A Very Chrome-y Christmas

The Google Cr-48 on vacation in Marco Island, Florida

As I type this, I’m sitting on the balcony of my vacation timeshare in Florida overlooking the Gulf of Mexico.

I considered leaving all of my primary computing devices home this Christmas vacation, and seeing if I could “Rough it” with just our Android-based smartphones.

Instead, I decided I would compromise — this vacation, I left my prized iPad and my work laptop PC home, and brought along two unproven traveling companions — a Google Cr-48 Chrome OS-powered notebook and another device which I’ll discuss in a future article.

As it has been widely reported already by ZDNet and other technology news outlets, the Google Cr-48 is a limited-production notebook computer, of which only 60,000 have been produced and which will serve as a technology demonstrator and large beta test of Chrome OS, Google’s 100-percent Cloud-dependent operating system.

Beta testers are invited to apply to join the program, but only a lucky few will be chosen. Only certain members of the press were given access to these systems — as it turns out, I was not one of them and Google has declined my request to put the device through an extensive formal review process, citing availability.

However, one of my industry colleagues who received the unit to evaluate felt I could do a much better job putting the Cr-48 through its paces than they could, so we met the night before I departed for Florida to share dinner and drinks and to transfer the goods. My wife and I have been using it as my main computing device on vacation ever since.

Chrome OS is little more than a Linux kernel which allows the core services and the hardware drivers of system to function and to support its only “App” and central UI, the Google Chrome 9.0 browser. The Chrome browser in turn acts as a front-end to Google’s services, such as GMail, Google Docs and their Web Store, which allows a multitude of other web-based apps to be plugged into a centralized menu screen.

If you know how to use a web browser, you pretty much already know how to use Chrome OS.

While Chrome OS’s source code is Open Source, you can’t easily install the latest Google build of the software on any old x86 hardware yet like you can with a typical Linux distribution such as Ubuntu or Fedora.

It is expected that other vendors are going to pre-load this software on future mobile computing products. This begs the question of whether or not Google is actually going to provide some sort of installable “distribution” for hobbyists and OEMs, or if a 3rd-party is going to take the source code and run with it, such as with the “Hexxeh” builds that were made available by a UK-based college student this last year.

As it is a pilot program, the Cr-48 is meant to be something of a reference platform for future mass-produced Chrome OS netbook hardware, but to say that the template is set in stone is highly unlikely. There will almost certainly be different variations on screen sizes, battery life and connectivity options for Chrome OS-based products. Additionally, it would not surprise me to see this platform installed on desktop Thin Clients, such as on devices produced by WYSE and similar manufacturers addressing that vertical market as well as for end-users in the consumer space.

The hardware on the Cr-48 is specifically designed to take advantage of this minimalist approach. The Pegatron-produced device resembles a standard Intel Atom-based netbook computer with a 12″ screen, but that is where the similarities end. For example, the keyboard has been optimized to discard all legacy keys from Windows and even the Mac.

All of the function “F” keys have been eliminated, and replaced with browser action keys, such as Page Back/Forward, Page Reload, Fullscreen toggle and Window toggle. The Caps Lock key has also been eliminated, perhaps as a nod to conventional Internet messaging etiquette.

For mouse navigation, Google has provided an oversized trackpad which can either be lightly tapped to confirm dialog activity or mechanically depressed. There are no mouse buttons, and the “Right-click” function is done by either using “Alt-tap” or “Alt-depress” or actually hooking up a two-button mouse to the notebook’s single USB port. Cursor navigation is provided using arrow directional keys.

In my use of the device, I’ve found that an inexpensive two button optical mouse with scroller (which I purchased from the local Radio Shack down here in Marco Island) goes a long way towards preserving your sanity and vastly improves the overall experience using Chrome OS, particularly if you are used to working with PCs and/or the Chrome browser on Linux or Windows.

Another area in which this laptop differs from its Windows and Mac cousins is how it normally handles storage, or rather the lack of storage. While the system has an integrated 16GB flash storage device, it is primarily used to store and boot the OS and to cache user profiles, browser data and Web Store apps — the end-user by default has no control over and no ability to manipulate the local Linux file system. This is done by design as to force the user into a 100 percent Cloud-enabled experience and also to prevent end-user tampering of any kind.

[EDIT: The Chrome OS file system can be manipulated, but only after enabling a special developer setting in about:flags]

In addition to tamper-free local storage, the Cr-48 has no optical storage either. So if you need to upload data, such as digital photos to Cloud-based services such as Picasa Web Albums or Flickr (which I needed to do in order to produce this article’s artwork) you have to access it via the USB interface.

This is where we get into some very interesting usage scenarios, and where Google and its hardware and service partners need to put a bit more thought into how their customers will interact with Chrome OS devices.

