Please Don’t Take My Beta Chrome Away (2): Cloud Castles

This is the second in a series of posts flowing from my conversation with blogger Adam Rothstein about his experience testing the Chrome Notebook.

I�m curious about the impact of a hardware beta�it strikes me as unusual, maybe even unique, the very public way that Google has rolled out this notebook. I wonder what kind of impact that move is supposed to have, and whether it has played out the way Google hoped it would.

We should note, perhaps, that the Cr-48 is officially called a �pilot-program�, not a beta. Maybe to distinguish that it is meant to be a one-and-done operation. But because Google is so famous for their beta tests, it seems like an apt description. The one thing I consistently feel about Google is that they are much smarter than most pundits credit them. I don�t know why they went the route they did, but I�m sure they have a plan. Even when they roll out something that isn�t widely adopted, you can see they were really studying it. It�s the whole scientific method thing–design your experiments so you learn something whether the hypothesis is proved or disproved.

To me, what is even weirder than a hardware beta, is that it is just a test, and will never be anything more. Again, I return to the mystery of the �product that is not a product.� Hardware is nothing. It�s just money. It costs a certain amount of money to make a number of machines and hand them out. Whoever is crunching the numbers must be seeing something in the �plus� column that justifies the cost. It also costs money to develop and run free webmail. With Gmail, they are gaining ad placement. Search placement too, I suppose. What else? A hell of a brand. But what about the Cr-48? There�s no product to hype. This thing is a void of branding; as mentioned, there isn�t a mark or symbol on it. They get the standard ad and search placement via Chrome, but no more so than if I was using Chrome on my other computer (which I already do). Is my extra usage of Chrome via web apps tipping the scale to increase their profit above the cost of my individual Cr-48? I�m reading the New York Times through Chrome these days, and not on my iPhone. I see more ads when I search in a new Chrome tab then when I open up mobile Safari. Maybe the numbers of these small migrations work out so it�s profitable. Maybe not. How long until such a thing is profitable? How cheap must computers get, and how valuable must user-base be? How long until Bing and Google hire folks to stand at the subway entrance, handing out free netbooks so you will look at their ads on the way to work, just like free daily papers? How long until that is a reasonable thing to do, or at least until the profit someone earns off of it makes it seem like a reasonable thing to do?

Or it could be something else entirely. Maybe I�m a collaborator in Google�s new thought-war. A hamster in their lab. Perhaps it�s just a �free calendar� that I earned for being a loyal Google product adopter over five years. Or they could be secretly learning my typing pattern to test a new security product they�ll roll out next year. Some other more nefarious beta than that? Or maybe they�re just �accidentally� snooping on my neighbors� Wifi, like they did with the Streetview panopti-cars. Could be be all of these things. Or none.

But here�s another interesting note we should factor in to the concept of a hardware beta. Google has given out �official� instructions on how to jail-break your Cr-48 and install Ubuntu. It�s not exactly user-friendly, but there it is. What the hell are we to make of this? Do they imagine that this will enable developers to test the hardware and OS in some way that they wouldn�t if it wasn�t jailbroken? Did they want to encourage hackers to focus their efforts on testing the security in other areas of the manufacture? Or is this just a beta test of a potential real-world use: third-party dual boots? Will we see this option on consumer releases? What if part of having a Chrome OS device is picking two different OS: a web app OS, and a hard app OS? That would change the OS market much more than the browser market. Maybe giving out these instructions is a way of ensuring the hardware can be �recycled� when the beta version of Chrome OS is deprecated. Is it really �jailbreaking� if the manufacturer teaches you how? Man, that�s a lot of questions.

I�ve been calling it a hardware interface with the web, but is that really correct? It�s an interface with Google applications, right? The whole question of browser choice disappears with a successful transition to this model. Should Firefox develop a notebook?

No, it�s definitely an interface with the web. All the Google apps are there, but they always are when you�re in Gmail or any other product, just like the rest of the web is, no further than a bit of typing into the address bar or a bookmark click. Though in the past few years of using Google products, I�ve noticed they�ve upped the connectivity between them. They used to be overlapping in function, contradictory. Certain products would be discontinued (my poor Google Notebook, may it rest in peace). But the leaders have emerged, and there is some growing continuity. Again, the future is a question of what Google�s goals are. Maybe if they made their product more seamless, more �exclusively Google�, (more �Facebook,� if I may coin a slur) then some of their experiments would work better. If they had replaced Gmail with Wave, for example. But then maybe that would have sunk the whole boat. Facebook is only one major slip up away from losing their whole integrated product, whereas Gmail doesn�t seem like it is going anywhere.

