The story behind my Platforms Rant

Hey, everybody. Welcome back to Stevey's Tech Talk. I have another story time for you today. The story today is going to be about this blog post that I wrote back in 2011, which went around the world like a bullet. And people still talk about it today. It was my Jerry Maguire moment — some journalists actually did say that. And I don't think I've ever properly told the story of how it happened, and where my head was at the time, and the aftermath and the fallout.

I know I'm talking about something that happened 10 years ago, but I still hear about it every week. A buddy will say, oh, our CEO mentioned you on stage during an all-hands last week. I had another buddy who went to work for the NSA, and one of them told him in his first week there, we have a Bible here — and he pulled out my post.

It had a huge, huge impact. It was intended to have a huge impact, but it was legitimately intended only for a Google audience. There are telltale signs in the post that it was for a Google audience — I use some Google terminology that clearly I would have explained or changed had I meant for this to be for an external audience. So I just wanted to tell the story of how it happened, because it is pretty crazy.

So I am a platforms guy. My first job out of college was an operating systems company called GeoWorks, and an operating system is a platform. A platform is just a computing environment that provides an all-inclusive set of facilities for you to write programs that can talk to the network, talk to the disk and store data, or talk to other processes. A platform has a programming surface. It has APIs. A platform is something that invites you to program it, as opposed to a product, which is a piece of software that you're supposed to use but don't write programs for.

And so I was a platforms guy. At Amazon, we wound up building their web services platform, their service-oriented architecture, which ultimately paved the way for Amazon Web Services. And I was in that small core group of people that designed and built that and migrated people to it. All the while I was using tools like Emacs that are platforms in their own right, and I was writing a game that was a platform. Platforms are everywhere. They're ubiquitous. They're not that hard to write and they're fairly well understood.

Building a big distributed platform with multi-tenancy, so that you can have Home Depot and Disney and Starbucks all running their computing in your data centers? That's work. There's a lot of nuance and difficulty, and it's certainly not easy. This is why Google Cloud Platform and Microsoft Azure are having so much trouble keeping up with AWS. But by and large, people have been doing platforms since the seventies, at least.

So I didn't think it was any big thing — except at Google, they didn't do it. It just wasn't really a thing. Everywhere you looked, everybody was building a thing that was going to bring in billions of dollars or millions of users. That was their yardstick, their metric for whether they would work on something or not. But they never thought about developers. They never really thought about programmers. That was completely secondary to them. It was just not in their DNA. And to this day, they still don't seem super platform-oriented.

And back in 2011, when Google Plus was starting to become this big thing, I was mad. I was actually mad about Google Plus in about nine or ten different dimensions. The person behind Google Plus — let's just say who it was, it was Vic Gundotra — showed up from Microsoft and wrapped Larry Page around his finger. And Google Plus was 100% his idea, his implementation, his design. And he was doing a terrible job of it. And we all knew. So for those of you who weren't at Google at this time, and you're like, how could Google have done that — believe you me, we were all asking the same question.

There was an internal meme generator, and everything he posted got negative 2,000 or so downvotes, give or take. He was reviled. He was loathed. But he also was screwing up the company, with the whole triple threat. He was screwing up our brand. He was screwing up our culture. And he was screwing up our tech, our codebase, by forking things, shutting projects down and merging them into his project, and making other terrible decisions.

Google was very egalitarian — very meritocracy, very all-voices-matter, speak-truth-to-authority. And they meant it. Vic was trying to change all that. We'd gang up with the senior people and go make our case, and he treated it like he was royalty. That was the culture damage I was talking about.

So I was specifically very upset about a number of decisions that he had made. Ten years have gone by, so I would have to think hard about what all ten reasons were, but I can remember a couple of them. One of them was Real Names. I actually did an internal blog post on how stupid of a decision it was to use Real Names instead of letting people use fake names — have Google know the real person behind the account, maybe require some identification if they want certain elevated privileges, but by and large hiding people's names was super important. If you looked at Slashdot or any of the other forums that encouraged people to talk to each other, they were anonymous forums. And that was a huge differentiator over Facebook, which was using Real Names.

