Not a D language topic, but there is some data science interest in the community, so...

I and one of my eBay search colleagues are running a data challenge as part of the 2019 SIGIR eCommerce workshop (July 25, Paris). The task involves a common problem in eCommerce search applications: choosing a good recall set when a non-relevance sort order is used. The challenge is open now, for info see: https://sigir-ecom.github.io/data-task.html. No need to be a search expert to give it a try.

This is what I was working on while many of you were at DConf.

--Jon