Coding Horror

I am no longer a part of Stack Exchange.

I still have much literal and figurative stock in the success of Stack Exchange, of course, but as of March 1st I will no longer be part of the day to day operations of the company, or the Stack Exchange sites, in any way.

It's been almost exactly 4 years since I chose my own adventure. In those four years, we accomplished incredible things together. Stack Overflow is now an enormous bustling city, a hugely positive influence on the daily lives of programmers around the world, a place to learn from and teach your peers. And the entire Stack Exchange network, born out of the seed of Stack Overflow, is a reference model of high signal, low noise, no-nonsense Q&A that makes the internet better for all of us. I could quote traffic figures, but to me that's not what it's all about. I prefer to think of it building something awesome, because I know that if you build it, they will come.

And they did. I'll be damned if we didn't change our little corner of the Internet for the better. Possibly permanently. This is more than I could have ever hoped for, and I am honored to have been a founding and guiding part of it for the last four years. But I don't need to be a part of it forever – nor should I be, if I've been doing my job correctly. Stack Exchange was always about designing software and creating recipes for self-governing communities who love a particular topic. It is an honor to be a "just" a citizen of this community again, because as a citizen, I too have the power to shape its future. Just like you do.

Startup life is hard on families. We just welcomed two new members into our family, and running as fast as you can isn't sustainable for parents of multiple small children. The death of Steve Jobs, and his subsequent posthumous biography, highlighted the risks for a lot of folks:

For a long time, work was my only thing. I worked evenings, weekends, and Christmas. At those rare times when I wasn’t at work in body, I was there in spirit, unable to speak or think of much else. I wanted so badly to climb the mountain that I stopped asking why I was doing it.

I admire Steve for the mountains he climbed. At the same time, I wonder if he missed the whole point, becoming the John Henry of our time. He won the race, but at what cost?

Me? I may turn out to be a failure in business, but I refuse to fail my kids.

I've followed Brad Wardell's success for a long time, and he had a very similar reaction to Jobs' death.

In the last several years, the company has been successful enough to generate a substantial amount of capital. And with it, I have been fortunate to bring in people with great talent. And so I started thinking of all the amazing things we would do. I would put in crazy hours to do it, of course, but we would go and do amazing things.

Then Steve Jobs died.

And suddenly I realized something. What is the objective here? My oldest child just turned 15. My other two are no longer little either. And I have been missing out on them. And my wife.

For all the success and amazing accomplishments of Steve Jobs, in the end, nothing could save him. Death can come at any time. And I realized that if I found myself on death’s door, I would regret deeply not having spent more time with my kids when they were…well, kids.

You may have more discipline than I do. But for me, the mission is everything; I'm downright religious about it. Stack Overflow and Stack Exchange have been wildly successful, but I finally realized that success at the cost of my children is not success. It is failure.

I've met so many amazing people through Stack Exchange. First, the incredibly talented team of people who work for the company, many of whom I personally recruited. As far as I'm concerned, you are among the best in the world at what you do. That's why we hired you, and it has been an honor to serve with you. But more than that, the broader community that formed around a shared vision of making the Internet better through these beautiful public parks of curated, creative commons Q&A. I have continually been humbled by the brilliant minds that saw fit to work alongside us towards this goal, who selflessly contributed their own time and effort because they just plain loved this stuff as much as we do.

I will miss you all terribly.

What's next for me? I honestly don't know. I do know that I love the Internet, and I remain passionate as ever about making the Internet better – but right now I need to be with my family. In six months, perhaps I'll be ready to choose another adventure. I have total confidence that the team at Stack Exchange, and the thriving community that makes it so great, will carry Stack Exchange onward. After all, our shared voyage never ends, it just takes different forms.

Come, my friends.
'Tis not too late to seek a newer world.
Push off, and sitting well in order smite
the sounding furrows; for my purpose holds
To sail beyond the sunset, and the baths
Of all the western stars, until I die.
It may be that the gulfs will wash us down;
It may be that we shall touch the Happy Isles,
And see the great Achilles, whom we knew.
Though much is taken, much abides; and though
We are not now that strength which in old days
Moved earth and heaven, that which we are, we are —
One equal temper of heroic hearts,
Made weak by time and fate, but strong in will
To strive, to seek, to find, and not to yield.

