Webware

In the last three years, September has become a busy time for Web start-ups and other new companies looking to make their mark. Warring start-up conferences TechCrunch50 and DemoFall take place within mere days of one another, leaving a wake of more than 100 companies that are launched within just a week's time, all vying for media and consumer attention.

Last year it was even worse, as both conferences happened at the exact same time.

This scramble to get things ready often leaves companies not ready for the users they hope will flock to use their product. So, as a service to you, we've gone through and sorted out which of the products you can use right now. These are sites with open registration, and no special beta or invite requirements.

We've also sorted out which ones are aimed at business users versus consumer use. All of this information can be found in a spreadsheet embedded at the end of this post. But before we get there, let's take a look at the makeup of launches that were open versus closed:

Note: We considered sites that were listed as having "private" or "invite only" betas as closed. This is because there is no guarantee that you could get immediate access once you signed up to use them. For physical products, we counted whether or not you could purchase or download them. All data was gathered Thursday.

This year, there were slightly less products launching at Demo, although that's not including the 14 "alpha pitch" companies, which are neither part of the main program, or by definition supposed to be live at the time of presenting. That said, there was a higher percentage of companies that were live and ready to go in the days following the show than the year prior, which came in at 75 percent compared with last year's 67 percent.

TechCrunch50, on the other hand, stayed around the same as the year before. Last year's show had just 42 percent of the 52 launching companies open for public use, compared with this year's 48 percent.

Worth noting is that there are a few sites that are on the verge of being open but that were not ready to go in time of this article going live:

• HealthyWage, a company that launched at TechCrunch50 will be opening up to beta users on Monday.
• Spawn Labs, the video game place-shifting service that demoed at TechCrunch50 will be available in November. We played with a demo unit at the show and everything worked great. The company is just working on production ramp-up to get it ready for the holiday season.
• Twirl TV, a DemoFall launch is currently open, but will close after the first 10,000 users sign-up. We marked it in this list as open, since we were able to register.
• Weels' site said it would launch on Thursday, however it was having problems with Amazon's EC2 service, and was expected to launch later in the evening.

Consumer vs. Business

Also of interest, and something we didn't do when comparing open and closed companies for last year's shows, is the make-up of products that launched at the show. Were they for consumers or businesses?

Not surprisingly, TechCrunch50 had a better showing of consumer-oriented products and services. That's to be expected though. The show is pitched at this audience and as a place for companies to pitch new software and services. Very little hardware is chosen to present--and what is has historically been for consumers. Demo on the other hand had more than a third of the companies aiming their products at businesses, or business users.

How important is being open?

Tech trade shows are never likely to have a 100 percent live at launch factor, nor should they. Many of the companies that made their debut at these shows are coming out of stealth mode and have worked very carefully to keep information about their service secret, both for a competitive advantage and to iron out last-minute kinks. Others have services that just plain aren't ready for big audiences and need a small and eager test group to help see if their creation scales.

What's often more impressive is how quickly some of these companies that aren't live, go on to sort things out and open up. Although last year that wasn't necessarily the case. Just seven months after the conferences many of the companies that were not yet live or open were still shut (including a few that shut themselves down). Will it be the same with this year's crop? Check back here in six months.

Previously:
Post-launch frenzy: What can you actually use? (from 2008's show)
TC50/Demo revisited: What's alive, what's dead? (seven months later)

Outside the TechCrunch 50 conference in San Francisco earlier this week.

(Credit: Josh Lowensohn / CNET)

SAN FRANCISCO--At some point during it became evident that the Web 2.0 floodgates are no longer open.

Maybe it was when conference co-organizer Jason Calacanis asked one of the panels of judges what they'd thought of a round of pitches from just-launched social-networking start-ups like inbox aggregator Threadsy and photo-sharing iPhone app Clixtr. Sean Parker, the Napster co-founder and former Facebook exec who will be portrayed as a "Silicon Valley bad boy" in the film adaptation of ," leaned his elbows on the onstage table, slouched, and declared, "I'm a little bit bored with social media."

