 And open protocols and open source and open data are a better long-term bet for the success of your company and for retaining consumers for longer periods of time. But the days of developing standalone software are fast coming to an end. So here's an example we have. This is a MathLine event. This is a mashup, or a piece of software that basically pulls data and logic from multiple sources. It comes from competing sources and does something innovative and different with them. This is not a system that runs on one machine on the internet. It relies on a lot of different services. We're going to see more and more of that as we go along. So what does a liberated product look like? As we see it, it's in addition to things such as APIs where engineers and developers can access data and synchronize data and pull information and logic over to other services. On a simple consumer level, we think that you should have a big, fat export button, or a new import button. A simple way for you is to just say, I'm revoking my trust in you, I'm taking my data with me, and I'm going to bring it elsewhere. And that does work in the assumption that there are second sources available. But any data you create or put into a product, you should be able to take with you. There's a number of reasons why products are liberated. The most obvious one is the business reason of people, the companies fear they're going to lose their users. They focus on getting them, and then they focus merely on keeping them, and not on innovating and working to retain them to creating a product that they want to use more. So, but that's on the business side. More on the technical side, there's a lot of interesting difficulties here. Lossy conversion from one format to the next can be a problem, especially when you're doing it in proprietary formats, which is, again, when we seek to use open formats for possible. Speed of implementation is important. If you want to get a product out the door, you really want to focus on getting the product out the door, and not necessarily on liberating and making it easy for users to move their data from one place to the next. But quite frankly, data size can also cause difficulties. Hard disks have been increasing these sizes at a tremendous rate. It's an order of magnitude every four years, which makes it more as a law, like a jail, quite frankly. But our network speeds have been going very slowly. So if I want to get four terabytes from Chicago to Brussels, I'm going to put it on hard drives and put it in my luggage and take it with me in the airplane. The internet's not going to serve me for that. But lastly, again, internet resources are a finite cost, and you need to focus and prioritize. So that's why many products look at these later. And again, consumers, even though look at this typically at first, when you start using a product, people don't look at it and say, you know, am I going to get my data out of this? But again, these are not insurmountable issues. So liberation is important for a couple of reasons, and I want to point out a couple criteria that I use whenever I switch to a new product or look at a new product that's hosted in the cloud. The first question is, am I going to be able to get my data out? The second question is, what's it going to cost me above and beyond what I'm already paying for the service? Is there an additional fee? The third thing is, how much of my time is it going to take to get my data out? If I put 500 of my photos in a photo service on the internet, and I have to click save, download, you know, put here for all 500 photos, it's going to take me 12 hours, and who's really going to do that? Again, the door's not locked, but it's darn well-willed and shut. So to sort of tie it all together, lock-in is a short-term solution, liberation is more of a long-term solution. You're going to expect to own their data in the cloud just in the way that they own their data in their computers. And the ability to revoke your trust in a company or service is very important if you want to build and keep that trust. An interesting thing about the internet is that cost of distribution is rapidly approaching zero. This makes it very easy for new companies to get in the game, and it makes it even easier for people to try new products. This is very, very disruptive. We think this is a good thing because it makes competition much more intense and much better for consumers. So the way you compete in this situation with distribution zero, cost of entry is very low, is to innovate and not to lock people in. So just to briefly touch on, we have a nice grid of examples of some of our products and some of the different ways that we allow users to take our data out. My team in specific has worked on iGoogle Blogger, Google Notebook, Google Finance, Maps, History, that sort of thing. But an interesting, I'll just use one quick example, is we liberated Notebook about nine months before the company decided to cancel it. And Notebook is this little browser plugin where you can put notes in, copy the little excerpts from web pages and save them and annotate them. Users were obviously, who used this were very disappointed that we were closing off this product because it's very useful to them. But we had a competitor, a small company that did a very similar thing and they released a press release and a blog post and a video showing you how to take your data out of Notebook and put it into their product. This was a huge win for our users, myself included, who had a number of Notebooks. It made it not only safer for me to use this, but I felt more comfortable and certainly less angry that I was going to lose my data. We're currently working on two other products, Google Sites. We're trying to get users to put more of their data into a website through our product called Sites, which is sort of like a super wiki of some sort. So we're providing a way for people to download the whole site as HTML using micro formats that you can just drop in any Apache web server behind there. And beyond that, Docs and Spreadsheets. Remember that example of photos, 500 pictures to click a button? That's currently the case with Docs and Spreadsheets. You want to download it, you open it, you go to File, you go to Export, you choose your format, you save it. We were working on a way to basically select multiple documents, if not all of them, and create a zip file or zip archive and let you download that all at once. So again, we think the liberation and the ability for users to take their Docs with them, their data with them is important, and lock-in is definitely not the long-term solution. Thank you very much.