Thursday, June 17, 2004

More e-voting nonesense

Dan Gillmor, my friend and neighbor here in the Silicon Valley, writes in today's MercuryNews that California's Secretary of state "finally sees the e-voting light." (if that link is blocked by a registration form, try here.) But as I told him thru his weblog:

Another load of bullfeathers, Dan. But I'll only pick out two points - source code and paper trails.

Source code: Holding some source code in your hand tells you nothing about how the DRE terminal will perform. Only a well designed validation test on the DRE terminal itself will do that. After all, who's to say that the source code in your hand is the same as that compiled in the machine? Or that the compiler itself doesn't introduce a flaw?

Paper trail: As I've said before, anyone smart enough to manipulate the votes on a DRE terminal would find it child's play to manipulate the print out. Only well designed validation of that particular machine is useful in preventing incorrectly reported ballots whether mechanical error, electronic failure or willful manipulation.

While there's no reason not to also provide source code and paper trail, they are not, in and of themselves, any guarantee of accuracy whatsoever.

Validating each and every voting terminal is a lot harder than waving around a few pieces of paper, admittedly. But getting accuracy in balloting is well worth the effort. Not doing so is the real first step to risking our democracy.

Comments: Post a Comment

© 2003-2006 The Virtual Quill, All Rights Reserved


[Powered by Blogger]