Google I/O 2012 - Chrome/OS Fireside Chat

Uploaded by GoogleDevelopers on 10.07.2012

>>Jeff Chang: All right. Can everybody hear me? Cool. Well, good morning and welcome to
the third day of Google I/O. Thank you everybody for coming to this. And I'm excited to see
so many bright faces here. And this is the Chrome fireside chat. We try
to do this every year as an opportunity for you guys, the developer community, to engage
with us, the Chrome team. So today we have a few product managers and a few engineering
managers here today to take your questions. If you look over on that screen, there is
a URL there that you can go to which will bring you to that page that you see on the
left over there where you can submit questions and vote on questions. So that will automatically
refresh throughout the session, and I'll take questions from there and I'll alternate with
questions from the audience. And so there's two standing microphones, one's
over there and one's over there. I would recommend actually moving close to the edges so that
you have easy access to the microphones if you want to ask any questions.
So that's it. We have an hour. Let's start with some intros. Each of us will tell you
our name and what we work on, and we will try to say something about either a recently
launched product or feature that we're excited about or something that's coming up.
So I'll start. My name is Jeff Chang, and I'm a product manager on the Chrome browser.
And I'm currently responsible for end user features and I also help coordinate the releases
that we do every six weeks. And the recently launched feature which I'm
the most excited about, which hopefully you've heard about by now, is tab sync, the synchronizing
of tabs across the many devices where you use Chrome.
We had previously already had the ability to sync your bookmarks and settings and other
stuff like that, but tab sync in particular allows you to see all the tabs that you have
open on all of your other devices and open any of them at any time, which I think is
really useful, especially across mobile and desktop.
And the tab even includes the navigations history, so you're right back where you left.
But that's me. And I'll pass it off to Alex now.
>>Alex Komoroske: Cool. Hi. So my name is Alex Komoroske. I'm a product manager on Chrome
as well. I focus on the open Web platform, so all the technologies that Web developers
use to build cool apps. My favorite stuff -- and if you saw my talk
on Wednesday this won't surprise you -- is Web components. So Web components is a new
technology, a new standard that's being developed. So relatively early stages, but it promises
to allow people to totally change the way they architect their Web apps and be much
more productive. So I truly believe this is one of the technologies that four years from
now we'll look back and say how did we build stuff without this? I'm really excited about
the Web components in particular. >>Erik Kay: Hi, my name is Erik Kay. I'm an
engineer on the Chrome team. Actually, I've been with the Chrome team now for a really
long time, almost since it started, which is actually getting close to six years ago,
believe it or not. We were secret for almost two years at the beginning. And so even though
we've only been out for about four years now, it's been a long road at this point.
Anyway, I've done everything from the auto update system to Chrome extensions, and our
app system, and I also now help lead the Native Client effort and a few other things.
The things I'm most excited about -- I'm actually really excited about Web components too, Alex,
also. I think that's pretty awesome. But selfishly the new stuff we're doing with apps I think
will really be a big deal for people who want to use Web technologies to do more advanced
applications going forward. So I'm really excited to see what people are going to build
with that. >>Ian Ellison-Taylor: Hi. My name is Ian Ellison-Taylor.
I'm the director of product management for the Web platform.
And I'm just realizing I probably should have sat at that end since we already covered a
couple of my favorite features in Web components, and apps.
I think just to echo Erik's comment, I think that what is coming with the new generation
of apps is going to be a game changer. We kind of teased it just a little bit in the
keynote. We tried not to really oversell it because it's very early days, but there's
a ton of very cool stuff coming there. But just is to make Pavel's life worse and
not repeat, probably the other thing -- the other major area I'm really excited about
that's coming up in the next sort of 12 months is around all the tools work.
We've got -- I think Pavel showed some amazing stuff with our own dev tools in Chrome, but
if you also saw Adobe did a fantastic talk on some of the work that they're doing around
Brackets and Shadow. So -- and they're doing -- have done some great work with Edge and
designer focus tools for the Web platform. And then there are other lots of companies
as well, Cloud 9 to name but one. So I think with a big investment in tooling as well as
on top of the platform we can make it much easier for people like you to build some amazing
experiences. So I think that's really going to help. So I think that's pretty exciting.
>>Pavel Feldman: Thank you. I'm Pavel Feldman, software engineer working on the developer
tools. And the areas we are currently focusing on
are better authoring so that you can develop Chrome within the tools, its memory. As the
platform matures we see a lot more requests on the memory. As your apps are running for
days and weeks, you better make sure you have no memory leaks. So we have a new heap profiler,
we have new tools for native memory measurement, lots of stuff there.
The other area that is important for us is support for mobile. We are working on the
remote debugging for the mobile and as well as emulating mobile handsets on your desktop.
We do support the -- we've created the remote debugging protocol that we are happy to see
other big players using now around WebKit. The other area that is important for us is
the rendering smoothness, and you've seen a couple of sessions on that yesterday. That's
where we make sure your apps are buttery smooth. And we are happy to see more and more requests
there. And overall we are happy to see the way the platform matures and the way you have
more and more demand in the tooling. >>Felix Lin: I'm Felix Lin and I'm on the product
team for Chrome OS. I spend most of my time working with hardware partners to bring Chrome
devices to markets. I guess I'm most excited about the recent Samsung Chromebook and Chromebox
that were just released, and the close working relationship over the past couple of years
we've developed with Intel and Samsung and Qualcomm and all the other folks in the PC
ecosystem building devices. I'm a huge fan of Chrome, as you can imagine.
I love all the things these guys have built, but what we're really trying to do with Chrome
OS is fully integrate Chrome directly with the hardware and deliver the best possible
experience in a fully integrated package. And so the benefit of Chromebooks and Chromeboxes
and other devices is that, you know, when we can control the entire stack all the way
down to the hardware, we can deliver the best possible experience for Chrome and the Web
on any hardware that people have. And so it's not about replacing PC's or replacing
other devices, but for the people who are spending more and more time on the Web, what
we want to do is just deliver an awesome, magical experience for all the rich, immersive
experiences that you guys are building for the Web.
