Make Sure Donald Trump Can’t Tamper with Public Data

I’m written a small, hopefully useful, tool to allow for decentralized monitoring of public data (coding against Donald Trump). It’s called minitrue, and you can get it here.

What is it?

Python scripts to monitor public data and prove that it’s been changed.

The idea: suppose Donald Trump goes full 1984 and tampers with government data. For example, modifying climate data to conceal evidence of human generated climate change. That’s bad. Wouldn’t you like to catch him? minitrue can help.

minitrue will take a list of URLS that you give it. Then it will go to those URLs, save their contents to disk, and tweet out a (sha256) hash of the contents.

Then suppose Trump (or some evil corporation, or whomever’s data you want to monitor) changes it. The next time minitrue runs (you should set it to run every day), it’ll see that the hash has changed, and it will tweet out the fact of the change.

Because you posted the original hash on Twitter, and Twitter maintains a record of dates posted and doesn’t allow editing, you can reasonably well prove that the file existed in its original form on the date you originally posted it. Take that evidence, plus the two files, and go wild — -media, lawsuits, whatever your little citizen-of-a-goddamn-democracy-no-matter-what-Donald-Trump-thinks heart desires.

The second nasty trick: the tweets never contain the name or description of the document being monitored. (A unique id is assigned to make it easier to correlate them later.) So nobody can tell just from reading your Twitter which documents you are monitoring.

The more people who use it, the more threatening it is: the ultimate idea is to deter document tampering by making it impossible to know which documents are being watched — -to create a digital panopticon to keep Donald Trump/whomever else in line.

Very much a work in progress, but it’s ready to go right now in alpha. Please download, test, and make pull requests. See the readme in the github repo for more.