How much should users be able to customize embedded products they buy? A whole lot, says one reader.
I found Sean Lamont's letter (``Seal It, Don't Hack It'', in the January/February 2002 issue of ELJ) unfortunate. Lamont treats the ability to study or modify a device as a pure business issue, a decision to be made once and for all by a manufacturer. But when customers buy devices, they own them--the manufacturer has given up possession.
While a mismatch of consumers' interests with suppliers' business models is unfortunate, a vendor has no right to insist that consumers use their property only in approved ways or refrain from examining it.
In my view, reverse engineers are heroes of technology, responsible for considerable innovation (sometimes, admittedly, innovation that is inconvenient for a particular manufacturer). They also can act as the ultimate consumer advocates, helping customers get the best possible value out of what they buy.
Every branch of technology has relied heavily on reverse engineering, and so have most major electronics manufacturers. Natural science itself reverse-engineers nature to supplement its documentation and to allow us to add new features.
No embedded operating system provides more ``protection'' than any other against the efforts of a device's owner to study how it works. Each operating system has its own set of emulators and debugging tools. Some designs add additional layers of obfuscation.
However, there's no reason to think that obfuscation will frustrate analysis over the long run. See Barak et al., ``On the (Im)possibility of Obfuscating Computer Programs'' (2001).
Embedded Linux Journal supports the right of customers to learn how products work, modify them and install new software. We do not support redistributing software copied from an embedded Linux device in violation of copyright law, and we don't think vendors should have to support modified products.