But he was like a little kid. He saw Facebook and he's like, I want that. And so he did everything he could to make an alternate Facebook. And he did it through fear, very much the way any politician that runs on a fear platform will do. He scared Larry into compliance by saying, oh, Facebook's going to kill us, Facebook's going to kill us. Just Wormtongue, whispering in Théoden's ear over and over again.

And he wasn't wrong — Facebook was doing some real damage in terms of eyeballs and advertisers. But at the time he was doing it, he was doing it wrong. Real Names was a stupid, stupid idea, because first of all, Facebook had already cornered that market. And second of all, that's not really what people wanted. People wanted to be able to have an avatar, have an identity, a secret alter ego — somebody that they could be online that wasn't them, so that they could say what was really on their mind. And yes, you can say, oh, when people are anonymous, they say awful things. Well, people aren't anonymous on Facebook, and you're saying people don't say awful things on Facebook? So that argument just falls apart.

So I wrote an internal blog post, and it didn't get a lot of traction because it was internal. I was the one guy speaking up against Google Plus, which was a no-no — which was already a sign that Google had turned a corner in a really bad way, because they would not allow people to speak evil of Google Plus. Just to criticize it. And that was scary, terrifying in its own right — that we were all getting behind this initiative that I knew was going to fail.

And the series of posts that I had planned out included one on platforms, and one on comment threads. Google Plus wasn't a platform, and I knew it, because when they first came out with it, I was like, oh, I'm going to write some cool plugins. I'm going to write some games, some quizzes. And they didn't have one and they didn't want one. It didn't even occur to them that it would be an important thing. They looked at Facebook and all they saw was, oh, shiny — look at all the people talking to each other. They didn't think about what Facebook had put into it to make it so that all of those distractions were available. There were huge games — FarmVille, which was the Animal Crossing of its time. And there was no opportunity for anybody to build anything like that on Google Plus.

I was also planning on doing one on comments. Look at how well Reddit handles comment threads with 17,000 comments in them, and look how horrible Google Plus's comment interface was. We had our own internal version of Google Plus that looked identical to the external version — this becomes relevant later. Once you got around 50 comments, the engineers would start complaining that they couldn't keep track of all the comments. There's no concept of threads. There's no comment replies. There's no indentation. Nothing that Reddit had already had down for years and years. They were just doing YouTube-style comments — just a big stream of comments. What good is that? There's no discussion there. Google engineers would get confused once we hit 50 to 80 comments in a thread, and yet on Reddit the same day, somebody's 97-year-old grandma is in there reading all the comments — admittedly with assistance — and you get 20,000 people commenting and somehow it manages to all stay manageable to humans. People deal with trees really well. And it was really obvious that that's the way we should have done it.

So Real Names, platforms, how we did comments and comment threads, and there were like seven others. I was going to write an 11-part series that ended with a final F-you, which was: you are doing this wrong. You are wrong, wrong, wrong, and it's going to fail. A lot of people agreed with me. Would it have made any difference? No. At best, it would have gotten me fired if I had managed to continue swimming upstream like that, criticizing them for the decision the company had made. Because as a leader in the company, you're supposed to disagree and commit. You're supposed to get in line. And if this ship is too big to sink, then you better just get on the ship and don't say anything about it sinking and not having enough lifeboats. And that's the kind of company that Google turned into overnight when Vic took over. And Google has never really fully recovered from that multi-year episode.

And I recognize that I'm burning bridges right and left, but I still care a lot about Google. I still have friends there, though many of my friends have left because Google isn't the place that it used to be. I think Google did a lot of things right — I could do whole episodes on the things that they did right. And yet they let that happen to them, and I was so angry.

So the platform rant was baking in my head for some time. My modus operandi for my blogs has been, I stew on things for months. I mull and I stew and I let it simmer — where I'm not full-on screaming angry, but I'm kind of mad. And then once enough time has gone by, I sit down and just cut loose and write for three hours and edit for three hours or whatever.

The thing that actually pushed me to write the post was, I was down in Mountain View, visiting some teams, and the code search team was there from Munich. Code search was a search engine for Google's codebase, and I had a knowledge engine for Google's codebase, and we married them later. But I was in an argument with them, because what I wanted to do was write an Emacs front end for code search, and all they had was the web front end. They had put a lot of work in, and in their minds the service offering and the web portal were one and the same. This is something we're offering you, take it. So you're going to work with a browser.

