Updating our blog
This week we ported our blog from Jekyll to (and I mention this anticipating a wave of WTFs) a custom rails solution we built in a few hours. The reasons for doing so were numerous, however the primary one was that it was causing a blocker for people writing posts.
Over time, and much to our surprise, we found that it is much easier to log into a web system to create a post than mucking around with the technical tools we use on a daily basis.
Instead of simply writing cool stuff, we were spending too much time getting Jekyll installed on people’s computers, running into permissions issues with our deployments and generally cursing loudly. The net result that people were saying they couldn’t be bothered.
In their defence, Jekyll-like solutions are likely to be more successful on a single-author blog rather than on a multi-user system, where the cost of set up is not multiplied across the team.
In the end we were risking mutiny and needed to make it easier for everyone to get stuff out there.
Another reason we needed to change system was that we missed the comfort of having a dynamic system with models we could add to and muck around with. This meant we were unable to easily add a few customisations such as providing an online home for some our pet projects, and giving authors profiles. We should be popping these things up shortly.
We also tried really hard to make the RSS feed match up so we wouldn’t bugger up everyone’s feed readers. Unfortunately we encountered a problem with date casting which happened when we piped our converted Jekyll posts to Heroku through their awesome taps system. Some data got a bit messed up and the result was that we probably irritated a lot of subscribers in the process.
Sorry about that, but we should resume normal scheduling as of now.