Maybe it was when the TechCrunch50 conference winners were finally announced and the grand prize went not to a slick and shiny app filled with Ajax interfaces and social-media mashups, but to RedBeacon, a mundane-looking local services start-up that aims to offer an alternative to Craigslist and the Yellow Pages if you're looking for somebody to paint your house or cater a party.

Or maybe it was when several Facebook execs took the stage to announce, among other things, that the social network--the subject of perpetual hand-wringing over how it would possibly make money--achieved a cash-flow positive status for the first time in the second quarter of this year, earlier than its 2010 goal.

Web 2.0 has grown up, after three years of investment, start-ups, and media hype, and it couldn't have been more evident at TechCrunch50, a two-day parade of start-up launches that sometimes feels less like a conference and more like a fraternity reunion. By this point just about everybody knows just about everybody else; the launch demos were just as likely to come from established industry players as from hopeful young newcomers.

Not so long ago, the Web start-up landscape was dotted with dozens of small companies with a legitimate shot at getting huge. As recently as last year, it wouldn't have been entirely ludicrous for an ambitious entrepreneur to take the stage at TechCrunch50 and announce that he or she was hoping to build a new start-up into the next Facebook. But the big guys have gotten bigger, and everything in comparison appears to be niche, peripheral tools.

Innovation on the Web these days comes in the form of fine-tuned features and tweaks, not big and lofty new schemes. TechCrunch50's lineup showed that while there are very promising ideas out there, the new stuff is about improving existing concepts, not creating something off-the-wall new. ToyBots, a new Web-connected toy company, takes the kiddie Webkinz craze from a few years ago and infuses it with the thinking behind "hackable" household gadget Chumby. Winner RedBeacon, as well as used-car marketplace Mota and job-hunt site LocalBacon, all pitched themselves as better options than traversing the Craigslist jungle. iMo and Spawn Player are both add-ons for gamers, the former an iPhone controller app and the latter a Slingbox-like place-shifter.

And when something popped up at TechCrunch50 that was pretty darn original, it was met with some restraint. There was plenty of excitement over AnyClip, a new database site that indexes and deep-tags short clips from movies, but the judges rightfully expressed concerns over the difficulty of wrangling with copyrights and content owners.

It's a far cry from the days when, even in the post-Napster era, into music- and video-sharing start-ups that weren't prepared to deal with the intricacies of big media. And likewise, VC dollars were once flooding into start-ups that hoped to be the biggest social network in the world. The economy put a damper on this, for sure, but so did the increasing dominance of the likes of Google, Facebook, and to a lesser extent Twitter and Digg. The big news in venture capital on the Web these days is Twitter's alleged billion-dollar valuation and , not in a huge rush of investors heading for the next big thing on the Web--which is exciting nonetheless, because it wasn't all that long ago that these companies were just as small as those presenting onstage at TechCrunch50.

For now, if we want genuine, holy-crap excitement in the tech industry, perhaps we should be looking at hardware, green tech, edgy mobile innovations like augmented reality, or perhaps even enterprise technology. TechCrunch50 seemed to have the right idea by --but the judges, whose backgrounds were in Web and software investments, admitted that this wasn't their area of expertise.

There were blunt words for some of the companies at TechCrunch50, especially community-based sites that require a critical mass of users to stay afloat; judges seemed skeptical that the social-media fever of the past few years can still pack enough of a punch. "Why would I leave Twitter for this?" asked Robert Scoble of one start-up--the same Robert Scoble who, in fact, , which had impressive technology but little mainstream appeal .

The next big game-changer in social media might be out there already, and we haven't even seen it coming yet. But watching more than four dozen start-up pitches in a row made it pretty clear that most of the biggest splashes of Web 2.0 have come and gone: we simply don't need another news aggregator, another discovery engine, another question-and-answer service, another blogging platform, or heaven forbid, another social network. This is good. It's a sign of industry maturation.

