The Setup Part 1: Website

1 min read

This series will have a focus on DIY computer stuff, privacy issues, automation and security related topics.

One thing I want to achieve this year, is independence from most customer cloud services. Customer cloud services meaning services that I use for personal stuff. That includes stuff like Google Mail, Google Calendar, GoogleDocs/Dropbox and many more.


As a first step, one thing I recently did was moving my blog away from GitHub Pages. This blog is now hosted on my own server. This costs a little money, but I have full control over the environment. The setup is different for each provider, but essentially it’s a git post-receive hook that triggers jekyll to build the page:

# $HOME/website.git/.git/hooks/post-receive
GIT_REPO=$HOME/website.git
TMP_GIT_CLONE=$HOME/tmp/website
PUBLIC_WWW=$HOME/html

. $HOME/.bash_profile
git clone $GIT_REPO $TMP_GIT_CLONE
$(which jekyll) build -s $TMP_GIT_CLONE -d $PUBLIC_WWW
rm -Rf $TMP_GIT_CLONE
exit

More information about deploying with jekyll can be found here.

The biggest advantage I see is being able to customize the post-receive hook. There are so many possibilities, like automatically posting a tweet once a post is published.

It’s important to note that if you are planning to move away from GitHub Pages, you don’t necessarily need a hoster that offers Ruby support. All you need is a webserver that can serve static files. You would first generate your blog with jekyll locally and then upload it it your server via FTP.

After moving the website away from GitHub, I also moved most of my GitHub repositories to GitLab. As a result, most of the links to my GitHub repositories have to be changed at some point. A mirror of this websites repository is available on GitLab.

I haven’t written any other posts for this series yet, but potential next topics include: Owncloud, Laptop Setup, Backups, Email, Password Management and Android liberation.

11 May 2014
blog comments powered by Disqus