There is a smell in the air tells me that people are starting to think that living and working in the so-called Cloud is already well on its way. In case you don’t know what this means, its simple. Google is a part of the cloud. If, when you login to your computer at your desk, you use all of what Google offers and never use a desktop application to get what you need done, you are working partly in the Cloud. The Cloud, after all, is an old term for the network. It really refers to the public network, more widely referred to as the Internet (or Interwebs by some dorks who can’t stop making fun of former Alaskan Senator Ted Stevens (R) and part-time stand-up comedian).
Therein lies the problem. It pervades everything we do in relation to anything computing. We login to our computers to get to the cloud. In a nutshell, as long as we use monolithic desktop environments to store files, run applications, communicate with others, and even to just access the web, we will never be completely in the Cloud. It is our desktop computers, like any other node on any network, which handles all of the network communications through user oriented applications, those of which are built on programmatic environments which require access to a load of data stored directly on your hard drive.
All of this happens locally. In a typical Windows XP installation we’re looking at about 1.5GB of data just for the operating system. If you are like 40% of computer users around the world, you use Firefox instead of Internet Exploder for Windows or Safari Mac OS X, so that involves another download of around 8MBs and it takes up around 28MBs of space installed. Do you use FTP? Need a client. What about email? Install Office for Windows or Mac for several hundred MBs to a little over 1GB. What about things which don’t work well on anything less than a ridiculously wicked fast network like audio or video editing?
If that’s not enough, there’s something that a desktop OS does which we pay no attention to, but would miss instantly were we to lose it. The very same was true with Apple’s iPhone. What did people scream and rant about for a couple years before Apple had to finally implement it? Copy and paste. If you were to work in the Cloud, what would you use as an intermediary between these services? How, for example, would you take an audio file, say an MP3 of a live recording you made on your phone of Weird All Yankovich jamming with BNL at The Grove in LA, and get that into an email in Gmail? Well, you could upload it to some Cloud storage service which supported your phone, then if Gmail supported URLs for attachments (assuming the Cloud file storage allowed you to make them) you could paste it in.
Think about how easy your supposedly ridiculously out-dated and code heavy monolithic desktop operating system is now in light of where we’re at with the Cloud. We are nowhere near being able to work in the Cloud. Sure, we can do a number of limited things, but these are just proxies for our real desktop systems. If there was one service which could be reasonably close to providing a workable Cloud environment circa-1999 it would be Google, except they don’t have Cloud storage. Besides, each of their so-called apps are very simple and mostly self-contained. If you have not had the pleasure of using Google Apps, you might be surprised to know that the only apps which do work are mail, calendar, docs, sites, and contacts. You don’t get anything else, and if you use them, you have to login to a standard account to keep your stuff.
That means that, even in the Cloud, you’re stuck with more than one place worth of data. Useless redundancy which serves no purpose. Even Microsoft’s hideously painful SharePoint server is more inclusive than that, but its still not usable because its horrible. So, if you had any illusions that we would soon be piercing the veil of our ancient Homo Floppicus ways to be reborn as Homo Fibericus, forget about it. About the closest we’re going to get in the next 10 years is to carry around little, network enabled versions of our desktops. In reality, we’re already doing that, but don’t expect that to change much any time soon.
![[del.icio.us]](http://images.del.icio.us/static/img/delicious.small.gif){kind=small}
Posts
Tyler, while I agree there is a bit of runway before the Cloud takes off, it is much closer than you think and certainly sooner than 10 years. First off, what you have mainly articulated is what is referred to as a “Public Cloud”. Obvious examples are Google as you mentioned, Amazon’s EC2, and many other Service Providers are in the midst of creating similar services based on a traditional utility model. Another form of Cloud Computing taking shape rapidly is referred to as a ” Private Cloud”, whereby a utility model is used to deliver applications within a given business and not by an external provider.
As businesses look for new ways to reduce their operational expenses, both of the aforementioned forms of Cloud Computing are becoming more compelling. We’re not talking about another failing business model of the late 90’s, but one which has a solid business foundation, and proven technology to assure it happens. Virtualization software delivered by VMware, Microsoft, Zen and others coupled with new powerful multi-core processor technology such as Nehalem and Istanbul will enable Enterprise and Mid Market businesses to radically optimize their efficiency.
Cloud services will reduce data center footprints and provide a level of business agility most companies have never realized. Company supply chains will reach their full potential, because they’ll have a level of dynamism, enabling them to move work loads immediately. Listen, desktops are not going anywhere, but the hypervisor or the virtualization abstraction layer allows you to choose what operating system you want. If you can get that in a utility form factor with the right amount of storage, compute, and memory at the right price, people are going to have at it. Will this exist for all desktop applications? No, not immediately, but every major ERP vendor is adjusting their business models.
So will Cloud Computing, both public and private be widely adopted in the near future, maybe not, but suffice it to say, it is a matter of time, and I believe that will be in less than 5 years.