(The power to understand your voting system.)
Sede (this program) subscribes to the Unix philosophy. The votes - the center of the program - consist of plain human readable text (ASCII). You (the user, the vote-administration) can configure the "magic patterns" to find the `vote-areas' in the returned plain-text data. Then your patterns cut these `vote-areas' to reveal the vote, the vote-code, and the comment (using sed(1) -e 's/...'). With this in hand, counting votes is as simple as running the returned data through uniq(1) --count (a little Unix program that removes unique lines from the input; with "--count" as an argument, it tells you how many times a certain line has occurred).
The vote-code is generated from random-data (/dev/[u]random), and stored both in a human readable/editable plain-text file (database, if you will), and on the ballots (in the aforementioned vote-area, or wherever you think you need it).
Voters can run sede on the data in the results, and generate a similar result as the original vote-administration. Almost the entire poll configuration can be downloaded.
Unix philosophy - Wikipedia, the free encyclopedia
It is nothing more complicated, then someone giving you a form where it says `vote here: ...' and `comment there: ...'. Those giving you the form then know, that to look for your vote, they should look under the entry at 'vote here: ...'. Then, these forms are cut with a scizzors to reveal only what voters have a written after `vote here: ...', and these little pieces of paper are counted. Only those that are exactly the same, are added together. Computers can do this fast and efficient.
What above is called sed and uniq, are Unix tools that can be used as "scizzors" and "counters". With "pattern" is meant `vote here: ...'.
You get a secret-code with your ballot form (one for each vote). It is used to make sure nobody is sending back any made-up ballot forms, and so you can find your vote again in the end-results.
next intro (...)
previous intro (unlimited)
back to title page