I'm not even going to try to link to the other articles already out there on Google's new notebook computer and the odd little OS that powers it. A quick search of the interwebs (via Bing, if you must) will turn up dozens. I'll only point you two places to start off: how to apply for the Chrome OS pilot program, and the under-the-hood specs of this black box (literally) notebook.
Let's get something out of the way right off the bat. I am exactly Google's target market for Chrome OS. I use Google Apps for my own domain (for example, I get to my documents by typing docs.kirv.in into any web browser, my mail by typing mail.kirv.in, etc.). I do just about everything in a web browser already, and that browser happens to be Google Chrome. I already sync my bookmarks, passwords, etc. between my Chrome installations on various PCs. I access Twitter via Seesmic's web client or twitter.com, same for Facebook. With very few exceptions, I live entirely in the cloud already. And even on my PCs, I adopted the Chrome OS lifestyle before there was such a thing. I run all my applications maximized, even on my 22" monitor. I never overlap windows because not being able to see part of a window offends my OCD.
I say this because a lot of the negatives I've seen about the Cr-48 and Chrome OS are actually positives to me. I'm glad there's no desktop. I'm glad there's no file management. On my Windows gaming rig, I have Chrome — obviously — Kindle, Calibre, iTunes, Teamspeak and my various MMOs installed. That's it. Well, that's all I use. That's only about half the software I have installed, because I also have to have Carbonite, Dropbox, Microsoft Security Essentials, CCleaner, Defraggler, nVidia video utilities and all the other cruft you have to have to support a Windows installation. Chrome OS doesn't need all that. There's no antivirus. There's no antimalware. There's no file system to maintain. And that adds up to getting more speed and better battery life out of the same hardware, because you don't have the overhead of Windows weighing you down when all you really need is Chrome and the internet.
I don't think I can stress this too hard. The revolutionary thing about Chrome OS isn't what it includes, but what it removes. According to Google product management director Caesar Sengupta:
As that trend proceeds, Chrome OS will be a fantastic experience for them, giving them all they want from the cloud but without the legacy issues of a traditional operating system. Backups, what happens if your computer dies? Viruses or malware. Those are the parts we’re trying to solve, a machine they can use and don’t have to worry about.
What I'm saying is that most people either get Chrome OS right away or they don't get it at all. For me, it's exactly what I was looking for in my day-to-day computer experience. But people like Mary-Jo Foley and Matt Rosoff don't get Chrome OS. They're too deeply embedded in the current view of what a computer is "supposed" to be. They can't make the paradigm shift of living completely in the cloud. No one's forcing them to. But for those of us that grok Chrome OS in fullness, it's a freeing, refreshing take on computing. All the stuff we want, and none of what we don't.
Now, let's get down to business. The first thing you notice about the Cr-48 (so named because while Chromium's atomic number is 24, Chromium 48 is an unstable isotope) is nothing. Specifically, this is an anti-Macbook in terms of style. No logos, no stickers, no markings of any kind except the letters on the keyboard (which are lower case). The shell is a matte black finish with a slightly rubberized or "soft touch" feel. It's twice the weight of the 11" Macbook Air with only one additional inch of diagonal screen size and a similar solid state construction (though the battery is removable, and in fact you have to remove it to flip the dipswitch that enables “jailbroken” developer mode). Frankly, the minimalism of the design appeals to me. You can tell from the very first glance that this is a different kind of notebook, that it is what it needs to be and nothing else.
Starting up the computer takes 10-15 seconds to get to the login screen, and signing in is easy if you already have a Google account. I had heard that it didn't yet support Google Apps accounts, and it didn't let me sign in with my day job's Google Apps Premiere account, but it had no problem whatsoever with my email@example.com Google Apps account.
Once you get logged in, you’ll probably notice right away there’s no desktop. This is disorienting to some people, even if they’re used to keeping things maximized in Windows. Where the min/max/close buttons would be in Windows, you instead have time, signal and battery. You can click on each for more detail, but you can’t make the browser go away. The browser is all there is.
