too many hardcoded things that made it a pain to run anywhere other than on the original computer where it was deployed. I had used a bunch of SQL Server stored procedures, which meant that you needed SQL Server to run FogBUGZ, which was expensive and overkill for some of the two person teams that wanted to run it. And so on. So I would tell my friends, “gosh, for $5000 in consulting fees, I’ll spend a couple of days and clean up the code so you can run it on your server using Access instead of SQL Server.” Generally, my friends thought this was too expensive.

After this happened a few times I had a revelation — if I could sell the same program to, say, three people, I could charge $2000 and come out ahead. Or thirty people for $200. Software is neat like that. So in late 2000, Michael sat down, ported the code so that it worked on Access or SQL Server, pulled all the site-specific stuff out into a header file, and we started selling the thing. I didn’t really expect much to come of it.

In most of the English speaking world it is not considered polite to open letters to a Mr. Joel Spolsky by writing “Dear Spolsky.” One might write “Dear Mr. Spolsky,” or “Dear sir,” or perhaps, “Hi Joel!” But “Dear Spolsky” is usually followed by some story about embezzled funds and needing to borrow my bank account.

Remember the definition of an operating system? It’s the thing that manages a computer’s resources so that application programs can run. People don’t really care much about operating systems; they care about those application programs that the operating system makes possible. Word Processors. Instant Messaging. Email. Accounts Payable. Web sites with pictures of Paris Hilton. By itself, an operating system is not that useful. People buy operating systems because of the useful applications that run on it. And therefore the most useful operating system is the one that has the most useful applications.

The logical conclusion of this is that if you’re trying to sell operating systems, the most important thing to do is make software developers want to develop software for your operating system. That’s why Steve Ballmer was jumping around the stage shouting “Developers, developers, developers, developers.” It’s so important for Microsoft that the only reason they don’t outright give away development tools for Windows is because they don’t want to inadvertently cut off the oxygen to competitive development tools vendors (well, those that are left) because having a variety of development tools available for their platform makes it that much more attractive to developers. But they really want to give away the development tools. Through their Empower ISV program you can get five complete sets of MSDN Universal (otherwise known as “basically every Microsoft product except Flight Simulator“) for about $375. Command line compilers for the .NET languages are included with the free .NET runtime… also free. The C++ compiler is now free. Anything to encourage developers to build for the .NET platform, and holding just short of wiping out companies like Borland.