Skip to content

Mastodon/Twitter bot that posts videos showcasing how random locations in the world have changed since 1984.

License

Notifications You must be signed in to change notification settings

doersino/earthacrosstime

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

63 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

earthacrosstime

Mastodon/Twitter bot that posts videos showcasing how random locations in the world have changed since 1984.

In a bit more detail, whenever the bot runs, it...

  • loads a shapefile from disk,
  • generates a random point within the bounds of this shape,
  • figures out which video tile needs to be downloaded to the point and an area around it,
  • downloads that video from the repository underlying Google Earth Timelapse,
  • reverse geocodes the chosen point using Nominatim to figure out the location's name,
  • edits the video, annotating it with latitude & longitude, area covered, and a named pin on a world map,
  • saves that to disk,
  • and tweets and/or toots the edited video, optionally with a geotag.

Much of the code has been adapted from ærialbot, a previous project of mine that basically does the same (and more!) for static maps, and CMU's Time Machine Viewer, which contains reference implementations of the required coordinate projections.

🐦 Check it out at @[email protected], where every sixth post or so is made by this bot!

Here's one of the videos posted by this bot, showing the construction of Incheon Airport and various developments on land.

earthacrosstime-example-incheon-airport-construction.mp4

Features

Here's why this bot is a Good Bot:

  • Configurability: Take a look at config.sample.ini – you can supply your own shapefile (or instead define a fixed point), control output verbosity, set a different Time Machine repository, scale the result videos to your preferred size, define the text of the tweet, and more!
  • Correctness: Because neighboring meridians are closer at the poles than at the equator, uniformly sampling the allowable range of latitudes would bias the generated random points toward the poles. Instead, this bot makes sure they are distributed with regard to surface area.
  • Automatic zoom level selection: Simply supply a maximum allowable number of meters per pixel – the code will then take care of dialing in a (more than) sufficient zoom level.
  • Comes with batteries included: The assets/world-shapefile/ directory contains a shapefile providing the outlines of the continents and most islands. More shapefiles, along with a guide on how to convert third-party shapefiles to the correct format, can be found here.
  • Cares about typography: The text that's superimposed onto the result videos is aligned with utmost precision. Just in case you were wondering.
  • Geotagging: Tweets will be geotagged with the exact location – you can disable this, of course.
  • Logging: Keeps a log file – whether that's for debugging or reminiscing is your call. Again, you can disable this easily.

Usage

Setup

Being a good Python 3 citizen, this program integrates with venv or similar packages to avoid dependency hell. Run the following commands to get it installed on your system:

$ git clone https://github.com/doersino/earthacrosstime
$ python3 -m venv earthacrosstime
$ cd earthacrosstime
$ source bin/activate
$ pip3 install -r requirements.txt

(To deactivate the virtual environment, run deactivate.)

One of the dependencies, Shapely, requires the GEOS library for performing operations on two-dimensional vector geometries, which you may need to install first as described here.

Configuration

Copy config.sample.ini to config.ini, open it and modify it based on the (admittedly wordy) instructions in the comments.

See here for advice regarding finding shapefiles of the region you're interested in and preparing them for use with ærialbot.

Running

Once you've set everything up and configured it to your liking, navigate to the directory where earthacrosstime.py is located (this is important – the bot won't be able to find some required assets otherwise) and run it:

$ python3 earthacrosstime.py

That's basically it!

If you want your bot to post at predefined intervals, use cron, runwhen or a similar tool. To make cron work with venv, you'll need to use bash and execute the activate script before running the bot (in this example, it runs every eight hours at 50 past the hour):

50 */8 * * * /usr/bin/env bash -c 'cd /PATH/TO/earthacrosstime && source bin/activate && python3 earthacrosstime.py'

Pro tip: If you want to host multiple instances of this bot, you don't need multiple copies of the code – multiple config files suffice: simply run python3 earthacrosstime.py one-of-your-config-files.ini.

Uber pro tip: Run python3 earthacrosstime.py --help to learn about some secret CLI options!

License

You may use this repository's contents under the terms of the MIT License, see LICENSE.

However, the subdirectory assets/ contains some third-party software and data with their own licenses:

About

Mastodon/Twitter bot that posts videos showcasing how random locations in the world have changed since 1984.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages