The announcement last week that Google is developing (another) operating system caused a frenzy of excitement, unfortunately spilling over into some hostility towards the Linux community. There has been some rather silly talk about "real people" and "typical users", but the problem with these terms is they are used (in my experience) by people who don't consider themselves in these categories, but are somehow able to speak for them. In an era of general purpose operating systems everybody is a "typical user", and in thinking about potential new operating systems this provides a useful basis on which to evaluate new ideas. I've been thinking a lot about Google's Chrome OS and can see a future for it, but one very different from the ideas currently being talked about. I think it has no future in the home, but a bright future in the workplace.
Let's think first of home use. There is an idea floating around that you can have all of your applications in the cloud and that you can get by with just a browser, but that idea is fatally flawed for home users in many ways. Take a walk around you local computing superstore (yes the one normal people use) some day and you can see evidence of the diverse ways people use their computers in the home. Of course they are using their machines to shop online, browse the web and read email, but they are also using their computer for other tasks, editing photos, making home movies, playing games, editing their parish newsletters and so on. Theoretically you can do all of these things using web based applications, and maybe one day people will do less and less of these tasks locally on their machines, but for now it is much quicker to engage in these, often data-intensive, applications using software located on your own machine. The economics are in your favour for working in this way too; the cost of hardware is in freefall - you can pick up a fairly powerful machine for a few hundred pounds (or if you take a mobile broadband deal a monthly payment often no more expensive than a contract phone). Add to that the possibility that the software might be inexpensive, or in the case of open source software free to the end user, you get a situation where you might as well use your machine to its full potential.
Take a look at what happens when you try to do all of this using applications in the cloud and the picture is less happy. You don't have to go far to find problems, in fact it all starts just beyond your telephone socket. Many people now get their broadband through an ADSL connection. A few lucky folk have cable connections, but ADSL is very common as it means that no new cable has to be installed to get a broadband Internet to a person's home, it can be piped through an existing copper wire connection. This is great, and has brought faster Internet to many people who otherwise would not have been able to benefit from it, but ADSL is essentially a workaround, and occasionally things go wrong. When they do it can be a long and miserable wait for your telecoms company to fix it, a wait potentially of days. In the thin operating system bootstrapping you to a cloud computing world scenario you would be left with a computer that would be no more use than a brick in this situation.
When your ADSL connection does work, like many other solutions it can be a bit of a bottleneck Once you add in such things as the contention ratio, a couple of teenagers in your house watching Gangsta rap videos, and somebody else making phone calls over the Internet you might find yourself not so much on the Information Superhighway but on a congested 'A' road crawling through a village. It's not a situation that is going to improve soon, the Digital Britain report in the UK recommended a universal service provision of only 2Mbps, not a high enough speed to avoid such issues. These aren't the only problems though, it is still unclear how a lot of online services expect to make money. It is often possible to use these services for free thanks to the current fondness for the freemium business model, but now some service providers are finding that this isn't working for them. These companies need to charge somebody something at some point so they can carry on meeting expensive costs of paying wages, developing their product and maintaining server capacity.
So it is clear for the home user the idea of "doing everything in the cloud" truely stinks, and Google Chrome OS (or any other OS implementing the same idea) might not be such a great choice. And for a while I thought that might be it, a bit of masterful communication to steal thunder from Microsoft (who apparently announced something the same day), but it only struck me, once the dust had settled that there is a whole different angle to this story and one where Google Chrome OS could really succeed - business use. Maybe we have been too consumed with arguing over the idea of "typical users" using this on their netbooks to see what would be a coup of an astonishing scale for Google - getting office users off Microsoft Windows and on to their platform.
This idea might seem ridiculous, after all aren't businesses quite conservative with their ideas about desktop computing technology? They like things that are tried and tested, that their technicians can do exams in, and people, to whom computing means Windows, can just get on and use (as long as you never switch the version of Windows you use!), and if free was such a factor wouldn't they swap to a Linux distribution straight away? It's easy to get lost in a present situation and assume that it will be the case forever. I remember a time when people talked of the battle between Commodore and Atari, and now neither exists in the context in which many of us grew up with those brands. Things change. If you want to find somewhere where people use their computers to browse the web and read email look in many offices. Many companies have been implementing their systems over the web, the only programs a lot of employees will be using are a browser, an email client, and an office suite. Google can provide all three even today.
The status quo is not an ideal position for business, it costs so much in so many different ways. The problems with Windows start right from a person logging in first thing in the morning. So your six person team is waiting ten minutes for Windows to start and do its thing - congratulations, you've just lost a man hour! Every computer in a business has a full operating system on it that will be need updating patching, virus checking. It will need a team of people on a help desk to help with problems, roving technicians to fix problems. If a computer breaks, you can't just swap it because that person has a particular software suite on it, they've probably stored some important files on it too, and while it is being fixed you might well lose that employee's productivity for the day. Not a great situation, and to add to the fun the company will be having to pay software licences for all of this, not to mention support costs.
So what would happen if we deployed Google Chrome OS in this scenario? Things get interesting then, for a start it wouldn't need such a powerful machine to run on, something like a nettop would be more than adequate. The individual machines would stop being so individual, if one breaks you would just reimage it or replace it. There would be no local customisation so you could swap the machine on the spot and the employee would lose no time waiting for their technicians to repair their machines or having to customise their machine again once it returns. As the OS would do little more than start a web browser there would be less code to patch, less to go wrong, IT departments would be spending less time making sure that regular events like Patch Tuesday don't break their internal systems. Efforts would centre on systems deployed on servers leading to a more efficient economy of scale. The company would feel the benefits right away, if Google are right about a very short boot time that lost time in the morning could be recovered. Multiply these benefits by the thousands of employees that large corporates have and you are talking about a factor that is not about doing what is technologically right or a better solution as such, but something that can give you an edge over a competitor. All of this is delivered with the comfort of a big brand name, after all, in a few years nobody will get fired for buying Google right?
A question that is often asked about Google Chrome OS is whether it will be bad for Linux? I don't believe it will. In fact I am no longer convinced that question is still an issue. The big prize is the humble office PC, a world where Lotus 1-2-3 and WordPerfect once mattered but Microsoft came to dominate, a world where paying for a support contract is the norm and paying license fees is done diligently. It fits perfectly with the idea of Chrome OS.
At this point I should dutifully point out that I am doing no more than speculating about an unreleased product, none of this might work once we see it. After all, we've been here before when Google announced a "fresh take on the browser" but ended up with something that was essentially the same as any other browser. None of this has dented my belief that the future is Linux.