And it's certainly not a bad thing that Silicon Valley's elite finally seem to be catching on to that.

Trying to weed through the 50 companies that launched over the past two days is overwhelming for you, and us. As done in years past, we've picked five of our favorites from TechCrunch50. These are all consumer-oriented services that bring innovative ideas to the table and have a good chance at succeeding.

Note: CNET's judges for this article were Josh Lowensohn (me) and Caroline McCarthy, both of whom watched all 50 on-stage pitches.

Story Something

Story Something is one of the few kid-oriented products we've seen in the past few years that doesn't have some crazy scheme with virtual worlds, virtual currency, or a way to suck kids dry of their hard-earned allowances. Instead, Story Something is set up to create personalized stories for kids that the parent can set up and read with minimal effort.

Much like Mad Libs, parents can insert the name of their child, and things they can relate to (like their friends, or parent's names) into an existing work. The service then goes through existing stories and inserts the names. Children also have the option of picking what happens next if it's a multipart story, akin to the Choose Your Own Adventure series.

Stories can be printed, and e-mailed for reading on mobile devices and e-readers like the Kindle. It even includes illustrations in these copies. As a time saver, the service can also be set up to e-mail you new stories as they come in, using information about your family that you set up in previous sessions.

The service is currently in private beta, and is planning to have a free version with a handful of sample stories, as well as a premium version which includes a full library of content for $3 a month.

ToyBots Woozees

ToyBots Woozees isn't a product as much as a platform. Think of it like Teddy Ruxpin, but Internet-enabled, and available in a whole slew of smart toys. ToyBots wants to get toy manufacturers using their standardized firmware, which lets toy owners swap personalities, and run downloaded software like games and stories.

This whole idea of shared content is hosted on a large network, kind of like Apple's iPhone and iPod App Store, so that games you buy for one device can work for another. It also lets third-party developers program new ways for consumers to interact with that old toy.

As a business this can bring something toy makers don't currently have with most stuffed animals: a revenue flow post-purchase. Instead of relying on accessories like additional outfits, and other characters, they can make money off software sales. Of course before this happens, ToyBots has to get toy manufacturers on board.

ToyBots envisions a future full of toys that share a similar hardware profile, letting users swap personalities among their Web-enabled stuffed animals.

(Credit: Josh Lowensohn/CNET)

Spawn Labs

Spawn Labs was pitched as a "Slingbox for video games" and that's exactly what it does. The $199 box is a one-time purchase that users hook up to their home game console and their Internet connection. It then lets them play video games from any Internet-enabled computer as if they were playing it on their home TV.

The service has three big things going for it: One is that it pipes the content in 720p HD, which is the proper size for most laptops and what its creators tell us is as good as you can get for real-time streaming without bulging the price tag to around $5,000. It's also a one-time purchase, which means there's no monthly service fee beyond whatever you're paying for electricity and Internet. And, the company tells me it plans to offer compatibility with future game consoles through updated hardware drivers, meaning that you can buy the box now and not worry about having to upgrade it when the next generation of consoles arrives.

Considering it costs close to what most current-generation game consoles do, it may be a hard sell, but after having tried it out at the company's demo booth we definitely want one--and think many other gamers, and people in one-TV households will as well.

Spawn Labs' box can hook up to your game console and let you play your own games over the Web, free of charge. All you need to buy is the box.

(Credit: Josh Lowensohn/CNET)

SAN FRANCISCO--By late afternoon on Tuesday, it was getting awfully hot in the conference venue hosting TechCrunch50. Blame it on the body heat, or maybe the scores of laptops humming away.

But the air was sure to get a little hotter when it came time for the "Social Media Streams" category of start-ups to present.