>>Jeff Chang: Great. All right. So now we're open for questions. Feel free to walk up to
any of the microphones and ask or otherwise we will trade off with the --
>>Erik Kay: We'll take precedent for the people who actually take some bravery and walk to
the mic. [Laughter].
>>Erik Kay: We'll just intersperse the ones from moderator here.
>>Jeff Chang: All right. >>> Hi. Good morning. I've got a question.
We're seeing yesterday that it's possible to open tabs on other devices, and sometimes
if you're like left at home or you have the devices -- the tabs open there and you open
on a mobile device, will it be possible to close the tabs that are home because I read
it on the mobile device, it's done, the task is done?
>>Erik Kay: Yeah, I want that, too. >>Jeff Chang: So the question was it's great
that you can open tabs that are on other devices from your other devices, but what about closing
them as well? That's a really interesting feature request.
We'll have to think about that. It might be possible if Chrome is running
on that device, but if that device has already been shut down, obviously you couldn't do
it immediately. >>Erik Kay: I think we could do it.
>>Jeff Chang: Maybe we could -- maybe we could figure something out there.
But yeah, I imagine that could be useful. But yeah, when we started designing that feature
we talked about tab sync, I think we realized pretty quickly on that you wouldn't want the
exact same set of tabs duplicated everywhere. So that's not the way it works. It's more
or less a dashboard almost. You can see all of the devices you have, what's open on each
one and you just pick and choose what you want. But we'll definitely keep that in mind.
>>> Great, thank you. >>> +1!
>>> Can I make a question on Chrome OS? >>Jeff Chang: Yes.
>>> What's the road map on Chrome OS and VPN? Sometimes to configure Chrome OS to VPN's
and getting to the site of a network it's been a little bit of a hassle. >>Felix Lin:
So Chrome OS today already supports VPN's and we're doing a ton of work through the
management console to enable I.T. administrators to centrally manage an entire fleet of devices
quickly and easily to support different authentication schemes, different access control mechanisms
and all that. >>Jeff Chang: Looks like we have another megaphone
question over here. >>> Hi. I'm wondering if anything is in the
works to improve drag and drop interoperability between Web apps. So A couple of years ago
we had the ability to do file drag and drop in HTML5 and it was fantastic and I would
love to see sort of the next evolution so I could, for instance, be in Gmail and just
drag an attachment out of a message over to another application that can do something
with that file and have that application be able to sort of interact with Gmail to pull
in the information it needs and do something, or vice versa, drag out of an application
on to a Gmail, et cetera. >>Alex Komoroske: So that's a great use case
there. And there isn't exactly precisely a standard
respect being developed to do precisely what you're asking for, but something that does
something sort of similar is Web Intents, which is a spec that we're driving that allows
applications to say I have an attach or whatever and allow the users to pick which application
they would like to sort of service that request with. So theoretically you could have a loose
coupling between Gmail and any other Web app that would want to view those images or what
you have. But that's a good point that users are familiar
with drag and drop and it makes sense to be able to drag and drop these things between
tabs, but that's not something that's currently on the roadmap.
>>Erik Kay: Yeah. I just want to emphasize that last point, is loose coupling.
So Web Intents is another one of these really exciting Web technologies that I think is
going to change the Web in a year. That you're going to look and say, oh, my God, how did
we interact without this? Because the idea is that now Gmail can interact with a whole
host of Web apps that it knows nothing about. It didn't have to be performed with the knowledge
that Oh, I want to save this attachment to Dropbox or G Drive or whatever else or that
I want to share this picture with Twitter or Google+ or whatever else. That it's up
to you in the apps that are provided. And just that fact that we can now have these
websites interoperate without having to hard code relationships between them I think is
going to be a game changer. >>Jeff Chang: Cool. Any other megaphone questions?
All right. Let's take one from the Google moderator page.
So offline it's pretty important for the Chrome OS story. What features are missing from the
Web platform that would help drive offline application uses?
>>Alex Komoroske: I'll take a stab at this one. There already are a suite of technologies
in the Web platform that allow you to do offline things. You've got things like Local Storage,
you have app cache, Index DB and others. They aren't necessarily a joy to use right now,
but they do exist as alternatives to build upon. The file API is a great -- another great
example. But I would argue actually that the way to
think about really driving offline application usage is the way you develop your applications.
So if you develop your application with an always on server -- always connected through
a server mindset, you're going to approach it in certain ways and architect your app
in certain ways. And then when you try to layer on offline it can be pretty challenging
because you don't have that server connection all the time. So actually orienting yourself
as a developer and saying I'm going to start at the beginning working in this world where
I don't know if I have internet connection, it sort of forces you actually to pivot your
thinking and to design your apps in a different way.
And that's a great thing about the new -- the next evolution of Chrome apps is actually
they encourage you as a developer to start from this world where you are sort of offline
by default, and connecting to the server is almost an exception. So that really orients
you to sort of the architecture app in the way that really works for offline.
>>Erik Kay: Right. So thanks for the shout-out to the Chrome apps, but the -- we really want
to emphasize that this is one of these key architectural things that we are building
in from the beginning for Chrome apps is to say that apps should be written to be offline
by default and we're doing a lot of things built in to them to make it possible to make
it easy for you to not get -- you know, get your app into a bad state.
One of the interesting things about offline is people sort of get hung up on, you know,
Oh, well, I've got connectivity everywhere. Why do we need offline?