The “home” screen, for lack of a better term, is the new tab page (or whatever you prefer to use as a homepage). By default, this is where you’ll see all your installed apps, frequently viewed and recently closed pages. It makes a decent app launcher, not unlike the how Apple is adding iOS-style app launching to OS/X Lion, but so far it’s cumbersome in that not only is there no way to sort the apps any particular way, but they don’t even sort in the same order from computer to computer.
Instead I use an extension called App Launcher that puts an alphabetized drop down in my tool bar up by the wrench icon. Everything launched from here opens in a new tab, so it’s a quick way to find something graphically.
More often, though, I use the search key. The CapsLock key has been replaced on the Cr-48 by a key with a simple magnifying glass icon. Tap it, and you open a new tab and place the cursor in the address bar. As I mentioned before, I use my own domain for a lot of things, so it’s often faster for me to type (search)mail.kirv.in(enter) than it is to use the launcher to find the Gmail icon. In a way, this keyboard-oriented app launching harkens back to my DOS days, but in a good way.
While we’re looking at the windowing system and grousing about how you can’t minimize, can’t tile, let’s look at what Chrome OS does offer. Windows in Chrome fill the whole screen, but you do get more than one of them. Similar to virtual desktops like Spaces on OS/X, you can group related tabs into separate windows and switch between them with either the “Next Window” key on the top row or the good old Alt-Tab combination. Ctrl-Tab switches tabs within a window.
I typically have three windows, or as I think of them, workspaces, open at a time. One holds my “permanent” tabs, pages I keep open by default: Gmail, Twitter, Google Reader, Instapaper, our ticketing system and wiki when I’m at the office. I have all of these pinned as well, reducing them to just the favicon to save space and docking them to the left side of the tab row.
My second window is my fiction workspace. Here I have Google Docs, the document window for the chapter I’m working on, my own wiki hosted on Google Sites, Wikipedia, Dictionary.com and anything else I think I’ll need quick access to while I’m writing.
My third window is my media workspace. I’ll typically have just Mog pinned here, but this is where I’d open up new tabs for Hulu, YouTube, etc. I’ll get to how well these actually work on the Cr-48 later.
There’s one more user interface element to introduce. You know those little “pinned-to-the-bottom-of-the-screen” panels you see in Gmail for chat windows, tasks, etc.? Those go system-wide in Chrome OS. Pop out windows dock themselves as panels on the bottom edge of the screen, and they stay there even when you switch windows. I usually have several open. I have a panel for a scratchpad that syncs with Google Docs, another for my Google tasks, and often a clock panel that includes a handy timer for writing sprints. Other panels that I use from time to time are notifications, media player, Google Notebook, downloads, even the Chromed Bird Twitter client. These are great for things you need quick access to but don’t need taking up space all the time. The scratchpad panel itself was particularly helpful in taking notes for this article.
Now that we know our way around, what’s it like to actually use? A lot better than I expected.
It really does wake instantly from sleep. By “instantly,” I mean it’s on and ready to rock before I’ve adjusted the screen to the proper viewing angle. Sometimes it takes just a bit longer to reacquire a WiFi signal, but only a few seconds. Between that and the fact that you don’t have to wait for a spinning hard drive to wind down, I open and close the lid a lot more than I would with any Windows notebook. This is the first computer I’ve ever owned that I’ll actually consider for a quick lookup of something, the kind of task I used to shoehorn into my phone. I can pull the Cr-48 out of my bag, pop it open, look up something or type something, close it and throw it back in the bag almost as fast as doing the same with my iPhone (replacing bag with pocket). Factor in the full size keyboard and desktop-sized page rendering, and the Cr-48 is actually faster at some quick reference tasks than my iPhone.
Battery life really does seem to be about 8 hours on a charge. This will only get better as the OS gets optimized for the hardware and Google implements things like hardware h.264 rendering. And I think part of the great battery life is that I could swear there’s an ambient light sensor mounted next to the camera (or they’re just using the camera) to fine tune the screen brightness. I’ve definitely noticed brightness fluctuations in the screen, and they almost always coincide with someone walking behind me or something else that changes the amount of light hitting the screen.
I’ve heard tons of complaints about the trackpad. As with so many things in life, this isn’t really that big a deal if you put just a smidge of thought into it and adapt your stubborn behavior accordingly.
First, turn off “tap to click.” Make the system only register a mouse click when you actually press down hard enough to move the pad itself. This removes just about all of the accuracy issues I’ve seen. If the trackpad seems to be too sensitive or “fast,” you can turn that down in chrome://settings/system.
Second, the default gesture for “right” clicking on the Cr-48 is to touch something with two fingers, then click. This is nearly impossible to do consistently. The fix? With your pinkie or ring finger, whichever is more comfortable for you, hold down the Alt key while you do a normal single click with the same hand. Problem solved. Now pipe down.
The keyboard is amazing. I wasn’t sure I’d like the Sony-style (look it up, Apple stole it from Sony) “chiclet” keyboard. But this is the most comfortable keyboard I’ve used, including desktop ergonomic keyboards. The keys aren’t quite as rubbery as the casing, but they’re just the right balance of soft and traction needed to type fast and sure. I don’t have a problem with my fingers “sliding” off the keys like they do on my HP Mini 2133. Nor does the keyboard get so hot it's painful to touch like the HP. In fact, I don’t know that I’ve even heard the fan kick in on this thing.
Several people have complained that the VGA out port doesn't work. It does, but you have to push Ctrl-Fullscreen (listed as "mirror" in the Ctrl-Alt-/ cheat sheet you better be using) to move the output. And despite the implications of the word “mirror,” it doesn’t show the same image on both screens. You’re toggling back and forth between the Cr-48 display and the external monitor.
Sometimes you’ll need something a little more “under the hood.” For example, at a friend’s house I needed to supply my MAC address to get on their WiFi network. You can find this kind of propeller-head detail by going to chrome://system in a new tab. Go to chrome://flags to enable “experimental” features — isn’t this a beta OS anyway? — like the panel media player and SD card support.
In theory, you can print from the Cr-48 by first installing the beta of Chrome 9 on a Windows computer and signing into Google’s CloudPrint service. Once you’ve done that, you can print to the printers available to that Windows printer from the Cr-48, no matter where you are. I haven’t tried this myself yet, mostly because I never print anything.
At 12 inches, the Cr-48 is a bit bigger than what we usually think of as a netbook, though the rest of the hardware looks familiar: Intel Atom processor, 2GB of DDR3, 16GB SSD, minimal ports, etc. In a way, this is the first time that word has actually been properly descriptive. A notebook running Chrome OS is literally a “netbook,” in that it requires the internet to function normally. While I expect more apps over time — including Google’s own Docs — to support offline use through HTML5’s local storage API, for now the Cr-48 can’t do much of anything without an active connection to the tubes.
Speaking of which, you’re going to need the 3G, and more than the 100MB/month Google and Verizon give you for free. 100MB will get you just about a whole episode of an hour drama (so say 45 minutes without the broadcast commercials) on Hulu at 360p. Even if you plan to use WiFi as much as possible, you will inevitably want to use the device somewhere you just don’t have a WiFi connection. It’s forward thinking that Google realized this and is requiring all Chrome OS notebooks to include a 3G or better cellular radio as a backup networking source. I ponied up for the 3GB per month plan for $35. We’ll see how much of it I actually use. I’ve used about 500MB in a week, but I’ve also watched a lot of Hulu at Chipotle.