The organizers of TechCrunch50 decided to save the last slot on the final day of the event (you know, right before everybody starts downing booze at the cocktail reception) to showcase new start-ups that deal with Silicon Valley's most hyped niche of the moment: real-time social media. As if Facebook and Twitter couldn't be dominating enough headlines here, there were six start-ups filling up the "stream" category: Threadsy, Lissn, Radiusly, Stribe, Clixtr, and The Whuffie Bank. And the panel of judges was joined by Twitter-savvy rapper Chamillionaire as a surprise guest.

Guess what? The judges, some of whom have been known to drink Silicon Valley hype Kool-Aid as though it were the world's finest wine, didn't think we needed most of these companies.

Oh, boy.

Threadsy's CEO Rob Goldman demos the site.

(Credit: CNET / Josh Lowensohn)

Threadsy, whose founders called it "the world's first integrated commnications client," was the best received of the bunch by far. It's a messaging client that aggregates e-mails, Facebook messages, Twitter replies, instant messages, and also "unbound" communications like general tweets and status messages that aren't necessarily geared to you. "We built Threadsy to pull you back together," CEO Rob Goldman told the audience, citing the rapidly growing percentage of Americans who are using more than one messaging client ona regular basis.

It's got a slick interface, can also aggregate automated profiles for your contacts' social-network feeds, and can track Twitter queries in an almost dizzying visual format.

"I think Robert Scoble's head was about to explode," conference organizer Jason Calacanis commented afterward, referring to the Valley mainstay's near-pathological obsession with social feed aggregation.

Scoble's response was remarkably pragmatic.

"I'm just wondering if it has the FriendFeed problem," he said, "which means there's not enough people in the world that care about aggregating all their friends' social networks," but added that he wanted to try it out as soon as possible. A few of the other judges raised questions about how Threadsy will make money, considering inboxes have never been a huge trove for ad dollars. Goldman's answer was a little bit convoluted, which this reporter took to mean that Threadsy hasn't quite figured it out yet.

Up next was Lissn, which appeared to be a combination of a news aggregator, a chat room, and a question-and-answer service. "Lissn starts with a conversation," founder Myke Armstrong said, and then demonstrated the app by posting the question "What would happen if the moon disappeared?" and watched comments and answers roll in. What wasn't really clear was exactly why anyone would use it, what with Twitter, Facebook statuses, and various "conversation" trackers out there already.

"Why would I leave Twitter to join this?" Scoble asked. Harsh words coming from the guy who loves to rave about the next shiny thing that streams words across your laptop screen.

Lissn lets people begin conversations about whatever they want.

(Credit: CNET / Josh Lowensohn)

Lissn was followed by Radiusly, which aims to solve scaling and communication problems for companies and brands that want to use microblogging and other social-media tools--many of which aren't terribly customizable. A company can build a Radiusly profile to create a directory of official social-network profiles for its employees, manage them internally, and share media like product images and videos for marketing and customer service purposes.

"I think you guys aimed at the right target but your dart hit the wall and not the target," Scoble said. LinkedIn founder Reid Hoffman chimed in, "In a rare position I agree with much of what Robert (Scoble) was saying." Ouch.

Next in the lineup was Stribe, which is in the same vein as Meebo's chat toolbar and Google Friend Connect--in other words, something that a smattering of established companies are already trying--adding social-networking features to any site by adding a chunk of code. Stribe can provide metrics pertaining to traffic and engagement, too.

Stribe's social network on a page (click to enlarge).

(Credit: CNET / Josh Lowensohn)

This was another well-designed one, but it was met with more skepticism. "I think one of the hardest things about these networks is actually getting the community to sign up," Facebook exec Mike Schroepfer said on the panel of judges. Dick Costolo gently reminded the Stribe team, "You can do too many things and then it becomes difficult for people to understand what they should use your product for...when you try to do a lot of things at once, it confuses people as to how they should use it and then they just don't use it."