And it's because offline isn't such a straightforward thing. It's not a question of am I on an airplane
with no connectivity? It's, Oh, well, sometimes I've got spotty connectivity on my cell network
or other times I'm at a trade show with 5,000 people all using the same Wi-Fi, you know,
connection or other times I'm at a hotel with a captive proxy or, you know -- there's -- offline
is -- has a wide spectrum actually, and making your apps that is resistant, resilient to
these network conditions takes a little bit of forethought. You have to build app, assume
that it's offline and basically you're operating on local data and syncing with the server.
And all these features that Alex was talking about with local storage, operating on that
local data, having your code running locally to begin with and then synchronizing the server
and as you get network connectivity is sort of the key to making offline work.
>>Jeff Chang: All right. The computer that was projecting the URL for the questions just
died and won't start up. So we're going to try to replace that, but hopefully we have
a pretty good queue of Google moderator questions there and audience questions, so let's keep
going. >>> Morning. So Chrome is a leader of the
HTML5 standard, no doubt. I was curious on the roadmap for the mobile Android version
of Chrome, when that would kind of catch up to the current platform. Specifically in my
case for the WebGL implementation, being that that's super hard in Tegra and Intel chipsets,
but I just wondered if it was on the roadmap and where that might fall.
>>Ian Ellison-Taylor: Is anyone from that team here?
[ Laughter ] >>Ian Ellison-Taylor: Oh, is there someone
in the back? Oh, yes. Oh, Mrs. Smith, good Lord!
Yes, Grace, come down and maybe answer this question, otherwise I'll make up something
like it will be there next week, maybe Tuesday. >>> Yeah. [Indiscernible] is also here.
So WebGL we are under development behind the flags. There's a performance issue that you
just mentioned and then there's a potential DOS attack. So we try to work with like OEM
device drivers to try to address these problems. >>Ian Ellison-Taylor: The high level strategy,
of course, is that all these features end up on mobile. There's implementation issues,
and also recognizing that the unique power management requirements make it a little bit
tricky, but generally when we do new features in the Web platform we're thinking about the
mobile scenarios. Almost -- I would say first and foremost generally. We worry a little
bit less about the desktop because that's an easier scenario. So pretty much everything.
I can't think of anything off the top of my head that we're not thinking about being in
mobile sooner or later. So hopefully we'll -- that lag between desktop
and mobile will get shorter and shorter. >>Alex Komoroske: So there's also -- there's
a little bit of this temporary lag, but the fact that Chrome for Android is relatively
new. So we're still going through and making sure we get all those pieces that we worked
on to make Chrome for Android work back into Chrome. So that process takes a little bit
of time, which is why it takes a little bit for us to catch up to where desktop Chrome
is right now, but that's a temporary problem. So there's a long-term problem of some of
these features that you have to make sure you design it in the mobile -- to work well
in mobile use case and other ones that it just takes awhile for us to catch up to the
desktop. >>Jeff Chang: Okay. Next question from the
moderator page is, OSX 10.8 has a power nap feature that keeps various bits of a user's
data synced while the machine is suspended, email, et cetera. Any plans for such a feature
in Chrome OS? Felix, do you want to try to take this one?
I guess, you can look at it on that screen, too.
Since that machine is broken anyway, maybe we can get that screen mirrored on that screen.
>>Felix Lin: So this is definitely a feature that we're looking at. I don't know exactly
when we'll have this feature in Chrome OS. Certainly as we work with Intel and other
hardware partners, these types of capabilities are much easier for us to implement because
we have full knowledge of the hardware that's running in these systems and we can actually
design them. The great thing about Chrome OS, though, is
that you could buy a machine today and if that machine has the capability at all to
make this possible, we'll roll out that feature and you will get it automatically.
So a lot of these features, when we spec the minimum requirements for Chrome OS boxes,
a lot of what we're doing is looking forward and making sure that we've got the kind of
head room that we need to be able to add new features, new capabilities, even though what
we release on any given day may not include this full set of capabilities.
>>Alex Komoroske: I should point out as well, this basic feature, this basic request, makes
sense in a lot of different scenarios. It makes sense, especially on mobile as well,
where you don't know, you don't want to constantly have your radio powered up. There's also been,
with these event pages. Generally we're moving towards a model where it's possible for Web
apps to sort of allow Chrome to wake them up, do a little bit of work, put them back
to sleep in response to events. That's really basically the core of what power nap is doing
as well. So we are looking into this on multiple fronts.
So while Jeff is figuring that out, I'll move on to the next one.
Okay. Will I ever be able to run Chrome OS on a tablet, maybe even a Nexus 7 tablet?
>>Felix Lin: That's a great question. [ Laughter ]
>>Felix Lin: Um, Chrome OS was really designed to run on a variety of form factors, so there's
nothing about Chrome OS that says that we can't run on a tablet or pick your device
form factor. The first device that we focused on, because of the complexity of use case,
is the calm shell where we just bring full capabilities of Chrome to bear. You know,
today, Google's OS for tablets is Android and we are able to bring all of the capabilities
of Chrome to Android for Jelly Bean and Ice Cream Sandwich. We're working on creating
a completely seamless, coherent experience across all devices, regardless of whether
the underlying OS happens to be Android or Chrome OS.
So from the user standpoint, whether you are buying content from the play store or whether
you are running an app, our goal is to make it such that you don't have to worry about
whether it's actually Android or Chrome OS running on that tablet; we want that to work.
>>Jeff Chang: Next. How will the two operating systems, Android and Chrome OS, eventually
converge? Yeah, we get asked -- we've been asked about this in the past, you know, one
thing to keep in mind is that Chrome OS has the goal of designing an end-to-end computing
experience that -- that embodies kind of the same core principles as Chrome does, the security,
and, you know, over time we've been thinking about what parts of the technologies between
those two makes sense to integrate, right? We are very happy to say that Chrome for Android
came out of beta this week, so Chrome for Android is now stable.