Farewell, Stack Exchange. I hope you can understand that if I was hard on you at times, it was because I wanted you to be the best you could possibly be.

It was because I loved you.

You know how interviewers love asking about your greatest weakness, or the biggest mistake you've ever made? These questions may sound formulaic, maybe even borderline cliche, but be careful when you answer: they are more important than they seem.

So when people ask me what our biggest mistake was in building Stack Overflow I'm glad I don't have to fudge around with platitudes. I can honestly and openly point to a huge, honking, ridiculously dumb mistake I made from the very first day of development on Stack Overflow – and, worse, a mistake I stubbornly clung to for a solid nine month period after that over the continued protestations of the community. I even went so far as to write a whole blog post decrying its very existence.

For the longest time, I had an awfully Fight Club-esque way of looking at this: the first rule of Stack Overflow was that you didn't discuss Stack Overflow! After all, we were there to learn about programming with our peers, not learn about a stupid website. Right?

I didn't see the need for a meta.

Meta is, of course, the place where you go to discuss the place. Take a moment and think about what that means. Meta is for people who care so deeply about their community that they're willing to go one step further, to come together and spend even more of their time deciding how to maintain and govern it. So, in a nutshell, I was telling the people who loved Stack Overflow the most of all to basically … f**k off and go away.

As I said, not my finest hour.

In my defense, I did eventually figure this out, thanks to the continued prodding of the community. Although we'd used an external meta site since beta, we eventually launched our very own meta.stackoverflow in June 2009, ten months after public beta. And we fixed this very definitively with Stack Exchange. Every Stack Exchange site we launch has a meta from day one. We now know that meta participation is the source of all meaningful leadership and governance in a community, so it is cultivated and monitored closely.

I also paid penance for my sins by becoming the top user of our own meta. I've spent the last 2 years and 7 months totally immersed in the morass of bugs, feature requests, discussions, and support that is our meta. As you can see in my profile, I've visited meta 901 unique days in that time frame, which is disturbingly close to every day. I consider my meta participation stats a badge of honor, but more than that, it's my job to help build this thing alongside you. We explicitly do everything in public on Stack Exchange – it's very intentionally the opposite of Ivory Tower Development.

Along the way I've learned a few lessons about building software with your community, and handling community feedback.

1. 90% of all community feedback is crap.

Let's get this out of the way immediately. Sturgeon's Law can't be denied by any man, woman, child … or community, for that matter. Meta community, I love you to death, so let's be honest with each other: most of the feedback and feature requests you give us are just not, uh, er … actionable, for a zillion different reasons.

But take heart: this means 10% of the community feedback you'll get is awesome! I guarantee you'll find ten posts that are pure gold, that have the potential to make the site clearly better for everyone … provided you have the intestinal fortitude to look at a hundred posts to get there. Be prepared to spend a lot of time, and I mean a whole freaking lot of time, mining through community feedback to extract those rare gems. I believe every community has users savvy enough to produce them in some quantity, and they're often startlingly wonderful.

2. Don't get sweet talked into building a truck.

You should immediately triage the feedback and feature requests you get into two broad buckets:

We need power windows in this car!

or

We need a truck bed in this car!

The former is, of course, a reasonable thing to request adding to a car, while the latter is a request to change the fundamental nature of the vehicle. The malleable form of software makes it all too tempting to bolt that truck bed on to our car. Why not? Users keep asking for it, and trucks sure are convenient, right?

Don't fall into this trap. Stay on mission. That car-truck hybrid is awfully tempting to a lot of folks, but then you end up with a Subaru Brat. Unless you really want to build a truck after all, the users asking for truck features need to be gently directed to their nearest truck dealership, because they're in the wrong place.

3. Be honest about what you won't do.

It always depressed me to see bug trackers and feedback forums with thousands of items languishing there in no man's land with no status at all. That's a sign of a neglected community, and worse, a dishonest relationship with the community. It is sadly all too typical. Don't do this!

I'm not saying you should tell your community that their feedback sucks, even when it frequently does. That'd be mean. But don't be shy about politely declining requests when you feel they don't make sense, or if you can't see any way they could be reasonably implemented. (You should always reserve the right to change your mind in the future, of course.) Sure, it hurts to be rejected – but it hurts far more to be ignored. I believe very, very strongly that if you're honest with your community, they will ultimately respect you more for that.