Some of you may be wondering, “why didn’t you just tether it over WiFi to your iPhone?” Well, actually that’s one of the first things I tried. The problem is that MyWi, the jailbreak app that allows you to use your iPhone as a hotspot, at current only creates an ad-hoc, or computer to computer, connection. The Cr-48 will only connect to infrastructure, or router to PC, connection. So right now, it’s just not possible to tether my iPhone to my Chrome OS notebook. That said, the folks that make MyWi are working on supporting infrastructure connections and Google is working on supporting ad-hoc connections, so at some point one or the other will make this work, and then I won’t be so worried about my Verizon data usage. Until then, I’m glad to have the Verizon 3G (actually a Gobi CDMA/GSM combo chip that will work anywhere) in the Cr-48.
If you don’t want to rely on wireless at all, you’re not completely sunk. Google says that the Cr-48 supports some — but not all — USB to ethernet adaptors. So if you find one that works, you can jack into the net over wire.
The last thing I want to mention in the networking section is that when I want to do something that I just can’t do no matter what on the web, like make changes to my Calibre ebook library or do something in iTunes, I still have the option of using LogMeIn in a fullscreen window on the Cr-48. This isn’t terribly fast, but it does work reliably for remote work. I’ve used it to do stuff on my home PC as well as remoting into the office (where I used Remote Desktop from my PC to remote into our servers; yes, I have a problem).
Okay, enough work. What about the fun stuff? The Cr-48 is no multimedia powerhouse, but I’ve been impressed so far at how good it actually is at entertainment.
Let’s get the Flash thing out of the way. Flash sucks on the Cr-48. There are a number of reasons for this. Adobe’s implementation of Flash is notoriously bad on Linux, and Chrome OS is Debian Linux under the hood. The Flash library on the Cr-48 is a special, extra-secure version. Linux has no native hardware decode libraries for h.264, the most common web video encoding standard. Neither Google nor Adobe has done anything thus far to optimize Flash for Chrome OS. And the Cr-48 sports a single-core Atom CPU. Add that all together, and you’re going to get a “sub-optimal” Flash experience.
Which is why one of my go-to extensions is FlashBlock. You’d be amazed — or maybe you wouldn’t — how much crap on web pages is Flash-based, most of it stuff you don’t even want to see in the first place. So with that extension, I only load the Flash content I actually want to see — web apps, media players, etc. — and the rest never gets rendered. Speeds things up enormously, and makes all the difference on the Cr-48.
Because Flash is so slow, you have to make certain concessions with web video. Hulu is fine and perfectly watchable windowed at 360p resolution. If you stretch the 360p video to full screen, it’s watchable, but choppy. I find turning down the “house lights” in Hulu for windowed video to work just as well at reducing distractions. 480p Hulu video is unwatchable at any setting.
Netflix doesn’t work at all on the Cr-48, because Netflix runs on Microsoft’s Silverlight rather than Flash and doesn’t support Linux — including Android — at all. I’m convinced this is less a technical restraint than Netflix being unable to prove to their licensing partners that people won’t be able to reverse engineer a Linux-based Netflix player and steal their content. And if they supported Linux, that is probably exactly what would happen. But between that, Hulu picking up more and more content, and disturbing reports about what Netflix’s corporate culture is really like, I’m not sure I’ll be a Netflix subscriber long term.
One of my big obstacles to moving into the cloud full time was iTunes. I have about 40GB of music in iTunes, so not a huge library by any means, but big enough that I couldn’t copy it over. What was I going to use for tunes? Pandora is nice and all, but sometimes you want to hear a specific song.
Then I found a review of the Cr-48 that mentioned Mog. They combine the best features of Pandora (like-artist radio, social discovery) and Rhapsody (subscription-based access to a huge catalog on demand) into a single service. They claim to have a catalog of 10 million songs, and I have to admit I’ve been surprised by some of the stuff they have. I’ve also been shocked and appalled at some of the stuff they don’t have (Only Hotel California by the Eagles? Really?). The service is $5/month if you just want to stream to PCs — including the Cr-48 — and $10/month if you want to stream to mobile iOS and Android devices as well. I’m going with the $10 plan and hoping they shore up some of their weirder content holes soon. One of the things I really like about Mog is that when you’re listening to the Pandora-like “Mog radio” for a specific artist, you can specify on a sliding scale how much to mix in similar artists, all the way down to not at all. Sometimes I want to listen to Kitaro and nothing but Kitaro.