And maybe Felix you have some other comments on this or --
>>Erik Kay: Well, I can add a few things. I mean, one thing is that you'll see, I think
you are going to see more and more convergence over time. You know, we -- first off just
bringing Chrome, you know, to Android was sort of the first big step. But there are
many places where we -- where we can integrate more. First off, getting the same set of apps
running, using the same cloud infrastructure. But I think the opportunities doing some of
the -- I think the Chrome-to-phone stuff that we had before, we're looking to -- to have
-- to leverage a lot of the same infrastructure and to have more and more integration over
time. >>Jeff Chang: All right. Any other microphone
questions? Don't be shy. Any plans for a messaging platform inside
of Chrome or Chrome OS, similar to C2DM or iOS notifications, which I believe are coming
to the desktop? >>Erik Kay: So we actually do have some -- effectively
with the sync we have a notification system built into Chrome and into Chrome OS and it
turns out that behind the scenes a lot of the, you know, infrastructure is very similar
and, again, as I said, that we're looking to unify some of the basic infrastructure
that we share across Android and Chrome. I would expect to see our messaging systems
converge as well. >>Alex Komoroske: You can imagine the solution
wouldn't just have to be about messaging, per se. It could be about notifications, so
different apps could decide different things that were worthwhile to do. In fact, actually
in the meet the Web platform video series we showed off a mock of what that could look
like. >>Jeff Chang: Cool. All right, audience question.
>>> A question about Chrome apps. What's the plan to get Chrome apps for Chrome on Android,
and how is it going to work? Is it going to the Chrome Web store, it's going to a (indiscernible)
to play? >>Erik Kay: Okay. Let's see, how do we start
with this? First off, you know, it's a common question that we get about what's our plan
for getting, you know, Chrome apps to mobile. So you heard Sundar say at the keynote, I
said the other day at the apps talk we are intending to get Chrome apps to mobile, basically
everywhere that we have Chrome. Now how are we going to do that? The details
of that? The timing of that? You know, it's too early to say. I can say that we're looking
for your feedback. We want to know from you what you want. What are the most important
features that you want to see brought to mobile? And just one last little plug, we are hiring.
[ Laughter ] >>Erik Kay: So if you want to help make this
happen, you know -- >>> Cool.
>>Erik Kay: Exactly. >>Jeff Chang: Cool, another audience question.
>>> Also about Chrome apps, you launched them from within Chrome but with a new version
that will act a lot like they are native to the platform. Any thoughts about setting them
up so you would be able to launch them as if they were a native app?
>>Erik Kay: Absolutely. That's one of the key features of the next evolution of Chrome
apps here is that we are pulling them outside of the browser. It's not just that they have
standalone Windows; it's that the way that you launch them will be without even necessarily
having to have Chrome running. So basically you will just launch them from the native
desktop. >>> When will we have private key SSH?
>>Erik Kay: Right. So I know that the engineers working on that pretty -- on the SSH thing
for Chrome OS is a top feature request. So I know that it's going to come. I don't know
what the timeline is. >>Jeff Chang: All right. Looks like we've
got someone walking up to the microphone. Go ahead.
>>> Why was the change to Aura made? I know when the Chrome was initially announced, the
launch video mentioned that you didn't even have a desktop background. And the change
on -- to Aura in the newer version of Chrome OS did make it a lot more like a desktop operating
system. What was the decision, why did that happen?
>>Ian Ellison-Taylor: It was 99% user feedback. I mean -- I think when we launched Chrome
OS in the particular UI model, it was something of an experiment. It was something that no
one had ever done before. We said that we were going to update it based on user feedback.
And obviously I think Sundar made the point about performance improvements and feature
improvements, but people just gave us their feedback that they wanted it to be a little
bit more like a -- you know, a "traditional" OS and particularly I think that model, I
mean this is just for me personally -- I will let Felix talk a little bit. For me personally
the tab model I think worked really well on the smaller screen. But once you get to sort
of Chromebox style and you're on a large monitor, that's -- that single large window gets a
little weird, you know. It's like you get a huge crick in your neck going from one side
of the screen to the other. So being able to like open multiple windows, switch between
-- then multi-tasking, now I've got three or four things going on at once, I want a
quick way to switch between them. There's just a use case that seemed pretty important.
And that you can very easily configure it so it behaves like it used to if you don't
like that. Then you just run everything full screen, keep everything in a tab. So if you're
in that model, you don't really lose anything, but if you want something a little bit more
traditional with multiple windows, you can do that, too.
>>Felix Lin: I think the biggest benefit of having a windowing system is that people can
do more than one thing at one time, right? So a very, very common use case is writing
an email and checking your calendar, if you have to flip between tabs and you only have
one pane, it's a pretty miserable experience. Also, as -- as Ian was mentioning, you know
with the Chromebox we are now able to deliver an experience on dual 30-inch monitors, if
you have got a huge desktop having one full-screen window is not a great use of space.
>>Erik Kay: Also, one other kind of funny thing that we discovered by having a full
screen browser, which is there's sort of an interesting psychological effect for users.
So users have sort of preconceived notions of what a browser can and can't do, it turns
out. And so they see tabs, they see an address bar, they see a back button and they immediately
assume, oh, well it's, you know, it's only a Web page, it can't do everything an app
can do. Or it certainly can't run offline, so there's sort of these interesting things
that sort of get in the way just by having it there. And as we move to more focus on
apps within Chrome OS and wanting to bring apps fronts and center, it was important to
pull those out of the browser and really give them their own windows.
>>Ian Ellison-Taylor: That actually reminds me since I tease this on Twitter about there
being an Easter egg. It reminds me of one of them. So Kan, if you saw the keynote yesterday,
Kan did the sort of the spinning window thing, so that's in Chrome OS. So that wasn't just
something we rigged up for the keynote. You can -- it shows the power of the window
manager, so one of the keys is very obvious. >>> Control alt and then refresh button or
something like that? >>Ian Ellison-Taylor: Sorry, what was that?
