#039

I fixed an issue with atom feeds for “finders”: pages which find things, like https://www.gov.uk/search/all and https://www.gov.uk/aaib-reports. Previously finder-frontend, the app which renders finders, would only serve an atom feed if the default sort order for the finder is by recency. We wanted a feed for /search/all, but that defaults to sorting by popularity, so no feed. I made finder-frontend fix the sort order for feeds, rather than just refuse to serve them.

The rest of the week had a lot of Elasticsearch 6 stuff, and the end is finally in sight. The ES6 cluster is deployed to our production and staging environments (previously it was just in integration); search-api is able to talk to both clusters (not my work); and we’ve got a plan for A/B testing the switch from Elasticsearch 5. There are only two blockers remaining before the A/B test can start: we need to import all the data from Elasticsearch 5, and we’re going to have a chat with some AWS solutions architects to check we have the right machine specs for our workload.

I decided to try my hand at some decorative typesetting, and picked the King James Bible. I’ve put it up on GitHub, and have now got the Pentateuch done with only one really ugly part. It’s been a fun learning experience as I’ve not tried anything like this before. A long time ago, in preparation for writing my Ph.D thesis, I read Butterick’s Practical Typography to gain some basic knowledge, but there’s still much to learn.

I’ve now been using DuckDuckGo as my main search engine for a few weeks now, and it’s pretty good. This came out of an experiment at work to figure out what a “good search experience” and a “bad search experience” is: turns out Google doesn’t have a monopoly on good search experiences.