Moving away from the cloud Part I - Introduction

Sat 23 March 2013 by Javex


This will be an ongoing series to describe the steps I took to pull as much as possible of my stuff out of "the cloud" to regain control of my data.

The idea behind this is that with your own server, you can selfhost most of the stuff you let others host for you now. However, there is a slight inconsistency here: Where is that server? In my case I will be using a small linode server. But this server is not under my direct physical control and it is a virtual server so we are in the cloud again. But without physical control over the server you will always need some level of trust (see: Reflections on Trusting Trust by Ken Thompson). I will go here with the following idea: Such a server is not free and companies host their data on it. Should it ever become publicly known that the company operating these servers has validated their customer's privacy, their whole business model is in danger. I personally believe that this is a level of trust I am willing to give.

Enough preface, let's talk about actually doing something. First, we will have to figure out what we actually have in the cloud and how we can replace it. I will list those things for myself and some of it probably matches yours or you have similar solutions.

The following things are what I have "in the cloud" and need readly accessible and synced between devices:

  • Blog
  • Contacts
  • Calendar
  • Mail
  • Arbitrary files
  • RSS Feedreader
  • ToDo Lists
  • Android
  • Read It Later (Pocket), What do you call their service?
  • PKI
  • Bookmarks & browser data (Firefox Sync)
  • Instant Messaging

The following criteria will be used to choose the best service: Ease of transition (i.e. imports, etc.), synchronization with multiple clients (Mobile & Desktop), licence (prefer Open Source) and functionality (try to not to loose too much functionality).


Since the blog is where I am writing this article, I will include it right away. Your choice here clearly depends on your preferences. But since I am a huge Python fan and have not really blogged before, I choose Pelican.

I can write articles in reStructuredText, generate static files and have a blog where no dynamic files are involved once it is generated.


For each of the mentioned services I will create an article that explains how I achieved it and what was necessary. All of those articles will be linked above so you either watch that list or subscribe to RSS.