After wowing the IO conference crowd with their Android prototype demonstration, the developers sat down for a Q&A session. The question on everyones’ lips was what exactly the demo hardware consisted of, and who made it; while the answer to the latter was under NDA, we were given some details of the former. Based on a 528MHz Qualcomm processor, the smartphone used a Synaptics capacitive touchscreen and the UMTS cellular standard; the demo itself was carried out using a 3.6Mbps HSDPA connection. Android’s memory requirements continue to be reasonable: the prototype had 128MB of RAM and 256MB of flash.
Q. Do you have a more specific release date for Android now?
A. [Android should launch in the] second half of this year. What you saw on stage looks pretty good, but we want to make sure it’s perfect and that people have a really good consumer experience with it.
Q. Will you have to buy the device with Android on it, or will you be able to load Android onto an existing device.
A. The software will be released as open-source when the first handsets are available … so people can do with it pretty much whatever they want. We don’t dictate how it’s used.
Q. Touch is heavily featured in the demo today, but Android is extensible to all sorts of different platforms. Were you just showing off the most flashy parts … how did you make those choices?
A. The platform is designed to be very generic, it has to work with touchscreen devices, D-pad devices and trackball devices. Some devices might not even have a screen. We chose to show a demonstration with a device that happened to have a trackball; we have other devices with a trackball. We could’ve shown that exact demo completely driven by the trackball. That’s pretty compelling as not a lot a lot of platforms have that flexibility.
Q. Could you tell us more about the hardware of the touchscreen handset?
A. The specific OEM who made built that hardware we’re under NDA with, but I can give you some details on components. It is a UMTS handset, the demo was on HSDPA 3.6Mbps. It was a Qualcomm MSM7201A [Ed: 528MHz] processor, it had a Synaptics capacitive touchscreen, 128MB of RAM and 256MB of flash.
When we announced the product and soon after we posted some videos, some of our partners demonstrated the platform at at MWC at Barcelona in March. This is the most recent stuff; literally the finishing touches were put on the device last night. The clock didn’t exist last night.
Q. What was shown here first?
A. I’m not totally sure if I’m right here, but notifications we hadn’t shown before, the desktop-like launcher of applications where you can put alias and icons on the desktop and get a spatial-representation of workflow. The compass and accelerometer based Street View was a new application. The unlock mechanism – letting the device learn the user’s gesture to unlock it for security and sleep purposes. And Pacman was shown for the first time too!
Q. The unlock gesture would only be available on the touchscreen?
A. You could select each of the circles using the trackball, but it might make more sense using the touchscreen.
Google IO Android QA Part 2[youtube]http://www.youtube.com/watch?v=wqq_HEx6yqg[/youtube]
Q. [Apple are planning on playing a central role in distributing iPhone apps] Do you see yourselves playing a similar role?
A. I don’t think there’ll be any one place for [distribution of Android software]. We’ll provide infrastructure to allow application developers to easily get distribution and post whatever business model they have in mind. Some developers just want to give stuff away for free, some developers want to allow you to try something before you buy it, maybe a demo version they give away for free, then try to upsell you to a different version. The infrastructure will be similar to how YouTube works, just a repository that things get hosted on and the technology that lets you put that on a handset.
Q. Geographically, are there regions that are more interested in Android or that you feel will help push it forward faster.
A. It’s been limited right now to the planet Earth! We have a lot of interest – it’s a global interest. We’ve done a lot of things to make sure that in the development of this effort with the 34 partners of the OHA that everyone was represented
, we weren’t drawing any maps and saying “this is going to go this way, this is going to go that way”. It’s really a collaborative effort, development is collaborative.
Q. Could you talk a little bit about why you’ve chosen to keep it in-house up until this point, and then once you do open it up, how much will be open and what specifically will be closed, and what those closed parts might mean for developers.
A. There are different types of open-source projects. There’s the type where a bunch of people get together from ground zero and say “we’re going to create an open-something”; it’s loosely coupled, it’s normally not a corporate entity sponsoring it, it’s a bunch of individuals. They grow it very organically and becomes something big, it takes a long time: that organic nature is rather unstructured. Some argue that that’s a benefit, that it’s a period of incubation, that’s a very interesting time. We think that there’s a window in the mobile industry where it’s very important that there’s a platform that’s open, that’s truly open. That’s why we gathered the 34 members of the OHA to together create this thing, a little more structured than the organic model, and then open source it when it’s reached critical mass, and that’s what we’d call 1.0. When the handset is capable of running the platform, and the consumer is happy with the experience, that’s what we call critical mass. So the other question is related to licencing – do you pick the licence where you insist to the community that they contribute back their intellectual property right. So it’s taking the position that, let’s create this collaboratively, but let’s not dictate business models or whether you have to give back your source code, so that’s why we chose the Apache licence. The apache licence is, essentially, you can use the software, you can use the copyright to create whatever you want, you can use pieces of it, you don’t have to contribute your modifications – we encourage you to, it would be great if you did, I think this whole open-source thing is going to catch on. There is open-source community contributions that we use in the stack, for example Linux which Android is built upon, but it’s pretty much everything you need to build a phone will be open-source. The only things that won’t be are some specific applications for Google services where the services themselves haven’t been opened up. And that’s just a roadmap thing, we’re on a mission to open up as many of our backend services as possible.
Q. Do you envisage the carriers permitting people to be free with their phone, to install software? Does that pose problems with management and reliability as we have with PCs these days?
A. Why wouldn’t the carrier want to allow any application to be put on a cellphone? So far they’ve not had a platform robust enough to give them a feeling of security, and I choose that word purposefully. The platforms today, a lot of them are 20 years old. There’s a lot of legacy, and security wasn’t considered then. Now, here we are in the internet generation, and things come to you – that didn’t happen – when you build a platform from a clean state you have a chance to think about those things, think about what’s going to give the industry, whether an OEM or a carrier, a safety net should malware be produced for these handsets. It’s a platform that will enable the carriers to do more innovative things with their services.
Q. Will they?
A. I hope they do!
Q. [What about MultiTouch?]
A. It’s typically a hardware function, and the handset that was demonstrated didn’t have a MultiTouch sensor.
Q. Does the software support it?
A. The software doesn’t need to support it, you don’t need to go out of your way to support any particular hardware sensor. When that hardware developer puts that [MultiTouch] sensor in their handset, I’m hoping the hardware developer will go out of their way to develop the driver.
Remember, you can see all of our Android demo videos here. Vincent is now off to hopefully spend some hands-on time with the prototype, so keep reading Android Community for more Google IO coverage!