The fifth company in the lineup received a somewhat better reaction. Called Clixtr, it's an iPhone app (and eventually expanding to more handsets) that combines photo-sharing with location awareness, turning the phone into what CEO Fergus Hurley called "the ultimate social camera." Clixtr's hook is event photos: The iPhone app lets you browse pictures from geo-tagged events, send photos instantly to other Clixtr users' phones, and find events near you.

"I think that was awesome," Schroepfer said, but expressed some confusion over exactly how geotagging could sync up to an event. Scoble complimented its sign-up process, but said "I'm not sure it causes enough gameplay, or enough something-else that gets me into this." He wasn't the only one to point out that getting people to use the app would be a challenge. "I would up the level of incentive for participation," Reid Hoffman said, and added that Facebook could easily build location-awareness into the photo feature of its mobile apps.

The last company was what Calacanis called "one of our wild-cards," The Whuffie Bank. Named after the deplorable term preferred by marketing-buzzword-loving social media consultants everywhere (basically, it's slang for social capital, a term coined by science fiction author Cory Doctorow), The Whuffie Bank is a non-profit organization for building a virtual currency around online reputation and influence. You can then use that currency to pay others with "whuffie," like tossing a bribe someone's way to ask them to retweet something you've posted on Twitter.

Note to the Whuffie Bankers: At the very least, please choose a different name for your organization. "Whuffie" sounds like something that would happen in porn movies. And the judges seemed to think that however cool of an idea it might be, it might be best if the currency stays in science fiction.

"The problem with these kinds of currencies is you generally need some kind of banking system to regulate them," Reid Hoffman said. "A lot of cool things...I think conceptually it's going to be extraordinary difficult."

"I want to hear in one line, what do I get?" celebrity judge Chamillionaire asked. "It seem like you've got to do a lot of work for them to raise your reputation...It seems like you can fake it."

And with that, it was happy hour. Or so everyone hoped.

SAN FRANCISCO--Two new companies are launching products designed to get the news to users faster--and from a wider variety of sources. Both are in private beta and not yet available to the general public but were demoed live at the TechCrunch50 conference.

Thoora is a new tool that clusters and aggregates news. It offers people a way to track the latest headlines with a simple ranking tool, ordering incoming stories by "Web reaction." It uses a mix of sources, including Twitter messages, blog posts, and breaking stories from more traditional news outlets. These stories are then filtered and pushed to a front page as well as Thoora's category pages.

One of the things that factors into what ends up on Thoora's front page is real-time chatter. The company tracks how many news-related tweets there have been about that topic in the last hour, as well as "Twitter impact," which is a percentage of density about that topic per 500 messages across all of Twitter over the past hour. It also tracks things like blog comments and linkbacks.

Thoora tracks hot news topics across a variety of chatter networks including blog comments, tweets, and news stories.

(Credit: Josh Lowensohn/CNET)

Insttant, on the other hand, cuts out traditional news sources entirely and uses Twitter's public stream instead. It takes these tweets and turns them into an interactive news page that covers people, places, and companies, including a way to track trending topics and user sentiment. All of this goes on a front page, which can be reordered and personalized with topics the user wants to see.

One of the service's more interesting tricks is that it automatically creates profile pages for people containing links and interests based on what they've shared in their tweets. This also happens for trending news topics, which makes for a more in-depth news-reading experience, since you can drill down on any topic and see things like recent mentions, related news and links, and a history of how popular it's been in the past few weeks.

Instant's front page is made up entirely of real-time chatter.

(Credit: Insttant)

Related:
Yahoo's Delicious adds a little Twitter
Full TechCrunch50 coverage

Perpetually is a new Web archiving tool demoed at the TechCrunch50 conference. It saves entire instances of Web pages, then lets users dial back to older versions. You just point it to a site or entire domain name then tell it what you want it to archive and for how long. It then does the hard work of saving pages to its servers.