All relationships are predicated on honesty. If you're not willing to be honest with your community, how can you possibly expect them to respect you … or continue the relationship?

4. Listen to your community, but don't let them tell you what to do.

It's tempting to take meta community requests as a wholesale template for development of your software or website. The point of a meta is to listen to your community, and act on that feedback, right? On the contrary, acting too directly on community feedback is incredibly dangerous, and the reason many of these community initiatives fail when taken too literally. I'll let Tom Preston-Werner, the co-founder of GitHub, explain:

Consider a feature request such as “GitHub should let me FTP up a documentation site for my project.” What this customer is really trying to say is “I want a simple way to publish content related to my project,” but they’re used to what’s already out there, and so they pose the request in terms that are familiar to them. We could have implemented some horrible FTP based solution as requested, but we looked deeper into the underlying question and now we allow you to publish content by simply pushing a Git repository to your account. This meets requirements of both functionality and elegance.

Community feedback is great, but it should never be used as a crutch, a substitute for thinking deeply about what you're building and why. Always try to identify what the underlying needs are, and come up with a sensible roadmap.

5. Be there for your community.

Half of community relationships isn't doing what the community thinks they want at any given time, but simply being there to listen and respond to the community. When the co-founder of Stack Exchange responds to your meta post – even if it wasn't exactly what you may have wanted to hear – I hope it speaks volumes about how committed we are to really, truly building this thing alongside our community.

Regardless of whether money is changing hands or not, you should love discovering some small gem of a community request or bugfix on meta that makes your site or product better, and swooping in to make it so. That's a virtuous public feedback loop: it says you matter and we care and everything just keeps on getting better all in one delightful gesture.

And isn't that what it's all about?

I enjoy my iPhone, but I can't quite come to terms with one aspect of its design: Apple's insistence that there can be only ever be one, and only one, button on the front of the device.

I also own a completely buttonless Kindle Fire, and you'll get no argument from me that there should be at least one obvious "Jesus Handle" button on the front of any gadget. I do wonder why Amazon decided to make the Fire buttonless, when every other Kindle they ship has a home button. Amazon has a track record of making some awfully rough version 1.0 devices; I'm sure they'll add a home button in a version or two. And, hey, at only $199 I'm willing to cut them a little slack. For now.

Microsoft is just as guilty. I sometimes struggle with the otherwise awesome Xbox 360 Wireless Microphone. It has only a power button and some lights.

In its defense, for the most part it does just work when you pick it up and start singing (badly, in my case), but I admit to being slightly perplexed every time I have to sync it with an Xbox, or figure out what's going on with it. Can you blame me?

When you turn on the microphone, the built-in lights shine to display the microphone status as follows:

• Power on: lights flash green one time every second
• Connecting: lights flash green four times every second
• Connection complete: lights flash blue, and then stops

When your battery power is low, the built-in lights shine to display the battery charge status as follows:

• Low: Lights flash amber one time every three seconds
• Critical: Lights flash amber one time every second

When your microphone moves out of the wireless range of your console, the lights flash green one time every second. The lights can also change color together with supported game titles.

If we can agree that no buttons is clearly a bad idea, I think it follows that one button is problematic in its own way. I have the same issue with the single button on the iPhone that I do with the single button mouse – it may be OK-ish at the very beginning, but over time it leads to absurd, almost comical overloading of functionality. Consider how many different things the single button on the face of an iPhone now controls:

(diagram courtesy Andrew Durdin, source)

The iPhone home button? Why, it's easy! You have your choice of…

• single-click
• double-click
• triple-click
• click and hold
• click and pause and click again

All of which have different meanings at different times, of course. In particular I spend a lot of time double-clicking to get to the active apps list, and I often mis-tap which kicks me over to the home screen. I have so many apps installed on my iPhone that search is the only rational way to navigate. This means I search a lot, which requires clicking once to get to the default home page, pausing, then clicking again. Sometimes I click too long, which is then detected as click-and-hold, and I get the voice search app which I am … er, not a fan of, to put it mildly.

I've gotten to the point where I dread using the home button on my iPhone because it Makes Me Think. And I get it wrong a significant percentage of the time. This isn't the way it's supposed to be.

You might be expecting me to turn into a rabid Windows Phone or Android fanboy about now and snarkily note how they get it right. I'm not sure they do. Either of them. They all manage to suck in their own special way.