I mean, come on — any programmer listening to this probably thinks this is really obvious. If you have a service that can give you results back, I give it a search term, let's say I'm looking for Dijkstra's algorithm, and code search would give me a ranked list of results back based on their ranking engine. But I had to do it through the website. So if I was in the middle of coding in an IDE, whether it's IntelliJ or Emacs or a command line, I have to interrupt my workflow to go pull up the code search page, do my search, copy the results out, and put them where I need them — when what I really wanted to do was have my program just call it, get the results, do the post-processing, give it to me. That's a really simple, obvious request. And they were like, nah, we don't really want people to do that, we just want people to go through the browser. And I'm like, what about T.V. Raman? The blind engineer who only goes through Emacs for everything — how does he use code search? They just didn't care. They were so apathetic about it. And it was one of those really frustrating situations where there was nothing you could do about it. They had cornered the market on code search — you couldn't write your own, or you're competing with them. They didn't understand the difference between a product and a platform. And it broke something in me.

What do you do when your colleagues at the place where all the smartest people in the world work are stupid? Well, I'll tell you what I did. You drink a lot and you rant.

It was the end of a long day, and I was tired and miserable and heated. And Ambrose Feinstein, who's truly one of the smartest people I've ever worked with, comes walking by and he's like, yo, Steve, what's up? And I'm like, have you ever had one of those days, Ambrose, where you feel like the whole world is against you? And he goes, no — you all right? I'm like, yeah, I'm fine. So you can thank Ambrose for my rant.

I left for the day and went back to my hotel room. I got a bottle of wine, because I was drinking a lot back then — I've quit, I quit five years ago. But I got a bottle of wine and I was just like, I'm just gonna write a love-hate letter to Google about this whole situation, because Google Plus has been bugging me, and this code search team — what the hell? Why are they so damn stupid? I was so angry. And I started drinking and I started typing, just unload everything. I typed for a couple of hours and then I started editing — going through and rereading it and drinking and drinking, and I got down to the bottom of the bottle and I was like, oh, that'll show them. I knew that my jokes were going to be hitting home because I was laughing at them. It's a good litmus test. And I was going to post it on Google Plus. That was going to be the ultimate insult — right on the internal Google Plus.

So just before midnight, it's ready to go. I've proofread it and everything. There was one actual bona fide typo in it, where I somehow stuck an extra G in "army" when I was talking about Rick Dalzell. We should probably get that G removed, but whatever — it added some flavor. If you read the post, you'll see that it was a little bit of a flavorful G.

So I went to publish it. I was used to Blogger, or Medium — share what's new, here's a little box where you can say, here's your post. And I'm on the internal site, going from tab to tab, and I couldn't find it. I'm drunk and tired, like, how do you post on this stupid platform, or whatever this thing is? And finally I found that it was, like, "what's new," some little box up in some corner somewhere. And I was like, oh, I get it, this is where I'm supposed to post. So I copied the text out of Emacs and put it into the what's-new box, and I looked it over and I'm like, okay, bang. And I sent it.

Now, what had happened is, during those 20 minutes where I was figuring out how to post it, I had somehow navigated from my Google account, stevey at Google, to my personal Gmail account — without noticing. And look, 10 years later, that still happens today. If you're not using Chrome profiles, the browser gets really confused about who you are.

So it was kind of natural that I would have somehow gotten logged into my personal account — except that all of the smart people at Google who built smart products like Gmail and YouTube and search, their internal ones would say sandbox.google.com, or they would say "internal Google," blah, blah, blah. There was some indicator that you were in the sandbox and not in the production live one. But Google Plus didn't have that. So I went and posted, and the last thing I thought before I fell asleep was, that'll show them.

About 40 minutes later, I got a phone call on my hotel room phone — not on my personal phone, but on the hotel room landline. I wake up, and I'm completely drunk, practically passed out. And I pick it up and there's this guy talking on the other end. And he's, oh, Steve, so sorry to bother you, this is so-and-so from Sweden, I was talking to the guys in Luxembourg and they're saying that the London PR folks — and I'm like, what? What? And he says, did you know your post went live publicly?