Included is a visual browser that shows you visual history of pages in thumbnail form. You can simply flip through these before viewing the full-sized version. There's also a way to pick specific parts of a Web page and dial back such as a breaking stories box. And for those who want to find a specific archived page, or piece of archived content, there's a search engine that lets you go back without having to browse.

The service is not free; in fact, it's not even aimed at consumers. The lowest plan costs $99 a month, all the way up to $499 month, each with a higher level of monthly archiving storage. Considering each page takes up some storage space, it can fill up quickly, which is why the pro plans offer more.

The company said it's aiming Perpetually at media networks, historians, and PR companies. It also butts heads with Iterasi and its Positive Press product whose core technology was first demoed in January 2008.

(Credit: Perpetually)

I recently talked to execs from two companies that do exactly the same thing but in completely different ways. In one corner, Yammer, the 2008 TechCrunch50 darling. It's a Twitter-for-the-enterprise service that's hosted by Yammer. Any company can get its employees on to the service, but all the data is run through, and hosted by, Yammer itself.

In the other corner, Presently (found at Present.ly), another Twitter-for-the-enterprise product. Customers can use Presently in the cloud, just like Yammer, but the company makes its money from, and has most of its users on, its software that customers can install on their own networks, "inside the firewall," as they say.

Which is better? Dave Naffis, co-founder of Presently maker Intridea, told me that most of his customers are on the self-hosted version. "They try the SaaS (software as a service) version briefly then install the on-premises version. Over 70 percent of our customers are hosting internally," he says. Presently is installed at more than 50 companies, he claims.

Given what Naffis' customers are saying and paying for, it's understandable that he's going in the software direction. Just announced is a new self-install kit: Companies can sign up for free online and get a 30-day free trial license for the Linux version of the software (for Red Hat Enterprise Linux 5 or CentOS 5). There's a predictable, one-time license fee of $2,000 (for up to 1,000 users) with maintenance fees running 5 percent to 20 percent a year on top of that, depending on services needed.

Naffis does believe that, "Over time, customers will get more comfortable with data in the cloud." But, he says, "a lot of these large companies are not there yet. Maybe in three to five years."

So the Presently strategy is to sell companies an architecture they can get behind today and will presumably be there for them when they drop their objections to running a corporate communications system through a hosting service. But the company will have to wean its customers off the installed software and back to the hosted service, and there's no way that will be easy.

Yammer keeps getting better, but its critical advantage is its architectural philosophy.

(Credit: Screenshot by Rafe Needleman/CNET)

Meanwhile, Yammer, which has more funding, has dropped its downloadable software strategy entirely. CEO David Sacks did tell me that early in the young company's life he heard the reservations of corporate IT guys about running a service like this off-site. His company announced that it would create a downloadable version of Yammer.

But over time, Sacks says, he came to believe that all his clients were all going to get with the program on cloud-based computing eventually. And, he says, Yammer-as-a-service started selling. The service costs $3 to $5 per user per month, depending on features needed.

Yammer has stopped work on the hosted version of the product and is now aligned completely around the SaaS service. Sacks cites a few Fortune 1000 customers using the product, like AMD. The market for his type of communications service is not as big as he wants it to be yet, but, he says, "It's big enough to support us today."

And, of course, Sacks is happy to rattle off the advantages of focusing on the service-based product model: No tedious and expensive upgrade distributions; no need to support legacy users who don't upgrade; the freedom to spend 100 percent of product development on one version of the product (and features for it); and the speed with which updates can be rolled out to customers.

Yammer has the strategic advantage in the hosted-vs-software model due to its deeper pockets (it raised $5 million in venture funds), which allows it to bet on the future instead of selling products based on today's architecture.

I didn't really like Yammer when it launched a year ago. In fact, I liked Presently a lot more and awarded it a special Webware 100 award in 2009. I felt Yammer was too spare. Since then, Yammer has added features (such as integration with Twitter), and rolled out more client apps, including one that embeds Yammer into enterprise e-mail client Outlook. It's a much better service now than it was when it launched.