As for the iPhone-syncing element of iTunes, I don’t miss it much. I sync or stream to my iPhone almost exclusively with cloud services already: Gmail, GDocs, Google Calendar/Tasks, Google Voice, Google Reader, Mog, Pandora, Facebook, Hulu, Picassa, Kindle, Netflix, Podcaster, Twitter, and of course Apple’s own App Store and iTunes store. I really don’t sync anything from a PC. I still haven’t updated to 4.2 because I’m jailbroken. So on the rare, once a quarter or so occasion when I actually do need to sync something from a PC, I can use my Windows gaming rig at home. (And of course, once I decide to give in and sell my soul entirely to Google by getting an Android phone, this won’t be an issue at all; everything will be OTA.)
As an irrelevant aside, I ended my 10 year subscription to Audible this month when they continually refused to charge my monthly fee to the new debit card I’d entered into their system. Despite multiple calls to customer service and them swearing up and down that the billing issue was resolved, they continued to charge my old card on a shut-down account and then acted like it was my fault the charge didn’t go through, like I hadn’t told them to use the new card multiple times. Couple that with realizing that I was listening to audiobooks as a way to avoid actually thinking during what limited downtime I have as a man with a smartphone, an always connected notebook and a Kindle, and I decided not to keep fighting it. And as it turns out, I’m paying less for Hulu Plus, Netflix and Mog combined then I was paying Audible every month. I think I’m getting a lot more value out of that money now.
Okay, fine, you’re thinking. So you can edit documents, presentations and spreadsheets, you can listen to music and watch videos, but you can’t do everything in the cloud. What about about Photoshop? Huh, smart guy? What about Audacity for recording podcasts? Ha! you may be saying.
For Photoshop, I prefer Picnik, which is really more like Photoshop Elements or Paint.NET. Simple stuff, but it does support layers, which is the vital feature I need for putting together book covers. More advanced graphics editing needs require heavier tools like Pixlr or Aviary. This last one in particular is an amazing web-based creative suite with raster and vector image editing, color pallete and filter creation, music mixing (Garage Band) and yes, even Audacity-style audio recording and editing. Like, for say, podcasts.
So is Google right? Is this the future of computing? Maybe.
It’s worth noting that Google will be the first to point out that the entire OS is like a web app. In that what you get when you first take it out of the box is the worst it’s ever going to be. The whole OS is going to get constantly updated, iterated and improved. Transparently, quietly, in the background. New features will just appear like they do in Google’s other products. I’ve already seen one Chrome OS update and it was just as quick and painless as it is in the Chrome browser on Windows (and you’re even notified about it the same way, a small gold dot over the wrench icon). Most PCs start off fast and clean and then degrade over time. Google is looking to reverse that trend. Not just stop it. Reverse it.
I’ve beta tested a lot of operating systems, going back at least to IBM’s OS/2 2.0 in 1992. As beta OSes go, Chrome OS is alarmingly stable. I’ve yet to get it to crash. I did push it over the edge into instability once, where extensions kept crashing in cascades and I had to reboot the notebook. But to get there I spent days installing and uninstalling dozens of extensions and scores of themes, had a half a dozen windows open, a couple with over 30 tabs each. And even then, the OS itself kept going, the engine of the browser itself never faltered. Just the extensions got squirrelly, and a simple reboot took everything back to normal and even restored the windows and tabs I had open when I shut down. This thing is solid, and it’s a beta. This is the kind of robust, bullet-proof performance you need when handing a computer to your computer-illiterate uncle. Yeah, you know the one. Him. He couldn’t break this. Think about that.
Some things are still missing, obviously. There is no official Skype client for Chrome OS, but imo.im supports Skype just fine. And obviously Google Talk video calling works. SD card and USB drive support is still experimental, one of the risky features you have to enable via chrome://flags and even then doesn’t work right. Yet. The system isn’t fully baked, and Google told us that up front.
