What Kind of Cloud Do We Want?

IMG_2956

(T) Thirty years ago, Ken Olsen, founder, and CEO of Digital Equipment, never believed in the PC. His argument was that “why someone would use a PC when much more computing power is available to you if you just connect a VT 100 terminal to a VAX”. What Mr. Oslen, although an outstanding engineer and entrepreneur, missed was that for users “personal” computing was more important than “powerful” computing. And over the years because of Moore’s law and better software engineering, Mac and PCs have become very persuasive in our daily personal and work life.

Eleven years ago, I was sharing a taxi going to a meeting in Chicago with a sale engineer, and that was the first time, as far as I remembered, of someone going to a meeting without his notebook but just with his BlackBerry. Over the last ten years, we have moved from “desktop” computing to “mobile” computing simply because “mobile” computing has just become more convenient than “powerful” computing.

I can only find two roots in cloud computing…

The first one is the concept of “thin client” that Sun Microsystems pioneered with the JavaStation. The major goal of the JavaStation was to lower the cost of operating clients in the enterprise by substituting PCs with JavaStations (whose OSs were a Java OS and had a Java browser). Google’s Chromebooks are following the same fundamental concepts of the JavaStation except that Chromebooks are designed to operate not with an enterprise server but with cloud services.  And the simple reason for that is that over the last 15 years most of the user traffic has moved from local networks to the Internet.

The second root of cloud computing that I can point to is Hotmail. Hotmail was the first web-based e-mail service and was very successful because it enabled users to check their personal e-mails at work, and at early Internet cafes. Over time, web-based services have exponentially expanded to consumers with iTunes, Facebook, YouTube, LinkedIn, GoogleMaps, Skype, Twitter and so on.

Today cloud computing services considered that the client is the browser. Furthermore, the browser can have the same fabric as its underlying operating system as in the case of Sun’s JavaStation or becomes the operating system itself as in the case of Google’s Chrome.

Do we want all our applications to have the browser as the only client?

Probably not. This is the problem that many of us have run on many occasions. There are some applications that can be designed for the browser, or more exactly for some applications, the browser can provide a sufficient user experience but there are many applications especially rich user interfaces or computing intensive I/O applications for which the browser does a poor job. A typical example is an enterprise network management application that might require to display some complex system configuration or network topology maps.

Do we want all our applications to be cloud-based?

Probably not. Even if the application can be done within the browser, I feel sometimes that I do not want the browser to do it for me. For instance bookmarking. I can use Delicious to have always my bookmarks available across my BlackBerry, Playbook, and Mac. Unfortunately, Delicious breaks the organization of the bookmarks that I have with my browser. Therefore, I have to export my bookmarks and import them across my devices: smartphone, tablet, and notebook through a USB – which is nonsense! Ideally, we want the cloud to keep not only our data and back-ups updated but also the state of a work in progress that we might do over multiple mobile computing devices – and with those devices running native device applications that have some benefits over web-based applications. Those benefits can be features or better performance. Apple is starting to head into that path with iCloud. iTunes, by leveraging iCloud, keeps the songs that you purchased across Apple clients and enables to download those songs across any Apple device.

Do we want all our data in the cloud?

Last May about 200,000 Citibank credit card customers in North America have had their names, account numbers and email addresses stolen by hackers. Earlier this June, Google announced that the personal Gmail accounts of several hundred people, including senior U.S. government officials, had been breached. Last April, Sony’s PlayStation Network was shut down after a massive security breach that affected more than 100 million online accounts! Also in April, hackers penetrated a network operated by data marketing firm Epsilon that handles email communications for JP Morgan, McKinsey, Best Buy, Target, Walgreens, Marriot. And, the list goes on…

 Even RSA, a leading security firm, had a security breach earlier this year that disrupted the use of their security tokens to authenticate access to corporate networks for many of their customers.

Until Citibank, Google, Sony, RSA start encrypting their customer data in their clouds, the best way to guarantee that if data is stolen, it will not be accessed, do you want to have all your data in the cloud? Probably not.

And, if you have sensitive data in the cloud, make sure that you did encrypt them yourself before giving it to the cloud provider.

Note: the picture above is from Point Reyes south of San Francisco.

Copyright © 2005-2011 by Serge-Paul Carrasco. All rights reserved.
Contact Us: asvinsider at gmail dot com.

Categories: Back-End, Cloud