>>> Isn't it like control alt and then the refresh button?
>>Ian Ellison-Taylor: It may or may not be that.
[ Laughter ] >>> I have done it before.
>>Ian Ellison-Taylor:. [ Laughter ]
You give it away. But there's another -- another Easter egg that was in the keynote as well.
I don't know if anyone spotted that one, but I'll maybe talk about that later, if the questions
are good enough. [ Laughter ]
>>Jeff Chang: Go ahead. >>> Yeah, so the current Chromebook runs on
Intel or Celeron, but all of your Android devices currently run on ARM, video, Qualcomm,
TI, respectively, if you're ever going to get to a point in the future where you are
going to be able to dual boot a device, whether it be a tablet or a laptop, desktop, or even
a smart phone down the road, into either Chrome OS or Android, you have to get on the same
CPU hardware platform, is there anything inherent that makes it difficult for Chrome OS to run
on say an ARM chip or is this just a matter of timing that this is basically something
that you should expect at some undescript point down the road?
>>Felix Lin: I think -- we can't talk about any future products obviously. But there's
nothing about Chrome OS that limits it to a particular processor architecture. Completely
open source. In fact, many of the ARM vendors have brought Chrome OS up on different ARM
SOCs, and it's really a matter of, you know, timing and price point and, you know, a variety
of other things with respect to, you know, packaging a product and bringing it to commercial
release. So I don't -- you know, basically there's nothing holding anybody back. Other
than, you know, commitment to ship the product. >>Erik Kay: There is one thing that I will
toss out. Which is it's not a limitation of Chrome OS. It's more just sort of what our
chip sets have been building -- what the ARM chip sets and manufacturers have been building,
so their main focus is on phones and now tablets, and what's interesting is that phones and
tablets have smaller screens. And it turns out that -- that the thing that you need that's
one of the biggest performance bottlenecks on computers in general is memory bandwidth
between the CPU and the display architecture. And the bigger the screen, the more memory
bandwidth that you need. Turns outs the smaller the screen, the less you need, right, you
know, conversely. So it's an optimization that you can do to make it use less power,
to be cheaper and that sort of a thing. So many of the ARM chip sets are designed to
really be only efficient up to a certain screen size. So it turns out, this is one of the
things that held ARM back getting to the desktop, bigger monitors, you know, hi-res. But now
as tablets have had -- now get an increased density, increased size, increased number
of pixels, they've had to solve those same memory bandwidth problems. So I think we are
right in the cusp of being able to see really awesome desktop and laptop ARM form factors
coming and you will see it probably for a variety of systems.
>>Jeff Chang: Thank you, next question. The moderator page. Speaking of sync, any plan
for app syncing as well, with an API that would allow apps to sync their data as well?
>>Erik Kay: So, yes. That already exists. There's an extension API people can start
playing with today, storage API,, basically it's a simple key value store so
you can sort of maintain your state and your preferences and that sort of a thing. It's
an easy way to sort of move app and extension settings from -- from machine to machine.
We have much more planned for doing more cloud-based storage and making APIs available to apps.
>>Ian Ellison-Taylor: And obviously there's a quick plug for our colleagues on drive that
for applications to be able to store their application, the user data, that's obviously
one of the goals for drive and getting drive onto lots of devices. So you get sort of both
-- both parts of the equation, your settings and your data that can travel with you. So
sync is a huge part of the overall strategy about what we think is really interesting
about having lots of devices, but a Chrome experience on one of them. So I think it's
pretty key. >>Jeff Chang: Okay. So regarding Chrome OS
and the enterprise, what plans do you have for enhancing the management and performance
of Cloud Print, particularly around management of over 1,000 printers at a given company?
I know that in Chrome, we don't usually talk about version numbers, but Chrome 20 stable
-- just went to the stable channel this past week.
Actually, I take that back. I'm mixing up my versions.
We are always managing multiple releases at the same time, so we have four different channels,
I'm actually thinking about 21, 21 I know has Cloud Printing more directly integrated
into the Chrome's print dialogue itself, so it's easier to find your printers more easily
when you are printing. And know that that dialogue has a search box that you can search
across all of the printers that are available to you. So I hope that's -- that the UI team
has considered the issue of -- of when you have lots of printers there, although I think
this question might be around more from the enterprise.
>>Ian Ellison-Taylor: Yeah. Actually, I wonder if any here -- if they have any from Cloud
Print here. They have a session today, so if there's no one from the cloud print team
right here, definitely go to that session and ask them if they have done a lot of work
there. So they can probably answer the question. >>Jeff Chang: Cloud Print is actually the
last session in this room today, I think. >>Erik Kay: Actually, just so you know, I
mean one of the big strengths that we've seen with -- with Chromebooks and Chrome OS has
been enterprise management. This ability to, you know, drop in, you know, a thousand new
devices and just hand them out. And you not have to image them and configure them on a
per user basis and deploy a whole bunch of stuff to them is one of the key strengths
we see of Chrome OS. And so we've been investing pretty heavily
in the enterprise management console to really make that better. So if there's any limitations
on that, we want to hear about it. File bugs, you know, talk to us. Because we think that
this is an area that we can really own. >>Jeff Chang: Yeah, we have an audience question.