When there's one button on the device, at least it's clear what that button is supposed to do, right? Well, sometimes.

There is one theme I agree with here – the clearly marked back button on both Android and Windows phones, just like a web browser. I mostly use my iPhone as a platform for the Internet, and the simplicity of the Internet is its primary strength: a collection of obvious things to click, and an easy, giant honking back button so you never get lost in its maze of twisty passages, all alike. It is true that browsers have a home button, but the latest versions of Chrome, Firefox, and Internet Explorer have all but pushed that home button off the UI in favor of the ginormous back button. While I'll tentatively agree that not all phone apps have to behave like the Internet, the Internet is becoming more and more of a platform for bona fide applications every day. The back button is a UI paradigm that works like gangbusters for webapps, and I'd argue strongly in favor of that being a hard button on a device.

But once you add three buttons, thinking starts to creep in again. Am I pressing the correct button? That's never good. And I don't even know what that third button is supposed to be on the Android phone! I could possibly be in favor of the hard search button on the Windows phone, I suppose, but I'd rather see good, consistent use of two buttons on the face of a device before willy-nilly adding Yet Another Button. I think there's a reason the industry has more or less standardized on a two-button mouse, for example. (Yes, there is that pesky middle button, but it's a nice to have, not an essential part of the experience.)

What about the one finger solution? Even with touch devices, one finger does not seem to be enough; there's a curious overloading of the experience over time.

On the iPad, there are a number of system-wide gestures, such as swiping left or right with four fingers to switch between apps. Four-finger swipes? That's convoluted, but imagine a virtual mixing console with horizontal sliders. Quickly move four of them at once...and you switch apps. Application designers have to work around these, making sure that legitimate input methods don't mimic the system-level gestures.

The worst offender is this: swipe down from the top of the screen to reveal the Notification Center (a window containing calendar appointments, the weather, etc.). A single-finger vertical motion is hardly unusual, and many apps expect such input. The games Flight Control and Fruit Ninja are two prime examples. Unintentionally pulling down the Notification Center during normal gameplay is common. A centered vertical swipe is natural in any paint program, too. Do app designers need build around allowing such controls? Apparently, yes.

Yes, our old friend overloading is now on the touch scene in spades: for all but the simplest use of a tablet, you will inevitably find yourself double-tapping, tapping and holding, swiping with two fingers, and so on.

SOPA and PIPA are two pieces of proposed legislation designed to "stop" Internet piracy… in the most hamfisted way imaginable. As Mitchell Baker explains:

Assume there's a corner store in your neighborhood that rents movies. But the movie industry believes that some or even all of the videos in that store are unauthorized copies, so that they're not being paid when people watch their movies. What should be done?

SOPA/PIPA do not aim at the people trying to get to the store, or even the store itself. The solution under the proposed bills is to make it as difficult as possible to find or interact with the store:

1. Maps showing the location of the store must be changed to hide it.
2. The road to the store must be blocked off so that it is difficult to physically get to there.
3. Directory services must delist the store’s phone number and address.
4. Credit card companies would have to cease providing services to the store.
5. Local newspapers would no longer be allowed to place ads for the video store.
6. To make sure it all happens, any person or organization who doesn’t do this is subject to penalties. Even publishing a newsletter that tells people where the store is would be prohibited by this legislation.

Just substitute "corner store" with "website" and I think you can see where this is going. These bills are so rife with potential for abuse and misuse, so clearly dangerous to the very fabric of the Internet, that frankly I have a hard time getting worked up about them. The Internet is under constant siege by large companies, and will be for the forseeable future. This is nothing new. These bills will be defeated, because they must be.

Instead, I'm scratching my head and wondering how such boneheaded bills made it this far in Congress. I can think of a few reasons:

• Average people don't understand how the Internet works and thus can't comprehend the danger.
• Nobody pays attention to what our government does until it hits them in the pocketbook (or below the belt).
• These bills were pushed through by highly paid lobbyists for the entertainment industry.

I often bemoan the state of Slacktivism on the internet, where changing your Facebook or Twitter picture is considered a valid and effective form of protest. But this time, I am happy to say, was indeed different.

