For those unfamiliar, Chrome OS is a new operating system from Google, but it’s not intended to be general-purpose; it’s intended to do little other than get you on the web as quickly as possible. Consequently, its UI is basically nothing but the Google Chrome browser, a time display, and wireless and battery icons. It’s questionable whether it can function as someone’s primary laptop (Google even has a handy quiz to help you decide); it turns out I don’t do much off the web, besides playing some games and writing code, and both of those things are better served by a more powerful and less mobile machine.

The Cr-48 Experience

So what’s Chrome OS’s web experience like? Well, just like any other machine’s, really; you probably won’t find any surprises. Flash is built in; we don’t get Silverlight (it’s built on a stripped-down Linux kernel, and when has Microsoft ever offered anything for Linux?), so there’s no Netflix streaming, but pretty much everything else works, including Youtube with HTML5 video, Flash games and toys on Newgrounds, Google Docs/Gmail/Maps/Voice–everything.

The problem is, the Cr-48 just isn’t very fast. Opening multiple links at a time can easily bring the machine to a halt for a few seconds, and it has quite a bit of trouble keeping up a decent framerate on Youtube videos (though it does appear to fare somewhat better with Youtube’s HTML5 mode). Even Chrome’s new tab page takes maybe a second to load all of its contents. Since I’m used to more or less instantaneous response times in Chrome on my work machine and laptop (both of which run Ubuntu with all the crazy graphical effects, the latter of which being a three-year-old Dell XPS), it’s hard to justify using the Cr-48 as a primary work machine. Still, I don’t blame Chrome OS for this; the Cr-48 is a beta product (with far from the best internals; they’re shipping a whole lot of these out for free), and I’m assuming that the upcoming stable Chrome OS machines will be quite a bit faster.

What you don’t get is access to the machine’s internals. I haven’t yet found any file manager, besides a list of downloads (EDIT: and a file selection box that showed the entire filesystem I got by accident that I’m pretty sure I wasn’t supposed to see); there’s no text editor, and I haven’t yet figured out how to use the media player (which is mentioned in a menu as “experimental”). The most you get is a list of technical files and logs that Linux users might recognize at chrome://system, such as process lists (which mainly lists a bunch of Chrome browser processes), CPU info, memory stats, etc.

Crosh?

There is a terminal, accessible by pressing Ctrl-Alt-T. It’s not like any shell I’ve ever seen before though; its prompt calls it “crosh” (presumably “ChROme SHell”?) and all it seems to allow is a few specialized commands for network and hardware troubleshooting–none of the standard file system tools (not even ls, cd or pwd). One boon that they did include, however, was (stripped-down) SSH. This allows me to easily remotely control any server I might want to (as long as I don’t need public key authentication), and I can see how this would be an amazing tool for sysadmins. Clearly this machine was designed by geeks.

All in all, this machine’s customizability is at about iPad level. Not that I expected more; it’s designed for nothing more than getting on the web quickly, and it certainly accomplishes that, and really, if all the machine is designed for is the web, SSH is frankly more than I expected (though quite welcome). But it’d be nice to be able to swap out the crosh thing for something I can use.

UI Quibbling

Google has tweaked the standard keyboard layout in ways that might surprise the average user:

For one thing, there’s no capslock key. In its place is a “search” key, with an icon reminiscent of Android’s search button, which really just opens a new tab (and lets you do a search by just typing your query into the URL bar, like in Chrome on any other platform). This is a delightfully welcome change, since I was never using capslock anyway (having, y’know, finished puberty years ago), and I’ve already become dependent on its replacement as of this writing. (Google even jokes in the setup notes that if you really want to, you can set it to function as a capslock key “so you can post an INSIGHTFUL COMMENT ON YOUTUBE.”)

The other notable change is the function key row. Since those don’t really have any use when the entire interface is a browser, Google has replaced them with a set of function-specific keys. In order, they are: Back, Forward, Reload, Fullscreen (all for web browsing, and welcome additions), Switch Window (for multiple browser tab sets), Brightness Down/Up, Volume Mute/Down/Up and Power. I just wish it had keys for switching tabs, since the keyboard shortcut for it is Ctrl-Tab, which is somewhat cumbersome.