>>> So I came in a little bit late, I don't know if you addressed this yet, but I was
curious about what's the state of eating your own dogfood within Google? If you could talk
about the transition of getting people to be full-time Chromebook users instead of regular
desktop users. Are people doing that? How successful is it? What kind of experience
have you had? >>Jeff Chang: Sure, yeah. The question is
asking, you know, about the phrase eating our own dogfood, which means using our own
products as they are being developed at Google. I can say for sure for Chrome browser, I definitely
eat my own dogfood. I run Canary, which is the nightly auto update, built side-by-side
with the beta installation on my machine. So I actually use Canary every day, so I can
stay with the latest engineering. But I don't recommend that to normal folks. It's painful
at times. And in terms of Chromebooks, I certainly carry
around my Chromebook with me. I have it right over there and I use it and it's really convenient
when I'm going to meetings and stuff, especially when I want to get out an email in like 60
seconds or something. Open it up on my lid, press the button and be able to fire off the
email and close my lid within 30 seconds while, you know, like my other Windows laptop might
be churning away on the anti-virus thing for like a minute because Google corporate makes
us run anti-virus on them. So, yeah, we definitely see that a lot around the office. But I do,
you know, use multiple devices and multiple operating systems. And for Canary on Windows,
I run that, because I want to get the latest code from our engineers as soon as possible.
>>> Do you have anyone going cold turkey, just Chrome OS exclusively?
>>Felix Lin: I can tell you that the dogfooders at Google who are using Chrome OS have skyrocketed
since we released the latest Chromebooks running on Intel Corp. You know the great thing about
Google -- two things, right? We have access to whatever computer we want. If you want
a Macbook Air or Macbook Pro, whatever high-end system you want, you can have it. If you want
a Chromebook, you can have it. We don't force anybody to dogfood Chrome OS, so everybody
who is using Chrome OS uses it by choice, and when you think about it, the fact that
you see people, more and more Googlers who have their pick of any hardware that they
want walking around with Chromebooks, typically which cost often a fraction of what they could
be spending, it's pretty remarkable. I can tell you that with the latest Chromeboxes
and the latest Chromebooks, these devices are just completely hassle free, battery life,
power management, superb, you know, even the -- the first generation Chromebook, just because
of the fact that we have so many existing customers on older hardware and the fact that
we're constantly updating even older hardware with newer releases of Chrome OS, you know,
I switch devices continuously, and so yesterday during a keynote when you saw Brian talk about
the fact that, you know, he's got 7 or 8 different machines and he just walks from one machine
to the next, he doesn't carry a single machine he has to worry about, you know, that is all
possible because of Chrome and Chrome OS. You know, when you're using a machine which
is a year old and you pick it up one day, and it's running 30% faster than the day before,
because you've got the auto update pushed to you, it's really noticeable, and users
really love it, and so even today, even though I've got, you know, you know, the latest Samsung
Chromebook, when I switch to the device that I've been using for over a year, I really
don't notice much of a performance difference for most of the things that I do. I'll notice
some things for, you know, high-end video and things like that, but the device is completely
usable, it's just great. >>Erik Kay: So let me give you another perspective.
I do use my Chromebook exclusively as my laptop, I have to use a different laptop, because
in engineering we to test on a bunch of different platforms, but I do live on my Chromebook,
and it's pretty common now within Google especially with the hardware, I'll say that the IT staff
within Google, you know, loves it and they want more people to switch over, as soon as
they can, and Google -- Googlers are a really demanding audience. It turns out that, as
Felix said, they can use whatever hardware they want, and -- and a lot of them are doing
heavy-duty development tasks and that sort of thing, and so making hardware that is going
to be good enough for the majority of Googlers to actually get their job done is a tall task,
and the biggest bit of feedback that we heard early on was that for Googlers the previous
generation of hardware wasn't quite powerful enough. And now the feedback that we've gotten
from people, is like, wow, this is actually -- this has crossed the line and it's great,
it's easy, and totally what they want. Now, they want more of course and so we're working
to make -- build even more powerful machines and a wider variety of form factors, and trust
us when we say that Googlers are some of the most demanding audience out there, and -- but
the rate that they're growing and they're selecting in and using it themselves is going
through the roof. >>> It is a great idea to have kind of your
computer as like an appliance or kiosk, and I feel very nervous if I have anything on
a single machine, that is my only copy, I like moving everything to the cloud, just
being able to -- but I know that there's lots of little, you know, dusty corners of our
lives where they're not all, you know, able to be pushed to the cloud on Chrome OS yet,
so it's nice that you guys are kind of filling in those holes.
>>Ian Ellison-Taylor: That's a great point. I'm the same way. Since I don't do development,
I'm 99% -- 99.9% of my time is spent on a Chromebook, and mostly that -- honestly it's
about I'm old and I need a keyboard, I love tablets, but I spend all my time in e-mail,
I just need to be able to type on a real keyboard, but I love the battery life, that sort of
8, 9 hours, I can get through the day and I don't have to carry the little power brick
around, so the battery life thing is huge for me when I'm running around campus. Another
thing that's really nice about Google, and I bounce between different offices, I'll often
not carry a machine with me, and I'll go to the other office, we have these little tech
stop places and I can grab a Chromebook from tech stop, I just log in, carry on from where
I left off, and then I just hand it back at the end of the day, and I just don't have
to travel with anything, so when I'm going through security, and, you know, just everything
gets so much easier when I can just pick up that machine, log in, carry on, hand it back
and not worry, like what did I leave on that machine?
What was I doing? Did I do anything weird today?
So for me, that's a hugely important use case, and I think we'll see more of that as time
goes on. >>> Thanks.
>>Jeff Chang: Let's take a question from this side.
>>> My question is on moderator but it seems to have dropped off. I just want to know about
international availability of Chrome OS devices particularly the new Samsung Chromebook, because
I'm from New Zealand and, yeah, we just don't have access to them there.
>>Ian Ellison-Taylor: What's the international access for Chromebooks, particularly the new
one. Particularly New Zealand? >>Felix Lin: So the new Chromebooks are currently
available in US and UK. We're looking at getting them certified in more countries as quickly
as possible, but, you know, as with all things, it's a matter of just figuring out where the
demand is and how to role out in more countries and get the certifications. You know, the
demand for the new Chromebooks, i think, have greatly exceeded the expectations and so,
again, it's just a matter of time, but we're working on getting it out there.
