NOTE: this is kept for historical purposes, I switched to using jekyll now.
Git Content Management System
How is this website run?
This is a django website with a twist: all content is stored in text files. I manage those files with git. A little script loads those files into the database and then django serves them.
In fact, all of the content and its history is actually publicly available (but you don't have to make your history available in order to use this system). You can all see this an example of how to use django-gitcms.
Where can I get the code for this?
Github, of course. The code is license Affero 3 (GPL for web applications).
I have also released it on PyPI if you prefer. (This also means that pip install django-gitcms or easy_install django-gitcms will automatically find and download the latest version).
Your website is using django. This means that it's easy to add any functionality you want.
- Out of the box, it supports:
- Restructured text articles (pages such as this one)
- raw files (for images and such)
- a conference system
- a blog
but the main point is that the architechture is easily extensible. I'm selling you an idea more than software.
Why is this worse than gitwiki?
Because there is no web front-end to the history or even an edit button. It's git all the way. It's not friendly to non-technical users (it's very friendly for technical users—those of us who live on the command line and don't want to have to learn a new way of doing things).
Why is this better than gitwiki?
Because you can store any type of data, not only hyper-text files.
For example, in this website, the conferences is a django app whose data is a text file that looks like:
name: International Symposium on Biomedical Imaging (ISBI) short_name: ISBI location: Rotterdam, The Netherlands start: April 14 2010 end: April 17 2010 submission_deadline: November 2 2009
followed by many more examples. Actually, the menus on this website are also saved as a yaml file, while the publications app saves its information as a bibtex file (I use exhibit for the display, but pre-process the bibtex file everytime it changes). I can write any functionality I want with django and store the information as a text file (I only have to write a load() function in Python, which can do whatever I want).
This is not really comparable to wiki-git-like projects. We just have different goals.
Why is this better than yst?
Because it uses django so you can do any Python processing you want. Also, the website does not have to be static, so you gain a lot of flexibility there.
Why is this better than nesta?
I don't know how nesta works because I don't know RoR. But I know Django on Python, so I wanted to use a system based on those technologies.