Three strikes and you're out is a good rule for baseball and startups. In the last six weeks, Zuckerberg has committed three management fouls that would have led to any normal CEO's firing.

The terms of service disaster. After Facebook revised its legalese to suggest it would keep users' information forever, even if they deleted their accounts, users got into an uproar. Rather than quickly roll out new language that satisfied their concerns, Zuckerberg introduced an absurd voting mechanism that would let Facebook user, in theory, vet future changes.

Facebook's awful redesign. After years of pursuing his own vision for Facebook and ignoring competitors' moves, Zuckerberg took his eye off the ball and grew entranced by Twitter, a message-broadcasting service which mimicked Facebook's status-update feature. He then introduced a new design for Facebook that was almost universally loathed, even among his own employees. And unlike the terms-of-service debacle, where Zuckerberg championed a user-run democracy, he told employees that they should ignore feedback from users.

Management chaos. The disgraceful, petty ouster of Facebook CFO Gideon Yu, a veteran of Yahoo and YouTube, on Tuesday, apparently prompted by a disagreement over strategy, is just the latest executive-suite blunder by Zuckerberg. He has prompted the departure of countless key employees over the past couple of years. At a time when he should be pulling the best talent in through its doors, Zuckerberg is sending it the wrong way. Facebook started with five cofounders. Zuckerberg alone remains.

Facebook first tried to claim that it fired Yu as CFO because it needed someone with public-company experience. (Never mind that Yu's resume includes Hilton, Yahoo, and Google.) Realizing the absurdity of that spin, Zuckerberg is now trotting out the line that Yu is "spending more time with his family," and trying to claim it's not a cliché.

It is understandable for a 24-year-old to be fickle, easily swayed, and vengeful. But that is the reason why we have very few 24-year-old CEOs running companies with 800 employees and hundreds of millions of dollars in revenues. The company is now conducting a search for Yu's replacement. But it ought to be looking for Zuckerberg's replacement instead.

Zuckerberg did a good job as Facebook's CEO during its startup years. He brilliantly conceived of Facebook's product, and pursued that vision unswervingly until recently. His hacker cred helped pull in likeminded engineers. When the Facebook application platform threatened to drown the site in spam, he reeled the appmakers back in, despite their howls of protest, proving he can sometimes correct his mistakes. But he has reached the limits of his abilities.

He has said that his most important job as CEO is to build the company's culture. And in that he has been an unmitigated disaster. It is a failing far worse than his three public strikes — though the depth of it is little understood outside Facebook's offices in Palo Alto. His executives are largely unhappy, consumed by infighting and backbiting while trying to accommodate a tempestuous boss's whims. Here's the Harvard man in Zuckerberg revealed: He has assembled a coterie of servants instead of building a team.

So what is Facebook's board to do? Firing him would be difficult. Zuckerberg owns 27 percent of the company and controls three of Facebook's six board seats. He could easily fight any attempt to oust him. And he is not unsalvageable. The right CEO could help Zuckerberg correct his failings and prepare him to take back the reins.

We think Netscape cofounder Marc Andreessen is the one Zuckerberg needs. Already a mentor to the young entrepreneur and a Facebook board member, Andreessen has few undiscardable obligations. He has started a venture-capital fund, but that is nothing more than a few business cards to shred. Facebook could buy Andreessen's also-ran social-networking startup, Ning, to remove that excuse, using inflated shares to give him a graceful exit. His services would be worth the price. (Or Andreessen could simply return what's left of the $60 million it raised last year to investors, an honorable way to give up on a company that's going nowhere.)

It would be difficult to pull Andreessen out of semi-retirement. At his companies — Netscape, Opsware, and Ning — he has preferred to let someone else hold the CEO title. But it might be good for Facebook to have a CEO who's indifferent to the job, as opposed to someone who wants it all too badly, but doesn't know how to do it.

On Microsoft:

Microsoft can build software, when they choose to.

On investing in startups:

I usually put in $25,000 to $100,000 per company. My philosophy is to put in a small enough amount of money that I won't get mad at the founder if I lose it.

Translation: Marc Andreessen is so rich that he can lose $100,000 and feel nothing.

On the failure of Friendster:

Friendster was very restrictive on what users did. You were supposed to connect because you know each other in real life, not, as [founder Jonathan] Abrams said, 'because you both like Reese's Peanut Butter Cups.' But sometimes you want to put your chocolate in her peanut butter.

Yes, he really said that.

On his deathwatch for the New York Times:

I don't want to become the crazy anti-New York Times guy. You have to do what Intel did in 1985. The Japanese chipmakers were killing Intel in the memory-chip market. It got out of memory chips and focused on the much-smaller microprocessor market. I would turn off the printing presses.

On his mentor and Netscape cofounder, Jim Clark:

I could tell you a lot of stories about his life [in Florida], but I won't. He's dating a 26-year-old Australian swimsuit model. I just ran into an entrepreneur who said, "I just ran into Jim Clark at a resort town in Italy. Jim was in a hot tub carved into the side of a mountain." I said, "Yes! That was Jim Clark."

On the iPhone's price:

Give it a year, it will be down to $99. Give it another year, it will be free.

