Head in the clouds

In ancient times beginning around the 1950s, the dominant paradigm for computer systems involved a mainframe that users accessed through one or several dumb terminals. Personal computers did not exist; the massive size and great cost of the central processing units typically required very substantial investments by large companies or departments of government.

Computing power came from another room or another building. Generally the people who had data to manipulate did not own or even have direct access to the programs that processed the input. Computers were relatively rare, and society did not depend on them for much, if anything. A network of computers that linked millions of people was an idea that only a few dreamers dared to entertain.

When personal computers arrived on desktops at home and the office, so did programs and processing power: they were all in the box, and a single user could input data, manage the processing, and obtain the output all in the same place--and rather quickly too. That user could retain possession of the data from the beginning, through the manipulation, and after the results emerged. While computers were removed from any network, the data on which they worked could remain private. Yet even before networking became widespread, the users of computers became increasingly dependent on these machines.

Clouds have appeared in recent years and with them a host of advantages. With them, too, there are some problems. Cloud computing offers the freedom of access to one's data anywhere there is a connection with the Internet. Furthermore, it offers freedom from responsibility for upgrading programs: clients may be able to access the latest versions of programs on the date of their release. And backups of data can be remote: local dangers--fires, floods, forced entry into physical facilities--do not threaten to destroy information stored in a cloud.

But while a cloud can yield many benefits, it also has its perils. When disconnected from the Net, users cannot access programs that exist in the cloud but not on local machines: work can grind to a halt. Disconnection can also prevent access to much needed information.

Additionally, there are some threats to privacy. Providers of online backups guarantee the security of the information they handle, but guarantees depend on the willingness of all involved to honor them. Even where there is a guarantee, there is still a need for trust--a dependence on another party to honor a commitment. Unfortunately, there can be no absolute certainty that breaches of privacy will not occur.

Data stored on remote servers can still be vulnerable to a hacker's attack. Vandals with no other ambition than destruction of information can ruin the work of a lifetime, if the information is not backed up somewhere else.

Moreover, the enforcement of certain laws, while hampering subversive communications and thwarting criminal plans, may also result in the loss of information managed by innocent parties. An instance came to light recently. Law enforcement officers contacted BurstNET, a web hosting company, on July 9, 2010 to learn who owned the server that hosted blogetery.com. The object of concern was some information posted to a blog--information allegedly supporting terrorist activities. BurstNET investigated the matter and determined that the blog did contain such prohibited material and disabled the platform. However the outcome was not only the appropriate suppression of forbidden communications but also the loss of harmless data for thousands of innocent bloggers.

These days, computers are more often part of a network. Presently a major reason, if not the only one, for owning a computer is getting connected with the Internet. More and more often, information is being shared. And online communications, being social interactions, are subject to all kinds of scrutiny, and they can elicit any or all of the responses that human communications in any other realm can produce. Online communications are also objects that are sometimes vulnerable to damage and sometimes in danger of being completely lost.

At least two conclusions come tumbling out of all this: One is the obvious recommendation to respect and uphold the law online. Computer networks do not exist apart from the rest of society, and they are playing a much larger role in it. Another is the recommendation to have an effective strategy for preserving data. Perhaps the greatest mistake in the Information Age is the failure to backup one's files. Even those who have committed no crimes can lose their information. Nothing in our worlds--real or virtual--is imperishable: clouds too can evaporate.