Nine to Noon: 12 August 2009

August 12, 2009 – 12:25 pm

I went through two telephones in this Colorado house and neither of them could hold onto the call. Now that’s frustrating! Here’s what I was going to speak about:
I will talk about recent American software company acquisitions and what it tells us about the economy and the future direction of cloud computing. Then I’ll tell you how to teach your kids to program.

Links: FriendFeed acquired by Facebook (BusinessWeek), SpringSource acquired by VMWare (InfoWorld), Poison for Venture Capital (NY Times), the Scratch visual programming language.

FriendFeed and Facebook

Facebook’s the biggest social network: more than 250m active users, with >120M logging on every day. They get a billion photos each month. It’s huge. But a person might belong to plenty of other social networks: Twitter, MySpace, they might have their own blog or two, they put photos on Flickr, …. It’s very easy to have a decentralised identity these days: my Internet presence isn’t just what I post on, there’s much more.

Friendfeed was an aggregator–it gave you one page to track all the things your friends do on many different sites. It looked a bit like the Facebook event stream, and so one of the rumours that went around after the acquisition was “Facebook bought it because they do a better job of presenting information to the user”. To get to the bottom of that, you should know why companies buy other companies.

There are four reasons: technology, HR, customers, and optics. A technology acquisition is where the company being bought has something new and hot, and the acquiring company wants it. Google often buys companies for their technology–they just bought one that does better video encoding over the web, which can save them millions in bandwidth bills because of all the traffic that YouTube does. An HR acquisition is where the company being bought has good people but no good technology. Buying for customers means you want to sell your technology into a new market and the company you’re buying has connections in that market. Optics is a fancy way of saying “we sold it so nobody feels bad about the company closing”.

Friendfeed was an HR acquisition. We know this because (a) they said so, and (b) the team is being split up. If the technology were good, they’d keep the team together working on the technology. So don’t expect Facebook to start looking like Friendfeed any time soon.

This acquisition is notable because FriendFeed was started by a big-name crew of former Google employees. They made Google maps hackable, they built GMail, they …. Google didn’t want to see them leave, but they were determined to set out on their own. Now they’ve been bought by Facebook, the company that Google worries more about than they worry about Microsoft. It shows that brilliance isn’t enough to create great products and make lots of money.

SpringSource and VMWare

Imagine you’re a website developer. You need to test your website on three or four versions of Internet Explorer and Firefox, several versions of Windows and Mac. Do you have one computer for each possible combination? No, you make your computer pretend to be a computer running Windows 95, or Mac OS 10.1, and the most popular software to do this is VMWare. It lets you run one or more “virtual” computers in your one real computer.

This is hugely popular with “cloud computing”, where a company runs a big data center full of real computers. They rent you just enough computer to get your job done, and they handle all the hard bits of backup, keeping the site airconditioned, swapping out broken hardware, etc. They use VMWare and software like it so give you a virtual computer that’s “just enough” of the physical computer for your needs. That way they can rent the same physical computer to several customers, each getting just enough compute for their needs.

VMWare just bought SpringSource, who made tools that make it easier to write big programs in Java. Lots of people used SpringSource. This was definitely a technology acquisition, because VMWare wants to own as much of the software bits of cloud computing as possible. So they bought the leading Java tools company, they’ll make sure it works VERY well on their cloud virtualisation software, and now there’s ANOTHER reason to use VMWare when you shop for or set up a cloud … you get more pieces of the puzzle, and they all work well together.


So you might think that this means it’s all good and there’s lots of money floating around in Silicon Valley and the VCs are all happy as Larry. You’d be wrong. The piece in the New York Times that I’ve put in the links refutes this nicely.

If you’re a VC, you put money in when a company is small so that it can become big and you make lots of money when it sells. Venture-funded companies often go through several “rounds” named after letters: you get the A round from the first set of investors, then a few years later you com e back and say “we’ve grown nicely, please pony up to get us to the next stage” and those same investors and maybe new ones will kick in more money in your B round, and so it goes until you’re bought by another company or you IPO.

At each round there’s a “valuation” — the rather arbitrary amount of money that the company is worth. It’s arbitrary because often they don’t have any sales. You’re forced to come up with a number so you can decide how many shares the new investors get. There’s an art to it, some rough guides (number of users, even if they’re not paying; size of the market), but there’s one thing you don’t want to see happening … and it’s been happening.

Valuations have been going down.

A “down round” is where a company comes back for a B round and the market says “yeah, actually, last round you were smoking crack and thought it was a $10M company. Now we look at it and we think it’s only a $5M company”. The investor that kicked in $1M in the last round got 10% of the shares (1M invested, 10M company worth), The same investor coming into THIS round gets 20% of the shares (1M invested, company only worth 5M). AND if the company sells for its valuation, the early round investor has LOST money.

This is, as the article says, poison — VCs take risks, but nobody wants to put money in if the investments are losing money.

As if that’s not bad enough, nobody’s cashing out — there aren’t IPOs, there are few and far between big sales. This is the VC version of the liquidity problem hitting the global markets: you need to be able to sell the company you’ve invested in, so you can get money back to return profits to your own investors and then to start the next round of investments. If nobody buys your company, and you can’t IPO, you’re stuck with your money locked into a company where, to be honest, it’s doing you no good.

In short, still grim times in Silicon Valley.

Kids Programming

For the last two years I’ve been using a system called Scratch to teach 7 and 8 year old kids to program. It’s ridiculously easy and fun. It’s free, developed out of MIT and it really is programming, it’s not a sham.

Scratch is visual: you don’t type in programs, you assemble them from little blocks. You can’t write something the computer can’t run: if two blocks aren’t meant to go together, their shape is such that they can’t be forced to go together. Compare this to typing programs in, which kids are slow at and hate, where a single typo can make your program unusable and the error messages are rarely useful.

You see a stage on which there’s a cat. The cat’s a “sprite”, a character, and characters can have costumes. You assemble scripts that make the characters do things: move, talk, change colour, lay down tracks, react to the mouse or keyboard. You can record and play your own sounds, draw your own characters, set up lots of characters interacting, and even write games.

Kids love it. Boys on the whole want to write games, and there are a lot of sample games that come with Scratch so they can see how to do it. Girls on the whole more interested in getting a lot of characters interacting and telling stories.

I teach classes in the same way: show them two or three simple scripts that make the cat walk back and forth across the screen, or change direction to chase the mouse, and then make it say “got it” when it catches the mouse. Then they’ve seen decisions, how to repeat things, how to get positions and other values, and that’s the basis of many of the programs.

Try it. If you’ve got kids and have wondered “what’s the best way to get them doing something more with the computer than just looking at YouTube”, show them Scratch.

You must be logged in to post a comment.