And in slow motion I'm like... noooooo. And I run over to my computer. My friend Nikhil Gore was the first one to comment, and people are commenting. But Google Plus was set up so that sharing something sent a copy of it. There was no undo in Google Plus — yet another one of its really, really good decisions. And there was no way to roll it back. The cat was out of the bag, and once it was out, it was out forever.

At that point it was all damage control. And I was talking to PR people for the rest of the night. I was very contrite, like, hey, I'll do whatever I can. I was on these war-room calls, and every once in a while I'd be like, is this really that big of a deal? And there'd be silence, and they'd be like, yes, it is a big deal. And I'd be like, yeah, but what's the worst that can happen? And they'd be like, you can't even imagine what's going to happen.

And they were right and I was wrong. It wound up on the front page of the Wall Street Journal. It was a big, big deal — look at this Google employee sticking it to Google Plus. And it was extra funny because everybody already knew Google Plus was doomed to fail. They already knew. So to hear a Googler saying it, and saying funny things, it really tickled a lot of people. The next few days were just miserable. My boss is like, thanks! She was a little mad. She was just like, you realize what you did here? I'm like, yeah, well, I didn't mean to.

Most of the company took it very well — except for Vic Gundotra, who wanted me fired. He wanted my head put up on a pike in front of the Google moat, in front of the portcullis, so that anybody coming to threaten Google ever again would see my head there. He really did want to fire me. But everyone was so scared of him — because these horrible things that happen in politics also happen in companies, and everybody's afraid to rock the boat.

Google really did have a policy where the only way they could realistically fire you and avoid a lawsuit — they didn't want to fire me for the rant. They didn't want to censor me. They didn't want to be like, oh, Google censors their employees. Now we've got double trouble. So they looked for ways. They looked to see if they could fire me for leaking confidential information.

But in my supreme idiocy — because I didn't know this was going to go public — through blind luck, I didn't say anything confidential. I knew about a bunch of confidential projects, and had I even given a whiff of one of them, they would have fired me.

Actually, my first interviewer at Google — the first guy from my Google interview story in the last episode — reached out to me. He goes, hey, Yegge, thanks for not leaking my project. And I'm like, yeah, man, my pleasure. But seriously, I could have very easily talked about some confidential project. And then it would have been over — I would have been fired. It didn't go that way. So I guess I'm lucky that I got to stay at this horrible place.

And it actually became a less horrible place in time, when Vic went to spend more time with his family. People literally sang "Ding-dong, the Witch is dead." He was brazen — he would lie to Wall Street. He'd just be like, oh yeah, lots of people are using Google Plus, look at all these people using Google Plus. And it was because they made you use Google Plus to do a YouTube comment. So it was obviously just garbage — padding numbers, trying to hold the investors off long enough until the thing takes off.

The funny thing is, with network effects like this, they really do take off overnight. It's like a rocket taking off — you can't stop it at a certain point. So the fact that Google Plus wasn't successful in its first week should have been a really good sign that it wasn't going to be successful at all.

The next few weeks were pretty chaotic, and everybody reached out to me. Vint Cerf reached out to me — Vint Cerf, who invented the internet. And he's like, I just wanted to meet you and talk to you and see if there's any way I can help. He said he's good at convening people — that was his strength, and he's right. If I needed him to bring a bunch of heads of state in to talk, he would have done it. Crazy times. In a way it's sort of — I wouldn't say it made my career, but it certainly made it so that there are opportunities for me now, if I ever want to go and do platforms somewhere. And platforms are still a lot of fun.

But it's worth pointing out that the platform rant was one out of eleven rants that I had planned. And it was actually number two in the series — number one never got published externally. And there was more. I had lots more pooping to do on Google Plus, lots more that I never got a chance to. So in a sense we were robbed of all that lovely, delicious anger — but it killed itself. The garbage took itself out.

So that's the story of my Google Plus rant. I was drunk. I didn't know that it was going public externally. And I don't regret a single word that I said in there — except for spelling "army" wrong, and maybe one or two other stylistic things. You can probably find it if you just type "Stevey platform rant" into Google. Something about it resonated with people. It really resonated. There was something about the anger there, and I think the humor, that got through to people and helped them understand that even if they weren't sure why I was so upset about this, I was definitely upset about it.

So anyway, that was a lot of fun to tell that story. Thanks for tuning in, and we will see you next week for Stevey's Tech Talk!