When Programs Get Complicated
My problem with this is that once programs get complicated, as they often do outside of the Unix command line, tracking all these privileges gets complicated. Users want their programs to do useful things like send e-mails and save to a flash drive. Expecting the "IT support" at the school to decide how to handle errors is not a reasonable thing to do. There are other complicated decisions, such as default numbers for rate limitation on network access, that are left to some authority I dont assume will be present. Bitfrost doesnt specifically address the issue, but I think it expects a lot of someone at the school administering systems. As we know from the sad story of Julie Amero, who was prosecuted because her class PC had become infected with a porn popup-spamming program, even in the filthy-rich United States of America school IT may not get the job done on security.And the basic issue of social engineering is not specifically addressed, except to the extent that technical measures in Bitfrost may prevent operations the user may be prompted to do by an attacker through a social engineering attack. But Im sure that still leaves some leeway for dangerous operations, and the possibility that the permissions administrator can be fooled into allowing the installation of a dangerous program. Its really not that hard to envision a school of OLPC notebooks becoming infested with bots, if a resourceful attacker tried to get them in there. And there are people at OLPC who envision such attacks. If enough of the notebooks are out there, they automatically become a target just because of their numbers. The standard configuration and relative naiveté of the user base make them more tempting targets. Other people at OLPC recommend that the activation and central backup projects be postponed, if necessary, in order to get more effective prevention in place against social engineering attacks. As highly as I value education, including computer education, I think there are better ways to aid the desperately poor parts of the world than to send them cheap computers. I think OLPC will fail because even such countries will see their money better spent in other ways. But I wouldnt be surprised to see some useful experimentation coming out of the project. Even if Im right about mistakes made the first time around, I suspect theyre not afraid to learn from them and fix them, and thats an effort worth supporting. Security Center Editor Larry Seltzer has worked in and written about the computer industry since 1983. Check out eWEEK.coms for the latest security news, reviews and analysis. And for insights on security coverage around the Web, take a look at eWEEK.com Security Center Editor Larry Seltzers blog Cheap Hack
More from Larry Seltzer
Check out eWEEK.coms for the latest news, views and analysis of technologys impact on government and politics.