A problem with self-destructing email is that the benficiary doesn't bear the burden, which is sociologically essential if the technique is to succeed. For example, it takes effort to set parameters on an email indicating when it expires, and this setting (the burden) must be done by the sender, while the beneficiary is the recipient.
Usually the security of the link (e.g. 802.11) isn't important unless you need to prevent denial of service attacks (perhaps that's a possibility where you work:-). Instead, you can rely on the security mechanisms that IP (and Mobile IP) provides.
The beauty of IP is that being software & readily available, it has been able to spread over all sorts of networks, providing connectivity, and consequent demand for features (such as security) that provide functionality irrespective of what functionality the lower layers of network hardware etc provide.
The main difference between CompEng and CompSci, IMHO, is the attitudes towards abstraction and complexity. In CompEng, the emphasis is on concrete things (e.g. hardware gates, device physics etc), whereas in CompSci, the emphasis is on software, which leads to a tendency for more abstraction and virtualization.
I had an engineering-based education, and now work in a software-based field. Usually when I read a technical paper, I can tell from the level of abstraction/virtualization whether the author comes from a CompEng or CompSci background.
Slashdot Mirror
A problem with self-destructing email is that the benficiary doesn't bear the burden, which is sociologically essential if the technique is to succeed. For example, it takes effort to set parameters on an email indicating when it expires, and this setting (the burden) must be done by the sender, while the beneficiary is the recipient.
Another alternative fiber structure was described in an article in The Economist in March, referring to work by a Danish company.
The beauty of IP is that being software & readily available, it has been able to spread over all sorts of networks, providing connectivity, and consequent demand for features (such as security) that provide functionality irrespective of what functionality the lower layers of network hardware etc provide.
I had an engineering-based education, and now work in a software-based field. Usually when I read a technical paper, I can tell from the level of abstraction/virtualization whether the author comes from a CompEng or CompSci background.
There are some photos of MEMS mirrors at CalTech and Bell Labs, and page 90 of the January Scientific American.
Given the CueCat logo's resemblance to an unhappy smilie :(, perhaps they foresaw disappointment regarding privacy.