Week 42

On User Needs this week, I made a bunch of fixes to Maslow for the minor issues raised from our IT Health Check (ITHC) last week. They were things like adding a logout link, enforcing Rails’ CSRF protection, and setting a couple of extra headers. I also added logging to the Need API’s organisation importer, and wired it up to our Kibana instances, so we should be able to more easily keep an eye on whether it’s running correctly each day.

I removed a field from Maslow which asked if a need was currently met by GOV.UK because it’s confused most of our users in testing, and the users who do need to know this are capturing it in a different way. I imagine it could return in the future as some sort of state, rather than a question.

With some help from Gemma, I’ve added a list all the ‘mainstream’ content for a user need on its page in Maslow. This is most definitely a first iteration, as we don’t currently return any information about the state of the content or whether it’s live or not, but it should be enough to indicate when there is content present for a need.

I’ve started a spike into how we handle duplicate needs. It’s a tricky thing to get right and opens up a ton of questions, like where the line is between a duplicate and two very similar needs from different organisations, and then how we actually go about 'closing’ a need.

Around GDS this week, I’ve also been helping run the pair programming exercise in a few interviews, which has been quite a bit of fun. When we interview new developers, we spend about 45 minutes with a code challenge and we pair to solve it. I’ve enjoyed seeing a variety of approaches to the problem.

And finally, I went to Hacks/Hackers London this week - an evening event at Wayra where journalists and devs turned up to talk about technology and data. There was an interesting talk by Roger Beecham about his work analyzing data from TfL’s cycle hire scheme.