blockr

Blockr

Block distracting sites from the command line

Features

I built this as a replacement for SelfControl. It has the following features:

• Cross platform (MacOS and Linux).
• Configurable with a human-readable and editable JSON file instead of plists.
• Easily unblock a single site (sometimes a site like Reddit, Twitter or Medium does have a useful article you want to read at work, but you don't want to unblock everything).
• Command line interface instead of GUI.
• Runs indefinitely until you turn it off, rather than having to re-run it every 24 hours.
• Offers an easier way to authenticate as root, which is needed for editing the hosts file.

Install

yarn global add blockr

Usage

Start by creating the file ~/blockr.json. You can use the file in this repo as a starting point. It should contain a single hosts key containing the URLs of sites you would like to block. This file can use JSON5 syntax, including comments and trailing commas.

Activate

Run blockr to block all of the sites in your config file.

Deactivate

Run blockr unblock to remove the block. Run blockr unblock [sitename] to unblock a particular site.

Authenticating

Blockr works by editing the hosts file at /etc/hosts. This requires root access. You can acheive this by running sudo blockr, however this will require you to type your password every time you use it.

You can also provide your sudo password via a flag, like blockr --password hunter2. If you're worried about security, use a command line password manager like 1Password to echo the password. Using a literal is not recommended as your password will then be saved your shell history.

Syncing config file

You can easily sync your config file in two ways: via a symlink, or a flag. First create your blockr.json somewhere else like ~/code/dotfiles/blockr.json. Then you can point to it

Via a symlink: run ln -s $HOME/code/dotfiles/blockr.json $HOME/blockr.json

Via a flag: use blockr --config-file $HOME/code/dotfiles/blockr.json. You could also put this in an alias so you don't have to type it every time.

Planned features

• Temporarily unblocking a site for 10 minutes or something.
• Require a one minute delay and a re-confirmation before unblocking.
• Configure schedules when the block should be applied.
• Show stats on how many times you've unblocked each site.
• Run a background process to put back the block if the user tries to delete it by editing /etc/hosts directly, like SelfControl does.
• Strict mode that prevents you from running blockr unblock for a certain period of time.
• Groups of sites to block eg "social" or "news"
• Support multiple URLs for one site, eg. bbc.co.uk and bbc.com

License

MIT