Android and ChromeOS will converge.
Like all ostensibly sensible ideas, this is probably a bad one. The whole point of Android and ChromeOS, right now, is that they're "playful experiments", tweaking Microsoft's tail, exploring new markets. There's no point Google having a big, boring engineering consolidation project to bring them together. That's just going to alienate internal developers ("but this is my baby!") and confuse third-party developers ("shall we wait for the convergence?").
Better to have a reasonably decent Android emulator running on the ChromeOS machine so that Chromebooks can run Android apps.
Update : read the comments on this blog-post. ("this blogpost" meaning this one, on Platform Wars). Lots of good discussion there.
Platform Blogs
-
-
From Curiosity to Exploration3 months ago
-
A Local AI Model Setup for Apple Macs4 months ago
-
A good enough map9 months ago
-
-
Progress on the Block Protocol1 year ago
-
-
The Long Night is Coming5 years ago
-
Still the best6 years ago
-
-
-
-
-
The betaworks 2015 book9 years ago
-
-
-
-
2012 Roundup: From GitHub to New Labor11 years ago
-
-
-
-
14 comments:
Seems like a good move to me. And note that there's no timeline involved.
Oops, I forgot to include:
http://webseitz.fluxent.com/wiki/z2009-11-24-BrinAndroidChromeFutureCombine
Here's link to Elliot Harmon at Tech Soup Ready or Not, Here I Chrome with a way to run Chrome on your computer. But the more interesting angle to me was what Chrome might mean in developing countries. Despite all the criticism of OLPC, the potential for education of digital distribution and content creation has clear advantages.
I, Cringley has a post up about Chrome, but if you haven't already do read the previous piece Pictures in Our Heads.
Your predictions about SAaS are coming fast.
Kaunda : I see the point about ChromeOS on old machines. Certainly will be useful there. Though I think old machines are actually of limited value when Moore's Law is driving down the cost of new machines so fast. (Especially mobiles as that Cringely piece points out.)
I'd guess that a 300 dollar netbook is more valuable than a three year old desktop or full-size laptop. It's more powerful, less weighty to ship and uses less energy.
I'm also guessing we could easily see the 100 dollar netbook within a couple of years. A friend was showing me a Bifferboard the other day (http://bifferos.bizhat.com/ ) and how easy it was to install Linux / Python on it. You could certainly make a netbook around something only a little bit more powerful than that.
BTW : my £280 Asus 1005 has been my only computer for the last 2 months, and I'm happy enough with it that I don't foresee myself buying a "full size" computer again. This is quite controversial at work, as my colleagues assume it must be hitting my productivity to use such a small machine. I don't think so [though I may have other productivity issues, of course :-/ ]
Bill : I'm going to stay contrarian on this one. Yes, there are advantages to merging ChromeOS and Android. I just don't think they're particularly compelling.
Whereas ...
a) Google's strength is its redundant architecture; and they can easily afford to support two operating systems.
b) Even a focussed company like Apple are keeping their options open on this.
c) The important requirements of ChromeOS are lightness, speed and browser standards compatibility. For the foreseeable future ChromeOS apps. just *are* web-apps - that must run equally well in ChromeOS / Chrome on Windows / Safari on Mac and Firefox on Linux.
OTOH, the requirements of Android (or any mobile operating system) are to give access to an ever increasing range of capabilities of the device itself : phone, bluetooth, GPS, accelerometers, compass, camera, multi-touch, stylus, etc. (Maybe moving towards Sixth Sense style video recognition and projectors etc.)
The requirements and functions of netbooks and "mobile devices" are diverging. Putting the same virtual machine and byte-code on both isn't going to change that (cf. Java) Though obviously it would be cool to be able to write Android apps. in Javascript. I'm all for that.
Your last comment really helped me to understand you point, which is pretty good because not geeky enough to really understand this stuff ;-) Old computers in the developing world are one thing, but I'm chomping at the bit for the $100 netbook--or less. And it isn't clear to me whether that will be more netbook-like or phone-like. The Bifferboard is wonderful. My sense is hackers in the developing world are the ones that will come up something that will really work for them.
Over on The Register they have a piece about this with the comment:
"But the ultimate irony is that after years of criticizing Microsoft for bundling its OS with its browser, Google has nearly made them one and the same. Yes, you can run third-party applications - but only web applications, online services of the sort offered most notably by, well, Google. And you certainly can't use a third party browser."
Personally I can't see why I'd ever buy a PC that just had Chrome when I could just as easily buy a normal PC that can have Chrome plus many other applications (e.g. OpenOffice, Skype, Spotify ... etc...etc..)
I'd only ever be tempted if Google effectively subsidise the hardware to the point where a ChromeOS notebook is costing < £50 or so and a normal notebook (with either MS or Linux) is still costing in the £200 - £300 range.
But to be honest, even at £10 I can't see why I'd want to buy such a disabled PC when fully fledged laptop PCs are so cheap.
To me it feels like very out of date "thin-client" thinking from Google, so I'm very curious to see what happens.
Oli : It's absolutely classic "thin client" thinking. The question is why you'd assume that that's out of date as opposed to an idea who's time hadn't yet come, when it was mooted, mid-90s.
