This blog has been lingering for far too long now. The reasons for that are manifold, but there are three primary ones which I would like to elaborate further.
Flick a Fruit, for iPad
Shortly after finishing this blog, in early January 2012, a good friend came over to visit me and we started hacking on an indie project which we were both very excited about: A fun game for the iPad. We estimated it to be around 3-4 months of work and spend the whole January devotedly hacking along. Sadly, we grossly miscalculated the required time, we kept adding features to it, and we both could only work part time on the project as we were also busy with other things. All in all, it took us almost 1.5 years to finish the project. I'm really proud about it and I learned a ton of stuff along the way, but working on this huge game in addition to managing my other projects (most notably InstaDesk nee PhotoDesk1) consumed so much time that I stopped even thinking about writing new entries for the Appventure blog. The game is currently in a market test in Canada, if you're interested, give it a try!
Lack of Focus
Also, 2012 and especially 2013 were transformative years for me. While honing my Objective-C knowledge, I felt the accumulated wisdom of 10 years of web applications (Javascript/PHP and later Javascript/Python) slowly fade away. Objective-C is ill suited for web apps2, and I'd grown weary of Python and wanted to try something new. So I spend most of 2012 and 2013 trying out various languages and playing around with different frameworks. All this resulted in a certain lack of focus of what to write about.
Jekyll Static Site Generator
The previous version of this blog ran on Jekyll, a static site Generator for Ruby. It sounded like a great idea back then. I could write my posts comfortably in Vim with Markdown and would not have to mess around with online editors or weird UI. Also, just having a static set of html would ease the burden on the server and conveniently expose no security issues3.
I quickly had everything running in Jekyll and all was good. However, after a some time, I constantly ran into issues with this setup. Whenver I wanted to write a new post, Jekyll would crash. I'm not very familiar with Ruby and I hardly use it for anything, but from time to time there're tools that I want to use that are written in Ruby, and then they need different Ruby versions, and different Gems, and then you need rvm or rbenv in order to manage all that, and somehow the entangled mess that I've created here lead to a point where whenever I wanted to use Jekyll, it would crash because something crucial was missing. I usually would spend one hour debugging this, until Jekyll ran again, and then, the next time I was about to write something, there would be a new error.
A couple of days ago I wanted to write a post again, only to be greeted by a new error, 'require cannot load such file, inflection error'. This was especially demotivating as I had spend a considerable amount of time updating Jekyll to the latested and greatest and getting everything running again a couple of weeks ago. At this point, I decided I needed a better solution. Something compiled which would always work no matter what happened to my system, something which was not dependent upon a myriad of gems and files in weird locations 4.
Everything new everything better
With the above in mind, I decided to address the problems outlined above and re-think and re-do this blog. Along the way, I also created a new layout that works better on mobile devices, offers more information and in general looks more streamlined.
- Clojure: I wanted to have a static blog solution that can be compiled and just works, even if I update my operating system. Clojure looked like a great solution for this. I forked Static, a static site generator for Clojure 5 and modified it to better fit my needs. I'm really happy with the result.
- Focus: I'm doing lots of research for the various projects that I'm working on, and I've installed and used several documentation systems over the years to store it. The solution that worked best so far was to write it as self-contained Org-Mode documents in Emacs. My main focus for the Appventure blog will be to share some of this research here. What's really great about the Static generator that I used, is that it allows to use org mode documents as input (versus, say Markdown). So I can just take my research docs, which already are Org mode documents, and publish them here on my blog.
- Time: I've also finished a lot of projects over the last year so that I can invest more time into other things - including this blog - in 2014. Most notably, I am back on schedule, which really wasn't the case for the past 6-8 months.
So all in all things look good and I'm looking forward to writing more. In addition to that, I really enjoyed working on the Static extensions in Clojure and I'm confident that a continous usage of this blog will help me extend it even more.
A change in Instagram's marketing guidelines forced me to rename the app in late 2013
Even though I tried by using it for frontend work via Cappuccino
I've been burned by WordPress more than once simply because I forgot to update to the latest version quickly enough.
I am not implying that this is a Ruby or Jekyll error. I think that the issue was that I know too little Ruby and that my environment was all messed up. In hindsight I should have gone with a Python static site generator, since I know Python.
I'm still in the process of finishing the final tidbits before I'll push my modifications to Github. The main difference is that I encapsulated the layout using Enlive templates compared to the original Static, where the layout was mixed between a Clojure file and the compiled sourcecode