>>> Okay, thanks. >>Jeff Chang: All right. Back to this -- actually,
I'll go ahead and let's do a couple of these. At least one of these. How far is Web GL support
for mobile versions of Chrome? I think we already answered that earlier up at the microphone.
Next question here is about Chrome on iOS. So for you guys in here, we launched Chrome
on iOS yesterday, we actually it's like the number one ranked in the app store now, right?
>>Ian Ellison-Taylor: Yeah, Chrome on iOS was number one, and drive on iOS was number
two. We'll see how long that lasts. >>Jeff Chang: So we're happy that --
[ Applause ] >>Ian Ellison-Taylor: Thank you.
>>Jeff Chang: We are definitely very happy that we were able to get that out to you guys,
especially with the think store we've been telling, it's very important that we have
Chrome available on as many platforms as possible. In terms of how Chrome runs on iOS, you know,
we obviously have to follow Apple's technical specifications there, so in terms of the actual
Web content it does use the UR WebView. I don't know if anyone else has any comments
there. >>Erik Kay: So there's two things: You have
to think about performance on a number of different axis, right? So one is how long
does it take -- how fast is it to just get around and get your work done and do stuff
like that and then the second thing is there's raw compute performance or rendering performance
within the page. I guess you could count those as three areas. So on the compute -- on the
pure compute side of things, like if you were to download just a JavaScript benchmark, we're
limited by what Apple provides us in that they don't allow you to run your own jitted,
you know, JavaScript engine. They don't allow you to provide your own rendering -- you know,
Web rendering engine, it's just not allowed by the terms of service, and so we have to
use UI WebView, and the UI WebView is actually more restrictive than what's in Safari. Even
though it's some of the own code, they actually turn off some features. And one of the things
they turn off is the use of the JIT in their JavaScript engine, so that slows down heavy-duty
JavaScript things. The second thing it does is it affects some of the GPU acceleration,
how that interoperates. So we have some limitations there, so if you wanted to do just sort of
the raw how fast is this particular page spinning in a loop doing stuff, we aren't going to
be as fast as we are on other platforms just due to limitations; however, the good news
is is that that's not the entire story. The usability, you know, in terms of how long
does it take you to get from point A to point B is how much we predict what you're going
to type, that helps with sync, your usage patterns, the Omni box, how we're able to
pre-fetch -- guess what you're able to do using our pre-rendering and some of our DNS
pre-fetching stuff. We have a lot of tricks within Chrome beyond just the JavaScript engine
and what we're doing inside the Web rendering engine. So actually if you read some of the
reviews that are floating around out there, if you see the ones that load up a benchmark
and compare them AB, you'll see that we're slower. If you see the ones that try to use
it and say I navigated to the following three sites and time that, you'll see that we're
actually way faster, that some people are getting things done twice as fast. Getting
to their web pages, so if what you're doing is surfing and navigating around, actually
I think you'll find Chrome on iOS is really fast.
>>Jeff Chang: Yeah, for me just the fact that all my favorite websites all ready auto complete.
It just helps a ton, and for me I use a lot of tabs. I tend to open a lot of tabs in the
background when I'm reading stuff, like read it or something. Being able to flip through
really easy, switch through them is a huge usability improvement for me. All right. Go
ahead. >>> On a similar note, one of the things that
makes Chrome so much more fast than anything else is the extensions, and in particular
for me, the Omni box extension capability of making completion work. When are you considering
bringing this to mobile? >>Erik Kay: So I think we talked a little
bit about this before, but basically, you know, Chrome is a pretty big set of features
and the Web has a very broad set of features and so we only just got, you know, Android
to -- Chrome to Android a few months ago, and we only just got Chrome to iOS today,
or yesterday. So we've got a little ways to go to get to feature parity, and extension
is one of those parts that we actually like to take our time with because it has, you
know, deep integration to all of the various hooks in the browser. We also think that in
mobile it's quite likely that how you want to extend the browser is going to be different.
Also the performance needs of mobile are different. So -- so some of the changes we've been making
in the desktop has have been anticipating our needs in mobile. If you look at the new
stuff we've done with event pages in extensions where basically there's no longer going to
be an always-running persistent background page for every single extension, that way
we can be a little more on demand, this is partly to reduce the resource function of
extensions, and we think will make it more adaptable to mobile down the road. Unfortunately,
you know, I don't have a good timetable to tell you, but it is something we're interested
in. >>Jeff Chang: All right, switch over to this
side again. >>> I'm curious if there's going to be any
advancements in printer support without Google Cloud Print. Right now it seems like you need
a separate machine in order to be able to print. Is there any way that it's going to
be possible with a Chromebook alone. >>Jeff Chang: Sounds like you're asking about
printing outside of Google cloud but specifically for Chrome OS.
>>> Yes. >>Felix Lin: Today people can use any of their
legacy printers by using Cloud Print through an existing machine on the network. On my
home machine I have windows XP set up, I've got Cloud Print installed and I can get to
any of my legacy printers on the network through that with pretty much every printer I'm aware
of on the market today. They're already Cloud Print enabled, so by simply going any new
printer that you get today pretty much can just be on the network and accessible to any
Chrome OS or any other device that uses Cloud Print.
>>> Oh, okay. >>Jeff Chang: All right. Back to this side.
>>> I found out I think next year that Google is giving my daughter's school a bunch of
Chrome OS netbooks and we live right next to Mountain View, so I was just curious like
what's the thinking behind that, and is it more than just a local thing?
Is it like going to be in other places around the country?