Perhaps because of the obvious danger of these bills, geek websites and communities banded together weeks ago to protect themselves and the greater Internet. Like many other technical communities, we wrote about it on our blog, talked about it on our podcast, and even put up a little banner on Stack Overflow for a day. Users were encouraged to call, fax, and write their representatives in Congress and express their concerns. And they did, in droves! But outside of our technical geek ghettos, there was precious little mainstream coverage of this dangerous legislation.

That is, until major sites like Wikipedia, Google, and Craigslist joined the bandwagon today. Most notably, Wikipedia actually went dark for all of today, January 18th, rendering all of English language Wikipedia inaccessible. That turned the tide, and transformed SOPA/PIPA into something that average people would actually talk about and care about. There's no better way to raise awareness of the danger these bills pose than blacklisting one of the greatest resources the Internet has ever produced.

While SOPA/PIPA are still alive -- barely -- for now, I think it's safe to say that they are well on their way to defeat. I'm glad to be a part of that, however tiny, but I cannot in good conscience celebrate.

Yes, we likely succeeded in defeating these two specific bills and galvanizing the political will of major Internet communities, including our very own Stack Exchange. These are good and noble and just and necessary things. They are things to be proud of. But instead of celebrating, let's take this time to reflect and ask ourselves a deeper question: how is it that these dangerous bills came to exist in the first place?

First, and again, this is a critical battle to wage and win. SOPA is just the latest, but in many ways, the most absurd campaign in the endless saga of America’s copyright wars. It will be yet another failed attempt in a failed war, and I obviously believe it should be opposed.

But second, and as you describe, this isn’t my war anymore. Not because my heart isn’t in it, but because I don’t believe we will win that war (or better, win the peace and move on) — even if we can win battles like this one — until the more basic corruption that is our government gets addressed. That’s the fight I have spent the last 4 years working on. That’s where I’ll be for at least the next 6.

For this is what I know: We will never (as in not ever) win the war you care about until we win the war against this corruption of our Republic.

Of course, as my book, Republic, Lost: How Money Corrupts Congress -- and a Plan to Stop It, describes, this is an insanely difficult, possibly impossible, fight. But whether difficult or not, it is the fight that must be waged.

We have done much. But in our celebratory enthusiasm, please take a moment to hear out Mr. Lessig, and appreciate just how far we have yet to go.

So yes, join us in fighting the obvious insanity of legislation like SOPA and PIPA that threaten the open, unfettered Internet. But please, please also join us in attacking the far more pernicious problem of lobbyist money subtly corrupting our government. If we don't deal with that, we will never stop fighting bills like SOPA and PIPA.

In November, I delivered the keynote presentation at Øredev 2011. It was the second and probably final presentation in the series I call Building Social Software for the Anti-Social.

I've spent almost four years thinking about the Q&A format, and these two presentations are the culmination of that line of thought. In them I present ten "scary ideas", ideas which are counterintuitive for most folks. These are the building blocks we used to construct Stack Overflow, and by extension, Server Fault, Super User, and the rest of the Stack Exchange network.

1. Radically lower the bar for participation.
2. Trusting (some of) your users.
3. Life is the world’s biggest MMORPG.
4. Bad stuff happens.
5. Love trumps money.
6. Rules can be fun and social.
7. All modern website design is game design.
8. Thoughtful game design creates sustainable communities.
9. The community isn’t always right.
10. Some moderation required.

It's not the same experience as attending the actual live presentation, of course, but you can certainly get the gist of it by viewing the slides for these two presentations online:

The Øredev organizers hired ImageThink to draw each presentation on a whiteboard live on stage as it was presented. I was skeptical that this would work, but the whiteboard visualizations came out great for all the presentations. Here's the two whiteboard drawings ImageThink created during my presentation. (Yes, they had two artists on stage "live whiteboarding", one on the left side, and one on the right side.)

It's not a bad approximation of what was covered. If you're curious about live whiteboard visualizations, ImageThink posted a great set of links on their blog that I highly recommend.

After four years, we've mostly figured out what works and what doesn't work for our particular brand of low noise, high signal Q&A at Stack Exchange. But the title Social Software for the Anti-Social is only partially tongue in cheek. If you want to learn anything online, you have to design your software to refocus and redirect people's natural social group impulses, and that's what these presentations attempt to explain. I hope you enjoy them!

Update: Part II is now available as a full talk, with audio and video courtesy of Oredev. Watch it now!

Discussion