I doubt Firefox would build an OS, because Ubuntu is their de facto OS. Maybe they would work with Canonical to try and get Ubuntu as a standard OS offering from computer OEMs again, to push a more brandable, native Firefox experience on light netbooks. Ubuntu�s ready for it, the question is are consumers ready to understand that they can�t download .exe�s? With Ubuntu�s new Software Center and a slew of new web apps, maybe. People sure do like big cartoony icons for their programs.

Apple and Microsoft already have a browser/OS combo. The key to preventing �defection� from the default would be integrating them even further so switching to a different browser is less attractive. It feels to me like Windows 7 does this just with the GUI. It looks unified, and so it is used as unified. Apple solved that problem on iOS by simply making defection impossible. But I think Apple is too fixated on hardware. They�ve been riding the click-wheel/multi-touch train for years. Safari sucks. Especially Mobile Safari. They�ve probably got more of a profit line on attracting the sorts of customers they are attracting via hardware and OS, and no motivation to really work on the software. I can�t believe that Apple actually advertises iPhoto as a feature. Talk about shitty native software that relies on GUI to prevent defections! I was in an Apple Store the other day, and I doubt most customers could tell me what a browser extension was. And Apple is making tons of money. There were like twenty-five employees on the floor! They probably will continue to make money on their hardware and OS alone. I wouldn�t be the first to suggest that they are the �new Windows�, at least product-wise, even if they�re technology is different.

So it might seem that no one is really in a position where developing a browser OS makes good sense�except for Google. They can only gain OS share, and judging from how well Chrome is doing, they will. (People say Chrome isn�t widely adopted; but how many new browsers are there on the market? From 0% to anything above 10% is �entering the market� in my book. That is a lot of people using Chrome, and a lot of people using Google as default search.) So maybe this is perfect for Google. Also, Jolicloud is very interesting. I think they have the right combination of a browser and OS. It appears they have replaced Ubuntu�s Gnome desktop GUI with Chrome, for all intents and purposes. Brilliant! With a product like that, they could really reel in a lot of people who are frustrated with iOS, aren�t sure about Android, but still want something slick and quick. I wouldn�t be surprised if market versions of Chrome OS are more like that. Android looks more and more like it itself was only a beta. An experiment to see how distributing an open-source OS works among lots of different OEMs. Note this: the frustration about lack of Android updates on all handsets? Guess how you update Chrome OS? The same way you update Chrome. You log out (close the program) and log back in again (start the program again). And it�s updated. I think Google learns its lessons really, really well.

But isn’t the Cr-48 curiously a throwback, in an interesting way? I want to call it a dumb terminal for the Web.

But a dumb terminal, in the era of mainframes, connected the user to everything s/he would want to do. I think it shows that trends in technology are a lot like evolution. Moles evolved eyes, and then became nearly blind. It�s not a nature-throwback. It�s evolution, baby! Everything grows in patterns, toward and away from niches, as they open and close. A little genetic drift here, a little natural selection there. I wrote an essay about evolution recently in which I compared natural selection, the time-honed survival of traits that serve to increase their expression in future generations, as R&D. Whereas genetic drift, the sudden swings that happen in traits due to factors like sudden environmental niches, random mutation, and population variations that amplify selection, are more like the sudden leaps of insight by a genius start-up. Or, like a small beta test that becomes a major product trend. Across the tech gene pool, neither is solely responsible for evolution. And neither could function without the other.

And as anyone who uses Chrome will tell you, there�s nothing dumb about it. It�s a really solid browser, that has gained traction for a reason.

Fair enough. But then dumb terminals weren’t really “dumb” either�they were sophisticated machines in their time, designed to serve a particular set of needs.