As I mentioned above, many apps don’t support offline mode, including Google Docs. In Docs’s case, this is because it used to support offline access via Google Gears, and they’ve removed Gears support to replace it with the more standard HTML5 offline storage. It’s just not ready yet. I’d say the smart money is on that feature working flawlessly by the time consumer Chrome OS notebooks hit the streets next summer. I’ve seen a few reports of lag when typing in Google Docs, but I haven’t seen any evidence of this myself, and this entire gargantuan article was written in Google Docs.
Some people object to Chrome OS on principle. Richard Stallman is worried that reliance on cloud services will leave people a slave to their service providers, unable to control access to their own data. While he has an academic point, I don’t see it as being that much of problem in the real world. It’s easy enough for me to backup my data outside of Google’s cloud if I so choose, and the convenience and utility of a cloud OS far outweighs any philosophical disadvantages. At least they do to me.
Marco Arment (of Tumblr and Instapaper) has a different issue. He’s concerned about enterprise adoption and that with their track record of killing products that they lose interest in — Wave, Buzz, Notebook, etc. — Google has a hard sell in front of them getting businesses to even try switching over to Chrome OS. And he has a point. Most of the computers at my day job are still running XP, for crying out loud.
But I don’t think that’s where Google is heading. While you could choose to see Chrome OS as the ultimate manifestation of Larry Ellison’s network computer vision, I think it’s more likely that as I mentioned above, Chrome OS will initially be marketed as a “computer for normal people.” They’ll go for consumer adoption first, and then IT departments will be forced to integrate the devices over time as they’ve done with the iPhone and iPad. Chrome OS will be an enterprise platform eventually, but it will come in the back door.
I’ve also seen several people assert that Chrome OS is just a transitional experiment, doomed to be eventually replaced by Android. Why would Google maintain two operating systems when one could do the job, they say. Well, you might want to ask Microsoft why they have Windows 7 and Windows Phone, or better yet, ask Apple why they have Mac OS and iOS.
Fundamentally different form factors demand fundamentally different platforms. The user interface between a keyboard-based device like a notebook and a touch-based device like a tablet have to be radically different to suit the form factor. This is why Windows on a tablet has been such an abysmal failure for nearly a decade. A user interface designed for a mouse and keyboard just doesn’t work on a touch device, no matter how you try to cram it in there.
And the converse also applies. A system designed around a touch screen, like Android as we currently know it — recall that initially Android was intended for Blackberry-like thumbboard devices, and diverged sharply after the iPhone was announced — is just not going to be an ideal experience on a notebook. And we’re not just talking about the system itself. We’re also talking about every single application. You know, all the apps that people give as the reason for ditching Chrome OS in favor of Android in the first place. Those apps, especially the really good ones, are designed for a multitouch interface. They’re going to be awkward at best, unusable at worst on a notebook form factor. There’s a reason Microsoft and Apple have different platforms for different form factors, and Google would be wise to continue following their lead.
Actually, anything but final. I’ll be posting more of my observations here over time, but I think it’s safe to say I’m floored by this thing so far. It’s a fundamental change in computing when coupled with the persistence of data in web apps like Google Docs. When writing this article, I would switch freely between computers and pick up where I left off. Write on the Cr-48 when away from my desk, pop back into my cubicle, toss the notebook in the bag, unlock my Windows PC and just keep typing. It’s amazing.
After a week of use, the Cr-48 is my primary, every day computer. But I can’t say I’m going to revise Revelation on it entirely because doing so would negate one of the coolest things that Chrome OS makes possible, the “write anywhere” versatility of Google Docs. But I will say that the Cr-48 has already earned a place in my default gear set. Everywhere I go, I take my iPhone, my Kindle and my Cr-48. I didn’t do that with my HP Mini. I haven’t done that with any notebook.
But this one’s different.