On his motives for giving away his money:

My wife teaches philanthropy at Stanford Business School. I would be in big trouble if I weren't hugely committed to it.

On his relationship with Microsoft CEO Steve Ballmer:

He's my Facebook friend. He's my Facebook 'friend.' [makes air-quotes gesture] I'll stop there.

And the Battle of the Sexes is on! In one corner, @michaellamb states the obvious: that the woman is getting the press because she's easy on the eyes, not because she's competent. @kimbjo wades in and shows her great vocabulary with this zinger: "And enough woman bashing you misogynist misanthrope." Oh, and for the less literate, she has just accused him of not JUST hating women, but hating ALL humankind! @leahculver joins in that said lady is edu-muh-cated, unlike most Valley CEOs????? (That's a story all by itself, Owen!) Oh, and she can't resist calling him a "jealous sexist asshole." @kimbjo also can't resist comparing the WidgetLab guys to a "disgrunted ex boyfriend," a high school one no less. (You don't have fond high school memories, then?) @skycut then confuses the issue by calling Gina a GUY (perhaps this is a creative attempt at staking out neutral territory). @michaellamb, undaunted by this very serious drubbing from the chicks, comes back and basically says, it isn't that she's a WOMAN, dumbasses; it's that she SCREWED UP. And @emnem follows up with the most beautiful, detailed heartfelt rant against feminism I have ever had the pleasure of reading. To which @raincoaster rejoins that she doesn't fuck her boss and none of her friends do either, and that @emnem must patronize two-bit whores. And @michaellamb makes one last plea: it's what she did, is anybody listening?

Terrific wank; good job everybody, and it's a very sad day when I have to satirize the Valleywag commenters. Please don't make me do this again. Thanks.

• Letter 1, August 2 From WidgetLaboratory cofounder Spencer Forman to Ning CEO Gina Bianchini: Widgetlaboratory wants to know changes coming to Ning before they happen and to not be blamed when things go wrong.
• Letter 2, August 2 From Bianchini to Forman: Ning and Bianchini want to talk on the phone clear up any "conspiratorial" thinking. "We just want you to succeed in a way that scales. Time and time again it feels like you are trying to threaten us into something that is never exactly clear." Let's work together if we can, if we can't let's move on.
• Letter 3, August 2 From Forman to Bianchini: We have 1,700 networks and millions of users, when we fail you fail. "Considering the fact that we are the only Network that provides any real products to your customers on the Ning "platform," do you really think we are being unreasonable to believe that Ning might keep us notified before you decide to pull the plug on using Dojo [a software toolkit used by JavaScript developers] in the header of every page?"
• Letter 4, August 3 Bianchini to Forman: I'm happy to talk on the phone, but the sniping has to stop.
• Letter 5, August 3 Forman to Bianchini: "Let's get to work."
• Letter 6, August 3 Bianchini to Forman: BTW, you were right we should have let you know about Dojo. Our bad.
• LetterLetter 7, August 7 Bianchini to Forman: Good talking on the phone. No we can't always alert you to when we're about to pull one of your widgets. No you can't ask your users their username, passwords or pins.
• Letter 8, August 7 Forman to Bianchini: No, please call us before you pull our widgets. Even at 3 in the morning. We have a million users! We're not phishers, please let us ask our users for passwords.
• Letter 9, August 7 Bianchini to Forman: Argh, I can't handle this anymore, I'm delegating.
• Letter 10, August 22 Ning general counsel Robert Ghoorah to Forman: You've been removed for TOS violations.
• Letter 11, August 22 Forman to Ghoorah: Our lawyers say: WTF? You can't do this.
• Letter 12, August 22 Ghoorah to Forman: You were booted. "Use of Ning is a privilege not a right. We do not intend to debate our decision."
• Letter 13, August 22 Forman to Ghoorah: Please, therefore, provide "any" specific details as to the "unduly degrading" of your network.
• Letter 14, August 22 Ghoorah to Forman: Your code breaks all the time. We called you last night about it. You were mean and unhelpful.
• Letter 15, August 22 Forman to Ghoorah: It took two minutes to fix the problem when you finally called at 3 a.m. last night.

An enormous, highly visible brand may not want to risk seeing its ad wind up on a page such as that run by the actual Facebook group "I've Had Sex with Someone on Facebook," which at press time had 59,353 members. Or consider the MySpace profile (turned up after about two minutes on the site) of 18-year-old "Nikki AKA Death Angel!," which is adorned with the motto "Don't fuckin fuck with ninjette bitch we'll cut ur fuckin head off an give it to ur momma."

When spending the majority of their time browsing content like this (or, more likely, content like this slightly more relevant to their friends), what are users thinking about? Checking out an ex's profile, we're more likely to remember the photos of the new sweetie, and not the "Last Minute Cabo Deals!" enticement next to it. If anything, that's salt in the wound. This new argument goes, if advertisers could sell based on users' messy passions, we users will stop playing Scrabulous — or dreaming of getting back together — and pay up. Sex does still sell.

I think exits are much bigger question mark today then they were 6 months ago. Think of the exits Ning, and Brightcove have to get to make their investors happy. It's a daunting, daunting number.