On another note, the touchpad is maddening. I don’t know why all the touchpads available nowadays are the kind with no separate “button”, but I have a lot of trouble with this one–I can’t click and drag properly (which is surprisingly important on today’s Javascript-heavy web, to say nothing of highlighting to copy and paste), or middle-click to open something in a new tab, or even right-click! This somewhat detracts, as it means I have difficulty using the machine while, say, reclining on a couch, when I can’t really use a mouse. It’d be nice to see a software update make it better (I’ve seen two OS upgrades today already), since I’m pretty sure this one is a software issue.

Speaking of mice, the Cr-48′s ports are fairly sparse. We have a standard headphone jack, one USB port, a VGA port for a second monitor, a power jack and an SD card slot. There’s no Ethernet (somewhat surprising, but then, I guess everyone has wifi these days), and I’m not really sure what the SD card slot is for since I don’t have any SD cards to test it with. Google shows again here that there’s a lot of cruft worth trimming on the hardware that we normally buy; nowhere here do we see the common perplexing set of colorful audio ports.

Wrap-up

Chrome OS represents a radical take on the web. It’s clear from feeling perfectly at home on it after a day of use that I sufficiently “live on the web” (at least for recreation). There’s no denying it has kinks to iron out, most notably the touchpad (it remains to be seen whether the hardware or the software is the issue) and its speed (probably the hardware). I also have nothing but praise for Google regarding this beta program–not least because I got a free notebook out of it–but also because it handles the issues Google tends to have with UIs, by aggressively putting the paper cuts on display for users to point out, criticize and work through before release.

I’ll keep using this machine and submit plenty of feedback like Google requested, starting with this post. As for the ethical issues surrounding it, those will have to wait for another one. Also, if you think you’d enjoy living on the web, you should sign up for the pilot program yourself–it’s open until Dec. 21, and who knows, you might be as lucky as I was.

Can you imagine a scenario where Apple releases a fully-fledged laptop locked down like the iPhone or iPad? A fully fledged computer on which the only way to install software is through Apple’s built in App Store?

To borrow an overused phrase, it’s more likely than you think.

Would you buy it? Would other people buy it?

Package management can be awesome

I’ve written about package management before. I believe that locking down a package manager such that the user can’t change the available software channels is unethical and flies in the face of years of innovation. We’ve always treated computers as though they were owned by their users, and I see the popularization of devices like the iPhone and iPad as a disturbing precedent. Might Apple be emboldened to try locking down full computers the same way?

We live in an age in which the technology is powerful enough to implement such a thing. Apple is certainly capable of running the servers necessary for supplying its users with all their software, and it would certainly be lucrative, given how successful the iPhone’s app store is and how much 3rd party pay software exists for Macs right now (which Apple would then be able to take a cut of the profits on). There is thus a clear financial incentive for Apple to set up such a system, and for developers to continue to target it (and submit to their harsh developer restrictions).

And let’s face it: such a system would be amazing for users who don’t care. They would enjoy Apple’s rigorous testing process for all their software, and thus enjoy machines that almost never become bogged down by slow or unstable apps. They’d have an easy way of finding and installing all the software they could ever want–even simpler than what they currently have on Macs, and that’s saying something–along with Apple’s approval of all of it, which carries all the quality control that their name has become synonymous with.

I can imagine other possible models. Maybe they’d offer people “simple” or “advanced” options when purchasing computers, the former being cheaper and locked down for the “I really don’t know how to use computers” crowd. Such a thing doesn’t seem like them though, given how well they’ve avoided fragmenting their product ecosystem. Or perhaps they’d simply release their own blessed way to jailbreak the system if a user cares enough.

Right, jailbreaking! Can’t you just jailbreak it?

A cursory glance at the history of the iPhone would make one all but certain that if Apple released such a system without offering a way to unlock it, their users would find a way (as they will undoubtedly with the iPad).

I find this scenario similarly unpleasant, for two reasons. One is that if your enjoyment of a machine depends on an unauthorized hack, you are entirely dependent on those hackers to keep it jailbroken the next time Apple pushes out a system update that invariably locks you down again–the ease of which is never a certainty. It’s a cat-and-mouse game that some users appear more than happy to play, but in the long run it definitely doesn’t seem worth the effort to me.

The other is that you’re ignoring the device’s biggest selling point. If a device is locked down, it’s locked down as a genuine means of quality control, which Apple has shown itself to be superb at. To my mind, jailbreaking an Apple device invalidates everything good about it–Apple’s assurance that you’re getting the best you can get.

This argument may seem self-contradictory–suddenly a device being locked down is a good thing? It’s important to remember that the reasons for locking down a device in the first place are, in order:

1. Making a cut off of lucrative third-party software sales.
2. Making sure that 3rd-party software enriches the platform rather than makes it look bad.
3. Making the system easy to use.
4. Making the system stable, secure and fast.