Google must essentially be thinking that the reason why the 'thin-client' idea didn't take off in the 90's is because the web wasn't good enough then.
I just think they're wrong on this.
The problem has always been that the economic benefit of buying a thin-client as opposed to a thick-client is so marginal that it's only large organisations who might ever be tempted to buy a one trick pony computer.
In hardware terms thin-clients are cheapest to build as a normal PC with a smaller hard drive. You still need everything else!
So, you can imagine schools or hospitals or government agencies, or call centres deciding that they only want to run application X anyway, so let's save £100 per machine by going thin-client and we'll save a bunch of money from the 10,000 computers we need to buy.
But for consumers buying just one primary computer (who wants to carry around 2 or 3 laptops) it's worth paying the extra £100 for a hard drive to have a thick-client for the benefit of having a general purpose computer.
I mean, who knows what will be the killer application that you'll want to run in 6 months time? Why limit your options with a smaller hard drive when a large hard drive is so cheap?
This was true in the 90's and arguably as computers have become cheaper it's even more true now. The sophistication of the web was never the problem.
Again, the only real benefit of buying a ChromeOS over buying a regular PC with Chrome is going to be money. Google will only be able to make this financial incentive big enough if they seriously subsidise the hardware - which is a business model that Google has employed regularly.
OK Google also thinks ChromeOS will start up so quickly that it's worth not being able to install other software on the machine. Given that this 'quick start' capability is something that everyone is talking about (and solid state disks are a major part of the solution here AFAIK), I'm not convinced that ChromeOS is really going to have such an advantage over regular Linux and Windows 7 for this one feature to be worth it.
One final point (for me anyway) is that I still regularly use my laptop in situations where I don't have a network connection (tunnels on trains is the hardest case scenario). What's the point in a laptop that starts fast but then goes dead when you lose the connection. Google gears will save some apps (and shows even Google sees the need for holding data client side!) - but at this point most web apps don't work with gears.
I mean really Phil, if I gave you back £280 would you swap your Asus for a ChromeOS laptop as your only computer?
Oli: You're right, that I wouldn't swap my general purpose computer for a ChromeOS machine. And, so, to an extent, I agree with you. For me, having the flexibility of a general purpose machine makes sense.
But it may not make that much difference to many people. And interestingly, this isn't necessarily about power-users or geeks vs. non-geeks. I do a lot of coding on a remote server over SSH. With Bespin, a programmer could comfortably live in a thin-client + cloud environment.
What makes the thin-client more plausible today than 15 years ago is obviously the speed and ubiquity of network access, but also the fact that the browser is so much more powerful as a container for applications. For most apps. you might want to install, there's a good chance that a browser-based version is available.
Having said that, I agree it's only going to be price or extra-security that drives people from PC + Chrome to an actual Chromebook. But to shift corporates to that model would be a major coup.
For organizations, the biggest savings might be support (lost worker time, support staff expense, anti-virus software, backup systems, etc.).
For people who might not be inclined to buy a "real" computer (e.g. grandma), maybe a thin-client makes sense.
Although the incremental cost of having broadband (or a cell-data acct) at home might be the bigger barrier for this market.
Then again, if your mom wanted a computer, and was prepared to pay for the data service, you might nudge her toward a thin-client to save those "support costs" of your time...
A possible use-case for a cheap thin-client is as a client (super-remote) for a home entertainment system, esp as people want to view internet-based content (home-video from your kids, etc.) plus traditional TV on the big screen they already have...
http://webseitz.fluxent.com/wiki/z2009-11-17-GoogleHomeServerPrediction
@BillSeitz : strictly speaking, my mom runs her own linux servers, both at home and remotely, and does her own technical support. ;-)
Re: Google home-server, I wonder if the business model really fits them.
Surely home entertainment is run by consumer electronics companies who provide the hardware + software bundle. That's potentially Apple's territory (as Cringely is often pointing out) or for the console-makers to fight over. A lot of it is about branding, distribution, signing up "content" partners etc.
Google are still an ubergeek company, who thrive behind the scenes and claim neutrality about content. Android and ChromeOS are tentative steps towards a different model. But I see them as basically technical hacks - a "better browser", a more open phone. They've yet to prove that they're the sort of company to do home entertainment.
(Hmmm... though, actually, if I ran Google and I wanted to get into this, I'd probably push the YouTube brand for consumer entertainment, rather than Android or Chrome.)
Yeah, I think you're right that Google probably doesn't want to own or service in-home hardware. Though I *could* see them cutting deals with consumer-electronics folks to license some stuff to run in TVs, set-top boxes, etc. like NetFlix did.
http://webseitz.fluxent.com/wiki/z2009-06-24-NetflixFuture
And YouTube might be a great brand-name for that. :)
(And, technically, I could see that box just being a proxy-server to cloud-based services... and a tablet being a good super-remote.)
Heh, your Mom is way ahead of mine, who doesn't really like movies made since 1955... so, now that we've moved away from her area, and want to let her see her grandkids, a netbook that can run video-Skype is something I could *maybe* see dealing with...
Post a Comment