Of course, Miller happily ignores his own advice. Spark just valued Twitter
— a company with no business model and no reliability — at just under$100 million, leading a $15 million funding round. This guy's almost as funny as his comedian namesake.

Previously:

• A is for Adelson, who cofounded Digg

Because Zuckerberg doesn't have to. Microsoft owns 1.6 percent of Facebook; Li, even after doubling his take, only 0.8 percent. Neither stake is large enough to merit a board seat. Andreessen is, like Thiel, the former CEO of PayPal, an entrepreneur-friendly choice; he bypassed Sand Hill Road altogether to raise Ning's $100-million-plus in funding.

Just yesterday, we'd heard that Zuckerberg, who owns 27 percent of Facebook, had the right to appoint two board members. That leaves him one more seat at the table to fill. Anyone want to take odds on the moneymen getting left out once again?

Andreessen's conclusion is worth quoting in full:

We are learning that hostile takeovers have arrived in our industry. This is the second major hostile takeover so far — the other was Oracle's takeover of Peoplesoft — but there will be more.

This is significant because historically hostile takeovers practically never happened in technology. Potential hostile acquirors assumed that hostile takeovers wouldn't work because the target company's employees would bail and the target company's business would collapse.

It turns out that as technology companies become larger and more mature, acquirors are becoming increasingly convinced that neither of these assumptions hold. Perhaps employees of large tech companies aren't that bonded to current management, and perhaps many of them would actually prefer to work for a larger, more dominant combined company. And maybe as a consequence, the target's business would do just fine in the wake of a hostile takeover — in fact, maybe it would do better, due to advantages of combined size and scale.

My bet is that hostile takeovers, particularly of larger and more mature companies, are going to become increasingly common in our industry.

The excitement may be just beginning.

What has changed in the near-decade since then? Yahoo, which grew up alongside Netscape — at one point, Netscape hosted Yahoo's servers — is that much farther from being a startup. Working there offers less risk, and less reward. Andreessen doesn't come out and say it, but he strongly suggests the place has become infested with careerists who would be just as happy working at Microsoft.

After the Netscape acquisition, Andreessen worked briefly and unhappily as AOL's CTO. For Yahoos, wheeling and dealing may be fine; but for him, it's the startup life or nothing. Andreessen may feign nonchalance at the prospect of more hostile takeovers in tech. But that doesn't mean he personally wants any part in them.

A recent Fast Company article on Andreessen's current venture, Ning, went no better. You can practically hear the writer propping his eyelids open as Andreessen goes on, and on, and on, about "viral expansion loops."

What happened to the Andreessen who once ridiculed Windows as "a set of poorly debugged device drivers"? Why, he's gone online. Andreessen's blog is relentlessly entertaining. His verbal fisticuffs with the New York Times are must-reads; the vitriol oozes out of every line. And he posts just infrequently enough to keep us hanging on every word.

The only surprise, really, is that Andreessen took so long to start blogging. This world was not made for him. In the Web, he created one to suit.

(Photo by mathoov)

Ning does have people in the Valley, as Fast Company claims, "buzzing," but not because of the "viral expansion loops" which Andreessen talks up in the piece. Penenberg's thesis: Andreessen has fused viral marketing with social networks, and therefore Ning's current fast expansion rate will continue ad infinitum, or at least ad acquisition.

This is a fashionable delusion fostered by people with something to sell. Supporting Andreessen's argument are Union Square Ventures' Fred Wilson and Sequoia Capital's Roelof Botha, both of whom make the argument for compound growth. Wilson is an investor in Twitter; Botha backed YouTube. Both profit from the notion that a site's current growth rate will continue unchecked.

The reality? Growth always slows. Facebook used to crow about how its user numbers grew 3 percent a week. By the time Microsoft sank $240 million into the company, that figure had already dropped; it may now be around 1 or 2 percent. Still impressive, and still fast-growing — but any projections based on 3 percent weekly growth are now dead wrong.

With absurdities about compound growth and viral expansion stripped out, Penenberg has little to offer in Ning's defense. According to figures in the piece, Ning is making roughly $1.7 million a year in the $20-a-month subscriptions some social-network creators pay. The rest of the money they make comes from Google's AdSense ads, the familiar fallback of hopeless startups. Bianchini admits as much in a blog post. And yet she and Andreessen commanded a $214 million valuation for their creation.

What Penenberg doesn't explore: The laughable reputation of Ning's software within the Valley. The piece quotes exactly one Ning user. Had Penenberg asked around, he'd have heard from scores of disgusted social-network creators who walked away from the service after trying it out. Pointing that out would get in the way of discussing the appearance of Ning's creators. Really, Adam, I thought that was our job.

(Photo by Fast Company/Art Streiber)

That spree could well end in tears. But that's not the Valley's problem; we make companies to sell them. Especially Andreessen who, despite his protestations of not building companies to flip them, is surely eager to unload Ning, his social-network startup. What better way to attract potential buyers than to butter them up with a post telling them how important they are? In the comments, I'll take odds on Andreessen selling to one of the companies he named within the year.