Friday Hug

FridayHug.com (former HugFriday.com) is a movement to help spread the goodness. Share your hug every Friday on twitter using the hashtag #FridayHug

Environment

• Ruby 1.9.3-p194

• MongoDB >= 2.0

• Bundler >= 1.2.0.rc.2 (supports ruby version on Gemfile)

Setup

Install Bundler 1.2.0.rc.2 or newer:

gem install bundler --pre

We need Bundler 1.2 to use Ruby 1.9.3 on Heroku.

Create database config mongoid.yml:

cp config/mongoid.yml.example config/mongoid.yml

This is the configuration file used by Mongoid, you can check more details here.

Create AWS config s3.yml:

cp config/s3.yml.example config/s3.yml

The file is self explicatory, just add your Amazon AWS credentials and your S3 bucket name to host the thumbnails.

Now you are pretty much set, the project should run but with no content.

Update Script

We have to update our database from time to time, watching a few keywords on Twitter and we won't do that on every user request, because this would be resource and time consuming.

So, in scripts/update.rb is where that comes in place, we run it periodicaly (every 10 minutes) using any scheduler you have available on your system. We use the the UNIX CRON for it and the command will look something like:

*/10 * * * * /bin/bash -l -c 'cd /my/app/fridayhug && RACK_ENV=production bundle exec ruby /my/app/fridayhug/scripts/update.rb >/my/app/fridayhug/last_update.txt'

Since we use Heroku, we run this script on a personal VPS, separeated from our production server. Why? We could use their scheduler, but it will use our available free dyno time and we would end up paying for it.

Tests

There's none. Yes, I know... what I shame. This was a one person and one weekend Rails Rumble-like project, hopefully there will be tests in the future.

Heroku

We use Heroku to host our little app and we use their awesome free tier. In order to get the app running on Heroku, there's a few extra configurations you need to do:

Since Heroku uses a very easy but peculiar git push deploy style, we can't host configuration files on our source repository, it's a bad practice (oh yeah, we did that in the past, when our code was still behind the curtains, shame on us).

How Heroku handles configuration then? Here come in scene Heroku Config Vars.

And our app supports that, if some key vars are present we will use that instead of using the YML files.

What you need to do is:

1. Forget about mongoid.yml and s3.yml, they are useless on Heroku.

2. Create MONGOLAB_URL var:

heroku config:set MONGOLAB_URL=mongodb://USER:PASS@ds0a4443.mongolab.com:37407/database_name

You can find the mongodb URI on your MongoDB service amin panel.

3. Create AWS vars:

heroku config:set AWS_KEY_ID='AKIAIEFGBBBESOJGQ' AWS_SECRET='qr1qdcEWpyCudzyZMdfg2344sdfsdfdsfdsf' AWS_S3_BUCKET='fridayhug'

4. Create RACK_ENV var: heroku config:set RACK_ENV=production

OPTIONAL

There's a very simple admin interface, that enables you to remove or hide hugs from the homepage. But to access it, you need to setup the admin vars:

heroku config:set AMDIN_USER=username AMDIN_PASSWORD=password

And then access the path /manage-hugs for the admin.

License

Friday Hug is licensed under the WTFPL license:

DO WHAT THE FUCK YOU WANT TO PUBLIC LICENSE 
                    Version 2, December 2004 

 Copyright (C) 2004 Sam Hocevar <sam@hocevar.net> 

 Everyone is permitted to copy and distribute verbatim or modified 
 copies of this license document, and changing it is allowed as long 
 as the name is changed. 

            DO WHAT THE FUCK YOU WANT TO PUBLIC LICENSE 
   TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION 

  0. You just DO WHAT THE FUCK YOU WANT TO.