For most users, a locked down system is protection from systems they don’t care about knowing how to use, and from software that might harm their systems. To users who care about customization, it’s a heavy-handed restriction on something they want to use that just gets in their way. Users in the former category genuinely benefit from a locked down system, and users in the latter category don’t.

So how likely is it?

I’ve always felt that Apple’s primary market is the former category, and that’s why I think this scenario is so likely. It worries me, because I’ve always clung to the idealistic notion that someone who uses a computer has a responsibility to know how to make changes to how it functions. An Apple user might counter that all Apple’s lockdowns do is simplify the method for those changes so more people have access to them. I would counter that such a thing is worthless if more fine-grained customizability is lost in the process.

But then, I suppose that’s why I’m an Ubuntu user. I have the tools I need to change where my software is coming from, and I use them. I just wish more people noticed or cared.

This most recent change consists of a revamp for Facebook’s already-substantial privacy settings. This change didn’t actually affect anyone’s settings who didn’t tell it to; users were greeted with an unavoidable menu asking them if they wanted to keep their old settings or switch them to new, simplified privacy categories–the default of which in a number of them was “Everyone”.

This is, to be perfectly fair, not something I really have a problem with. While “Everyone” was preselected for some users (UPDATE: apparently not all, since for some at least, it was set to Original Settings), they gave you all the information you need to decide whether that was actually a good thing. Plus, I get why they’re doing it; they want all the data they have on their users to be available to search engines and marketers so they can monetize it, and so they can position themselves as the search result that people might want to come up on Google searches for their users’ names. Monetizing profile data without incurring the wrath of privacy advocates is something they’ve been doing for years through, among other things, their own targeted advertising network (the one that’s known for selling sex to men and weight loss to women), the ill-conceived and ill-fated Beacon, Facebook Connect, and the notorious ads that use friend connections to make it appear as though a user’s friends endorse a product.

Many of Facebook’s users tend to forget that Facebook is neither free nor intended as a public service. They’re in the game to make money, and the data they’ve amassed on their users is worth a fortune to the right people. With that in mind, I don’t really blame them for encouraging their users to make their profile data public. (As long as, y’know, they’re not doing anything shady on the side in complete violation of their privacy policy.)

On the other hand, I imagine many of their users neither take the time to learn about how far their profile data can go nor care about the issue, and might very well absently click their way through the menu without thinking about what “Everyone” actually means. Frankly, I have little sympathy for those users, but the EFF disagrees. I would be quite interested in statistics on how many users actually changed their settings to Everyone as a result of this menu–those would presumably be the ones who just didn’t want to be bothered and thought “yeah, whatever”.

It just goes to show: as always, there is no replacement for a smart user. Internet companies will monetize however they can. It’s up to users to decide where they want their data to go, or if they even care.

Edited: Fixed a minor typo. Also, I’ve received reports that Everyone wasn’t always preselected, which is quite significant for the “yeah, whatever” cases.

No matter how secure your system is, there’s never a replacement for a well-educated user who knows to at least be wary of untrusted software packages. While a smart security framework can make a user pause before making a mistake, it’s ultimately still at the user’s mercy. (Though having such a system in place is certainly much better than not having one, or providing a false sense of security.) A Linux user who joyfully installs every package offered is really in no safer a position than… well, the majority of Windows users.

Maybe he just doesn’t want to. Maybe he just wants to go back to the good old days when a news establishment couldn’t be challenged by people working in their basements. Or maybe he genuinely does believe he can completely block off the WSJ’s content with a pay wall and have people actually pay for something they’ve always gotten for free and could just as easily continue to get for free from elsewhere.

But for better or worse, the Internet is here to stay. He’d be better off, y’know, innovating. Instead of shouting about pay walls and blocking his sources of traffic and trying to force his competitors to do the same thing to avoid losing his readership over it.

By the way, traditional news companies can innovate. The New York Times is doing an excellent job of this, even releasing public APIs for other developers making mashups from their content! The BBC even released an open source Javascript library for other web developers to benefit from their work!

But if Rupert Murdoch wants to try to wish away the Internet, I’ve no problem with that. I’ll just keep reading the NYTimes, giving them advertising revenue and link traffic. I might even buy a subscription. Maybe even through TimesReader, which runs on all platforms where Adobe AIR does.