Here�s another point of comparison, that I think shows that features (or, traits) are more important our understanding of the progressive evolution of product hierarchies. We are suspicions of the cloud�and perhaps rightly so. Since I can�t upload �hard� files to the internet with the Cr-48 or read them locally, I have to use my seperate desktop computer to put them into the cloud. Music, photos, whatever: into old-school Chrome browser, and up to the cloud: Google Docs, Dropbox, Ubuntu One, whatever. Then it can be accessed by the Cr-48. But with the iPhone/Pad, that has locally accessible storage space, you still have to transfer files through a desktop computer running iTunes. iTunes becomes the syncable desktop software, not unlike Chrome. (A merge of Safari and iTunes in the future? Hmm…) So which is more versatile? The upload-sync of Chrome, or of iTunes? I think easily Chrome. It runs on all platforms, can upload any file type, anything that can be re-read through a web app. It can be synced across an unlimited number of computers, both desktop and �mobile hardware browsers�. Also the Cr-48 can be anywhere, because it�s all through the internet. Whereas, iOS devices have to be synced to only one desktop, or else they delete everything on their drive (!!!). They can only use certain file types…. They still have to be connected with a USB cable. (and Apple is famous for being the cutting edge of eliminating slow and un-needed interface ports!) So, while the �cloud� may be uncomfortably remote to some degree in our conception, it is incredibly more streamlined and flexible than hard-wired methods of data syncing.

About Mohit


  1. Matthew, I enjoyed your perspective on the place and potential future of Chrome OS. Thanks for the reminders of old-school terminals, and reminding me about the intimate connection between a mobile iOS device and its desktop host. The price of DRM harmony is actually rather steep for Apple users.

    I’ve really started to use the Google Docs apps much more with the Cr-48, and it must have just been a catalyst. As you observed, it has been possible all along for some time to use GDocs in browsers, but the level of integration among the apps has been creeping up for a couple of years now, hasn’t it? From my earlier scattered use of them, I’d also say that usability in GDocs has gone up quite a bit, from mouse-responsiveness (I do use a USB mouse for creating Presentation and Drawing docs) to keyboard shortcuts that I learned years ago in Excel but can now use in a GDoc spreadsheet—without thinking about it being a web app.

    Two use cases have really shone with the Cr-48: GDocs collaboration (with different users on different platforms, in different places, and sometimes at different times), and single-author creation—where I may have a thought and open a browser at work to tweak a presentation for class, or do the same at home, and polish it up on a coffee break with the Cr-48, and then give the presentation from the Cr-48 without so much as needing to hook up to WiFi, much less a charger—just give me the VGA cable, thank you!

    But that convenience when collaborating is multiplied. I may find my own niche with browsers everywhere I go, but my colleagues can do the same, however best fits their lifestyle. And our collaborative product moves along all that much more quickly, and with better result—because it’s being composed in the cloud and can capture good thoughts easily and quickly after the thoughts occur, which improves the fidelity of the thought capture.

    As you noted with the solidity of the Chrome browser, I’d say that the GDocs integrated suite has the feel of many hundreds of person-years of dogfooding down there in Mountain View (in connection with collaborators around the world). I’m happy to use apps on my G1 phone, but I’m not expecting it to see the far side of Froyo. I work with apps on Windows workstations throughout the day, but gradually I feel Windows as more cumbersome and less productive for the simpler tasks. I see it sliding toward the nice once occupied by the Unix workstations of the 1990’s: high-powered, high-cost, lower usability, but best performance value for specific design applications. Otherwise, my desktop of choice has become Ubuntu since 7.10.

    And when is a beta not really a beta, anyhow? I try to find fault with Chrome browser, but it’s a hard task. On Windows, I’ve got the Google Chrome beta, the yellow Chrome Canary Build beta, and a nightly trunk of Chromium blue. I’m awfully hard-pressed to care which one I click, because they all seem to work well for me
    Last weekend I finally got a glitch, when I built a new Atom-board system, installed Ubuntu 10.4 and then updated to what I thought was 10.10—but turned out to be 11.4 Narwhal. When I installed Chrome beta, I couldn’t get the x86_64 Flash plugin to work. But it took a beta OS to make that happen…

    Anyhow, have you noticed how verbose Cr-48 comment writers seem to be? I’d say it has something to do with the really slick keyboard. But part of it is just how fun it is to have a notebook that is just right-sized: not too small to type well on, not too large and heavy to carry around many places with, and not so power-hungry that one needs to scope out a seat near an outlet to get a bunch of work (or commenting) done!

    Thanks again for posting your perspectives!

  2. My wife just received her Cr-48. It was an amazing moment in our home because our Mac just died and took with it anything & everything not backed up to the web. I enjoyed your perspective on this program but am basically writing to say that I dug the Pearl Jam – DTE reference.

Leave a Reply

Your email address will not be published. Required fields are marked *