I recently got an e-mail from a reader that asked, "Since when has the Internet become the cloud?" and it struck me how far from the truth the term "cloud" is (and I am talking about computing, of course). After all, there's nothing up in the air about this. Quite the contrary.
When facing something cool and hard to explain, we tend to give it a sexy name that might not have much to do with what it is. This is because we like it that way and because sexy sells. And Steve Jobs' recentonly proved that matters can get even worse, or better, depending on what side you're looking at it from.
So what's cloud computing?
Much as Jobs would probably love to believe it (and might even have already filed for a patent), he isn't the one who invented the term "cloud." He's not even the first to publicly use it. If memory serves, the first high-profile usage of the term "cloud computing" must have been Google's Eric Schmidt back in 2006 during a search engine strategies conference. The then-CEO of the Internet giant used the term to describe the company's approach to SaaS, or software as a service.
When asked about Google's development of Internet-based services, Schmidt said, "It starts with the premise that the data services and architecture should be on servers. We call it cloud computing--they should be in a 'cloud' somewhere..."
There you go, that's an abridged definition of cloud computing. Basically, the term refers to the use of computing resources over an IP network, instead of locally. For example, when you're on a plane and plug in your external hard drive to access information that your laptop hard drive doesn't hold, that's not a cloud service. On the other hand, when you sit in a cafe and access your Google Docs via the Wi-Fi network, you're using a cloud service. In this case, both the word-processing software and the data (the documents themselves) reside not locally but elsewhere.
2011 Apple WWDC roundup
QNAP TS-412 Turbo NAS server
First Take: Apple's iCloud
Examples of services provided via the cloud that take the place of functions that were originally handled locally are Google Docs, Amazon S3 online backup, and online digital streaming, which has taken the place of DVDs.
Physically, a cloud service requires a storage facility that hosts the information (for more on storage,) and Internet-based applications. Generally the storage facility is part of one or many data centers scattered around the globe. (Google, for example, by 2008 had about 40 of them all over the world.)
In other words, despite the lofty name, the backbone of a cloud service is definitely down-to-earth. The use of the word "cloud" is probably intended to hint that you, the user, don't really need to care about where the data is stored, just about the front-end application that's presented to you on the local device.
Soon after Schmidt used the term cloud computing, Amazon.com started promoting its online data services, Elastic Compute Cloud (EC2) and Simple Storage Service (S3), using the same language. The terms "cloud" caught on and has become mainstream and trendy since. Other storage companies realized that they could benefit from it and started using it to describe their existing local storage devices that supported remote data access over the Internet.
Examples of products with "personal cloud" features are the
Up until now, a cloud service, be it one hosted by a big company (Amazon, Google) or on a personal NAS device, generally allows you to use it on an opt-in basis; the features and the service are there on your device if and when you want to use them. In other words, you have control over them.
Apple's new iCloud, however, as personal as the name might sound, judging from how Jobs describes it, is going to be a very different beast.
Some things you should probably care about.
When watching Jobs introduce the iCloud service, the moment the he said, "It just works," I was both impressed and scared. Impressed because of the excellent and impeccable timing of his words and the changes of the presentation slides on the big screen. You can't create a better marketing effect than that. And you know by the applause that the audience members were sold. How could they not be?
The scary part, however, is the potential of what's going to happen if it actually works and, even worse, becomes popular. The profit for Apple and smile on Jobs' face aside, the iCloud will likely also create at least two troublesome side effects.
First is the excessive use of data connections. If the popularity of the iPhone has contributed to making AT&T's 3G network close to useless in many cities, to the point that many of us have to turn off 3G to hold a steady conversation, what will happen given that the iCloud will likely use much more bandwidth than the iPhone? Pushing digital content to multiple devices at the same time is much more bandwidth-intensive than playing games or downloading maps and e-mails on a single device.
Currently, a user first downloads songs, videos, podcasts, or photos onto a computer, then syncs it with other mobile devices. This is a little inconvenient but at least you just need to download the content once. The iCloud service, on the other hand, according to Steve, automatically "pushes" the content to all your iCloud-connected devices using the Internet. This means the total amount of data that needs to be downloaded would be the original amount multiplied by the number of devices you have. Now multiply this similar activity by millions of users, and we face a huge bandwidth problem and put a lot of stress on an already overstressed data infrastructure. All that just so you can save a few minutes by not plugging your mobile devices into the computer.
Many Internet providers now put a cap on the amount of data you can download per month. Comcast, for example, allows only 250GB, which is actually not much in my personal experience (though I am not a typical example as I have to test storage devices, which involves moving lots of data around). Nonetheless, having multiple iCloud-compatible devices would help you use up the data allowance faster.
The second side effect is the loss of control. As iCloud is integrated into apps and devices, that might mean users will not have control over it, and may even not be able to opt out. If you store your purchased digital content in Apple's iCloud storage space, that could mean the company has control over what you can view. It's unclear if you can even upload your own content, ripped music from your CDs or music purchased from other services for example, and store it in Apple's iCloud. Judging from the way iTunes syncs contents with devices, it's a safe guess that the iCloud service will offer users much less control over it than other existing cloud services.
On top of this, there's also a security risk factor that's present in all cloud services. After all, it's just another online data service that keeps the private information of millions of people.
Obviously, we'll have to wait to see how iCloud is implemented. However, judging from what Jobs said and what he didn't, the iCloud service does seem to have cloudy parts. For those hard-core fans of Apple and the iCloud who are superexcited, sorry if I have just rained on your parade. But face it: if you like the cloud so much, you should be ready to get wet once in a while.