I don't have enough information to judge the products on purely technical merits. Corporate customers will want to talk to Yammer and Presently reps themselves to see which service they're more comfortable with when it comes to scalability, reliability, and security.

Regardless of the feature comparisons, and even regardless of numbers of current customers, I give the strategic nod to Yammer, because it's not building for today's market, but tomorrow's, and it has the resources to wait it out. The whole enterprise Twitter space may collapse (if, for example, Twitter releases a strong business-focused service on its own), but if it does survive, over the long term, it looks like Yammer is better positioned to slide into the market that will be.

Facebook's Aditya Agarwal shows off its new Prototypes feature at TechCrunch50.

(Credit: Josh Lowensohn/CNET)

SAN FRANCISCO--Facebook took the stage on Tuesday afternoon at the TechCrunch50 conference for a "Developer Garage" event, to highlight just how important its team of engineers is to the company--and to unveil a new feature to let users play around with what they're up to.

Facebook engineering lead Aditya Agarwal unveiled a new offering called "Prototypes," which makes internal projects on the site accessible as applications on its developer platform. "Some of them are going to be buggy," Agarwal said. "Some of them are going to be super polished."

Prototypes, which is sort of like Facebook's version of Google Labs, had accidentally been unveiled in a company Twitter post earlier on Tuesday afternoon. "It's difficult to predict just what Facebook engineers will come up with next," Agarwal said of Prototypes, which has since been elucidated in a post on the company blog.

Many of Facebook's hottest new features were created in late-night employee "hackathons," Facebook Vice President of Engineering Mike Schroepfer told the audience. Its new iPhone app was created by a single engineer (someone from Facebook told me that this employee was actually a summer intern, which makes it even more impressive), its "Facebook Lite" low-bandwidth-friendly site option was created by three engineers, and the brand new status tagging feature was built in a hackathon.

Some of the new prototypes, Agarwal explained, are photo tag searches, desktop notifications, and a way to filter news feed items to see which ones your friends have recently commented on.

Considering TechCrunch50 is an event devoted to new Web start-ups, Facebook also had a pitch for the entrepreneurs behind them: employee Justin Osofsky then came onstage to talk about Facebook Connect and why start-ups ought to implement . He cited the power of being able to share information on such a massive network, the advantages of not requiring a separate registration process, as well as the proven jumps in page views and traffic that some of the 15,000 sites currently using Facebook Connect have experienced.

At TechCrunch50, Facebook conveniently was able to make the dual announcement that it's cash flow positive and just hit 300 million active users. There are 6 billion minutes spent on the site every day, Schroepfer explained, 1 billion chat messages sent, and 80 billion photos stored on the site (20 billion individual photos, each stored in four different formats).

Within an hour of the site opening up the floodgates to vanity URLs this summer, 1 million had been reserved, Schroepfer explained. He reiterated that the company's engineers were what kept it all afloat.

"The problem with this is, we (were) basically asking 200 million people to show up at the Web site at about exactly the same time," Schroepfer said. "Most people would call this a denial of service attack. We called it a product launch."

Oops! Facebook posted this screenshot of a new feature to Twitter and then pulled it.

(Credit: Facebook)

It looks as if Facebook was a little premature in using its Twitter account to announce "Prototypes," a Google Labs-like operation that lets members beta-test new features for the social network and offer feedback.

The tweet offering a screenshot of Prototypes was swiftly deleted--but props to The Next Web for snagging it before it was pulled.

So what are the "prototypes" in question? Facebook seems to be experimenting with desktop notifications, content discovery, and upgrades to its Events invitation service. Next to each test feature is a star-based rating system through which, presumably, users can offer their feedback.

Facebook plans to take the stage at the TechCrunch50 conference in San Francisco early on Tuesday afternoon to make a developer-related announcement. It's pretty likely that this has something to do with it.