A goddamn simple RSS reader, to replace Google Reader.
PHP Ruby CSS JavaScript HTML
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
config
logs
ruby
www-includes
www
.gitignore
LICENSE
README.md
UPGRADES.md

README.md

FUCK READER

A very simple RSS reader. I built it just before Google Reader died, and I've been using this every day on my own server. A couple of my friends use it, too. You can read more about it here.

Front-end is PHP, feed fetching is Ruby. Feeds, posts, and user post state all stored in MySQL or MariaDB.

Front-end installation

This was built with lighttpd and php5-cgi (running via fastcgi). It's also been tested with php5-fpm. Set the lighttpd document root to the www folder.

Also, include this line in your lighttpd config:

include "/path/to/fuck-reader/config/fuckreader-lighty.conf"

Obviously replacing the path to the file with the actual path on your server. That lighty conf file includes the routing for fuck reader. Make sure you edit the first line of it accordingly, since you won't be using fuckreader.com.

Edit www-includes/dbconn_mysql.example.php with the necessary settings and rename it to dbconn_mysql.php when you're done.

If you want to use Google Analytics, edit config/google_analytics.example.php and rename it to google_analytics.php when you're done. If you do not want to use Google Analytics, then comment out the last line in footer.php which mentions it.

I used MySQL, and then later MariaDB, and included in config/fuckreader.sql are the table schemas used by FUCK READER. Just make a database called "fuckreader" and import that using phpMyAdmin or something.

Feed Fetching Prereqs

Feed fetching uses the "feedjira" library from the folks at feedbin.

On Debian-based servers (Debian, Ubuntu), you need these prereqs:

apt-get install ruby1.9.1 ruby1.9.1-dev build-essential libcurl4-openssl-dev libxml2-dev libxslt-dev

Install rubygems to a specific version. At the time of this writing, rubygems 2.0+ won't install feedjira. If gem is already 1.8.x, ignore this.

git clone https://github.com/rubygems/rubygems.git
cd rubygems
git checkout v1.8.25
ruby setup.rb

Install feedjira and other Ruby prereqs:

gem install feedjira mysql2 nokogiri

Now you can use the getfeeds.rb file. Fun!

Crontab

Put this in your crontab file:

*/5 * * * * cd /path/to/fuckreader/ruby; /usr/bin/ruby getfeeds.rb > ../logs/feeds.log
0 3 * * * cd /path/to/fuckreader/www-includes; /usr/bin/php cleanup.php > ../logs/cleanup.log

You're free to set it to whatever frequency you want. (See note below about the cleanup script, btw.)

Notes -- please read

The biggest and most important note I can give is that this is an unfinished project and it has quirks. For instance, the cleanup script deletes stuff older than 30 days, but then if there are things older than 30 days in the RSS feeds you're scraping, they'll pop back up. That's annoying. Haven't figured that out yet, it probably means I'll need to cache things forever in some fashion. So for now, I've turned the cleanup script off on my own server.

Making new accounts on the site is invite code based, and when you want to make your own account, just add a new key to the user_invites table. It doesn't matter what the actual invite code is, because you'll use it to make your first user. So you could just run this command via MySQL:

INSERT INTO user_invites (invite_code, owner_id, tsc) VALUES ("butts", 0, UNIX_TIMESTAMP());

And then you can register a new account using the invite code "butts". If you'd like to give every user a bunch of invite codes, just use the script www-includes/give_everyone_keys.php via PHP CLI or by moving it to the www folder and going there in a browser. But don't keep it in www for long, as users might know to exploit it.

There are a lot of planned features. I have no idea what'll be breaking what in the future. However, since I do use this every day, I probably won't make any features that seriously breaks what's already here.

The getfeeds.rb script is not as efficient as I'd like, and with very dense RSS feeds (ones that update very often), it can choke a lot of memory. I plan on rewriting it in a better language, like Go.

There's some Redis functionality that is half-in here, but it doesn't actually do anything, so don't worry about the Redis stuff. It was a failed experiment.

There's a lot of MySQL indexing stuff going on which makes this all a lot faster, but in general if you have 50+ feeds, the site will be kinda slow. Be patient with it.

Change the index.html page as you see fit.