>>Jeff Chang: So I know for sure that we've had a lot of -- a lot of schools around the
country starting to adopt Chromebooks. Felix can probably tell you the exact numbers, but
I know that the feedback has been very positive. >>Felix Lin: Education is one of the most
-- most aggressively verticals adopting Chromebooks. And the beauty of the Web and education is
that, you know, the Web is really a great equalizer, right?
All the best content, apps, information, accessible to students, very, very low cost, always up-to-date.
Chromebooks are phenomenal platform for schools because schools have, you know, shrinking
budgets, no IT staff, and being able to take a device that has zero maintenance, zero management,
always up to date, doesn't require a ton of care and feeding, and gives people access
to all of the tools that they use is really just a great formula for education. We're
-- in addition to, you know, the hundreds and hundreds of schools I think in 48 states,
there are schools that are deploying Chromebooks, we're seeing huge initiatives at national
levels where in many emerging markets countries are looking to Chromebooks as a way to, you
know, transform education and really transform great economic transformation across the country.
And, you know, it's really all about, you know, the rich content that's available, all
the educational material that is available, and just the ability to get a situation where
instead of, you know, a 20 students to one computer type ratio in a school, giving students
a one to one opportunity where every student has a computer that they're using to enhance
their learning, is truly transformational. And so, you know, from our standpoint, and,
you know, this is the great thing about working with many of our hardware partners, in education
we're finding a great willingness on the part of companies to be able to foster education.
People don't seem to have as large a profit motive in education as they do in other verticals
where they need to be making, you know, an enterprise profit, but this is just a huge
opportunity for us and for -- for future students. >>Jeff Chang: All right, back to this side.
>>> Okay. Congratulations, by the way, on shipping Chrome for iOS. It's a good thing.
I was in a session yesterday afternoon, optimizing your HTML for -- for a mobile device, and
the presenter invited us to follow along, give URL for slides, and said we're going
to working on this interactively, so I pulled out my iPad and downloaded Chrome for the
iPad, and was following right along, and then he said, okay, now we're going to install
the page speed extension, so we can watch and see what happens, and I said, great, search
for page speed, right here, Chrome Web store, and it said umm, the Chrome Web store is not
supported on your OS. >>Erik Kay: I tried to touch on this earlier,
but extensions are something we definitely want to support down the road on mobile and
other platforms. It's just very early days and getting into mobile is going to take us
a fair amount of work. >>> Okay.
>>Jeff Chang: Go ahead. >>> So another Chrome on iOS question. I'm
just wondering if it does or is going to support URL schemes so that iOS developers launch
Chrome if they choose to. >>Erik Kay: Do we have anybody from the iOS
team around here that can answer that? Tell you what, why don't you -- if you want
we can look that up for you and leave your e-mail for us.
>>Jeff Chang: We can put you in touch with the right people.
>>> Okay, thanks. >>> Hi.
>>Jeff Chang: Hi. >>> I have a question about my Chromebook.
I would like to use it as a Web development platform with my HTML JavaScript files, and
I tried that yesterday and then I ran into two problems. One is that I don't have a text
editor for my HTML files, and I saw one of the presentations that was using an app, I
want to know which app that is. And the second problem is how to run my HTML files from Google
Drive. I only get a preview of the source code. I don't get a run.
>>Erik Kay: So I can talk to that. So that you're probably talking about our presentation
where we had some text editors running on Chrome OS, and this is using some of the new
apps features that we've been working on. They're pretty early on, so they're right
now available to people developing in Canary, but we hope to have them more broadly available
later this year. If you're a developer, want to ride the ragged edge, you know, you can
come check out -- we've got some sample code. I think that particular little micro editor
is checked into GitHub, and our Google Chrome repo, but a warning is that it's early, it's
very simple. It actually isn't as cool as it looked. We -- we -- we actually hit over
all the rough spots. But there are -- I'll tell you that just one of the cool, fun things
about the new apps stuff we've been doing, is just once you add these new capabilities
of seeing what people want to build with them, and within Chrome team, one of the first things
people wanted to build actually was code editors, and in fact we have like three of them already
that people are just doing different things with, and so we actually expect there to be
a plethora of code editors. >>Jeff Chang: I don't know if you mentioned
it already or not, there's definitely IDEs available in the Chrome Web store. A bunch
of them are pretty highly ranked. >>Jeff Chang: The number of cloud ID's emerging
and you can use them, and as the Chrome, we are providing them with the extension APIs
so they could plug into the instrumentation and get the dev tools integrated into the
cloud tools, so you have both complete ID with the remote persistence and the debugging
capabilities on the Chrome OS. But it's all coming. We're working on that. We're working
with base teams on these capabilities. >>> And then how do I run my HTML file? When
I clicked on the file from Google Drive, I just got a preview.
>>Erik Kay: So you have to ask the Drive team, so unfortunately none of us are from there,
so I don't know if they have any plans to allow direct hosting of HTML straight from
Drive. I assume that's what you mean for public consumption, but if you just want to do local
development, it should be possible to do -- to read the files locally and render them, you
know, in Chrome. >>Ian Ellison-Taylor: (inaudible)
>>Erik Kay: Yeah, we can check with the drive team.
>>Jeff Chang: Yeah, we can check with the Drive team. It looks like we pretty much ran
out of time. >>Ian Ellison-Taylor: Hang on. I forgot one
other Easter Egg, you probably already know, there's another Easter Egg in the video, the
History of Chrome, so you might want to study that video and -- and there's something pretty
cool in there so -- >>Jeff Chang: All right. Perfect timing. Yeah.
So thank you for coming. We enjoy doing this every year as long as you guys come, and so
hopefully see you next year, otherwise if you just stay in this room, there's a bunch
of Chrome sessions after us, looks like high performance HTML5, history and future of Google
Web tool kit, writing secure Web Apps, and Chrome extensions and the Google Cloud Print
section, so please enjoy the rest of conference, and thank you.
[ Applause ]