I'm reminded that political polling station systems are kinda like telephone systems. They didn't involve computing solutions at one time... and now they do. And when more complexity is introduced into an activity, more possibility for everything bad is also introduced.
I have a preliminary suggestion here for smoothing some of the issues being brought up by all sides regarding the recent US elections. EVEN ASSUMING NO IMPROPRIETY with any participating individual or group, the following problems have arisen or have threatened to arise with current electronic election systems:
server capacity and load issues
web incompatibilty issues like untested applets
web exposure issues like DOS or the opportunity for such
location staffing problems (some admins not allowed to fix problems)
wireless issues
supply problems
access code failures
weather issues
... and the hits keep right on coming. I contend that many of these problems could be minimized by imposing a CLOSED SYSTEM REQUIREMENT at the state level. No public internet or wireless. Modem communication over data phone lines. Not necessarily going back to 100 percent paper ballots but using current equipment in a less complex manner. I think this option would address many items both of outside interference and internal architectural problems. It would use older, more reliable and less overburdened routes of transmission and offer a smaller attack surface to miscreants.
Sure, this suggestion isn't extensively researched by me yet. Sure, it doesn't address the issue of the behavior of some county official somewhere. Sure, it isn't something that some state has 100% already done, such that I can point to it as an example. But this should get the party started among readers of this blog as to a discussion of where to go next.
Assuming the powers that be ask the IT people what to do. I offer the suggestion due to its simplicity: it involves not hiring a bunch of new people or buying a lot of new equipment. It would retrench by reverting to earlier solutions and curtailing the use of failed ones. It would possibly move us forward not so much due to what would be done, but what would NOT be done.