I wanted to be able to upload my vacation photos to Flickr, which is my preferred image hosting provider. At home, I use a batch uploader application on my Linux and Windows workstations to do this. On Chrome OS, you’re limited to what you can do inside a browser. This wouldn’t be so much of a problem if it weren’t for a number of limitations that have been introduced which effect user behavior.

To edit and view the photos, I inserted a small USB SD card reader with my camera shots into the Cr-48’s single USB port. Since I’m using a wired mouse, I needed to disconnect it. A word of wise to future Chrome OS hardware OEMs — please put multiple USB ports on your devices and provide for Bluetooth mouse connectivity.

On Linux and Windows, I’d simply drag and drop these photos directly into my batch uploader from my home PC, and then do the required photo editing on Flickr using a service like Picink, since at home I’m not concerned about bandwidth with my unlimited broadband connection. However, when I’m on the road with my regular work laptop, I’ll do some editing locally with a program like Picasa or GIMP, compress the photos down to a manageable size from several megabytes down to a few hundred kilobytes, and then batch upload.

However, with Chrome OS and the Cr-48, you need to give this a bit more thought. You can locally edit the photos on your SD card withPicnik installed as an app from the Chrome Web Store, but you can’t batch upload to Flickr or Picasa Web, you can only do it one at a time. Additionally, when you open the photos from the SD card into the editor, the Linux file system is fully exposed with all of its subdirectories.

If you were a complete Linux neophyte didn’t know that /media was where your USB SD reader was mounted, you’d probably be totally confused. Google needs to fix this so that only the essential directories relevant to the USB reader are shown.

Then there is the issue of bandwidth optimization and data consumption.

One of the things where the Cr-48 distinguishes itself is its integrated Verizon 3G connection. As part of the pilot, Google is guaranteeing 100 megabytes of free Verizon data per month for two years. There are other plans available, but I expect that many Cr-48 users such as myself are going to opt for the free connectivity rather than activate yet another cellular plan. The unit also has a Wireless-N Wi-Fi integrated as well, which is fine if you are in an area with good access point saturation such as in my home, at an airport lounge or at an Internet cafe like a Starbucks.

Unfortunately at the condo complex on Marco Island where I am staying at the Wi-Fi signal is awful, so I opted to activate the Verizon 3G.

Within two days, I saw my 100MB drop to 85MB — and that’s not by doing anything particularly bandwidth intensive, that’s just by using GMail and regular web browsing. Fortunately, I had also brought my two Android phones with me, both of which are capable of being used as mobile 3G/4G Wi-Fi access points. For the average Chrome OS user, I think that this is going to be a much better and more economical option than using the integrated connection.

However, it’s not just an issue of bandwidth consumption that needs to be addressed. Since this is an entirely Cloud dependent OS with no local apps whatsoever, and no “Store and Forward” or offline capability, I think Google needs to figure out how to handle data compression so that end-users don’t start saturating networks with these devices or bog down device performance.

For example, Google could proxy out all web connections and pre-fetch large JPG images and compress/cache them before the browser loads them, where you might have dozens of such images on a page. When you on the fly compress a typical JPG down from 500Kb or 1MB down do 80Kb, this could very well make the difference between 100MB-200MB of monthly data consumption offered at inexpensive price points by wireless carriers and 2GB-5GB of data at $50-$100 per month.

All it takes is a couple of careless content providers to forget to compress a whole mess of JPG files and you’ve chewed up your monthly 100MB allotment pretty quickly.

In addition to the issue of data compression that needs to be addressed with Chrome OS there’s the issue of how to handle notification tasks and routine connectivity. While my email notification and Facebook activity was handled by a GMail and FaceBook notifier applet, and I was able to do my Twitter tasks via the Twitter web UI and also via TweetDeck’s latest Chrome extension, I had no such equivalent for Instant Messaging. To run IM, I had to open up a separate window and connect to the Web-hosted IM+ app from ShapeServices.

Unfortunately, that service has been having intermittent connectivity issues. While Google may want to provide all of Chrome OS’s apps and services via the browser, I’m not sure that’s either practical or realistic — they probably should integrate IM into the OS, such as with one of the established Linux IM clients such as Pidgin or Empathy in order to provide multi-service IM support. And I think that integrating the Linux version of Picasa with the appropriate photo-upload and photo compression tools are probably a good idea as well.

Still, even with the connectivity kinks and the beta nature of the Chrome OS, I do believe that what we are seeing here represents the future of computing, once these issues are ironed out and Wi-Fi and 4G connectivity becomes a bit more ubiquitous. I’m really enjoying using the Cr-48 and I can’t wait to experiment with Chrome OS on other hardware once the software is in a more readily installable state.

Have you had any experiences with Chrome OS yet? Talk Back and Let Me Know.