Also: Defending telecommuting; serving up Java
OLPC unrealistic
Regarding Mark Gibbs’ BackSpin column, “OLPC: If you won’t contribute, step aside”: The real issue is whether we will be improving anyone’s life by providing a tool they may not need. I know it is an unpopular belief, but the U.S. is headed for trouble as we become a service and information society. Eventually, there will be an economic collapse, or hopefully just a more balanced distribution of the wealth, if we don’t begin to manufacture something we can export and make a profit on. Developing countries have an advantage without all the government oversight we have as they can do the manufacturing work we used to do. Providing them a tool that does not enhance their ability to take advantage of their export potential is not a help. So having said all that, what good is information or connections if you can’t use them in your society and more fundamentally, why do we need all this information for the masses anyway? I just see everyone around me getting dumber, fatter and lazier without any good basic work skills. And this is the good we want to share with the world? Or is it just a great plan on our part to make the rest of the world like us?
Ray Windisch
White Hall, Md.
Ever since man started questioning his environment — looking through a telescope to gaze at the heavens, examining objects under a microscope — he’s had to struggle with idealism vs. reality. It would be great to stamp out poverty, hunger, sickness and illiteracy. I don’t even think Americans would mind a slight tax increase if they knew these altruistic intentions really could be manifested. However, I’m reminded of the lesson I learned as a youth: “Give a man a fish; you have fed him for a day. Teach a man to fish; and you have fed him for a lifetime.” To me, passing out laptops to people who probably don’t even own fishing poles is fundamentally flawed. Next, we’ll be sending them wide-screen TVs.
Scott Mullins
Nashville, Tenn.
Regarding Mark Gibbs’ BackSpin column, “OLPC: If you won’t contribute, step aside”: It is completely unrealistic to expect millions of extremely poor people to be able to take advantage of networked computers, even if they are free. Even here in the U.S., the mainstream population does not adopt technology all that quickly. It’s one thing to go from a desktop to a notebook to a handheld, but if you’ve never even seen any kind of computer, a laptop will seem like a tool from another planet.
Much of the value of the Internet is derived from local content to which it can provide access. There is no infrastructure in place in the targeted countries that can make the Internet valuable for even 10% of the intended users. If there were no U.S. Web sites, how much would you use the Internet?
Then there’s the support issue. Who is going to tell people how to use these things? Worse, if they cost only $100, who will fix them? (I hope they’re not disposable; we already send so much electronic waste to Third World countries.) I’m not a Luddite. I love gadgets and technology, and I do believe technology can solve many problems — but not by itself. It’s a tool, and a very sophisticated one, no matter how spectacular the user interface. I just don’t think millions of dirt-poor people are going to make good use of these laptops.
I’d rather see these poor countries get technology that helps them grow food, make their own energy and build local economies. The computers, like any other tools they need, will follow.
Ken Goldsholl
Santa Barbara, Calif.
I disagree with Mark Gibbs’ stance on OLPC. The digital divide is nonsense. Will Googling “food” dampen hunger? Will Wi-Fi combat institutionalized corruption? Will instant messaging stop a bullet? Will e-mail clean up the raw sewage running past the front door? Better the money was spent on malaria — something we have known for generations how to control and combat, yet millions die from it every year — or health care, housing and sanitation. OLPC makes for good press and warm fuzzies, but it helps no one.
Vern Mastel
Technology coordinator
Veterans Memorial Public Library
Bismarck, N.D.
Defending telecommuting
Regarding “Advocates defend telecommuting against backlash”: This has nothing to do with teleworking at all. The root of this story is all about data integrity. What possible excuse do these corporations have for allowing what should be protected data to leave secured space?
Time and again I see businesses allow developers to use live data. This is uncalled for. There should be extreme accountability for personal information stores. Only when penalties soar into the millions and include jail time will any changes be made. There is just no excuse for live customer records to ever be carried around on a laptop — none.
To make it even more ridiculous, none of the data was encrypted in these incidents — it’s unforgivable. Particularly since it is trivial to encrypt such files.
Chris Esser
Middletown, N.J.
One simple thing that would help all these cases is making sure that there was encryption on the laptops and teaching the users how to use it. There are commercial offerings as well as very good open source and free offerings that make this information secure. I am not talking about encryption of your e-mails — you should be doing this anyway — but rather encryption of the entire hard disk or particular volumes/folders. The hardest thing to do is get users trained so that they buy into the technology. That is the battle.
David Butler
Nottingham, U.K.
Serving up Java
Your recent overview of server technology incorrectly states that many Java applications already provide support for multithreaded code and therefore will benefit from dual-core or multi-processor servers. Unfortunately, it is not that simple. The operating system and corresponding JVM must be designed to support multithreading along with the application code. In most situations, the first processor or core must reach 100% utilization before the workload is shifted to a second processor. This is why there is usually only a modest performance gain with multiple processor systems. The OS may use the second processor or core to run other applications, which may help overall system performance, but that is not the same as having an application’s code running efficiently on multiple processors and showing a scaling of performance. As has been the case for much of the history of computing, the hardware is years ahead of software able to provide for its full utilization. Even your article states that much of the X86 platform’s benefits will not be realized until Longhorn ships in 2008 (or more realistically in 2010), almost a decade after the introduction of the first multi-core chips. The Microsoft monopoly has cost dearly this industry and all the businesses that utilize its products.
Bruce Stenman
President
Lightsmith
Prunedale, Calif.
