Working with domain experts can be frustrating for all parties involved - it’s very hard for them to get the subtle nuances of their craft across, and aggravating for developers to constantly tweak and refine code to reflect ever evolving specifications.
There is, fortunately, a solution: Domain Specific Languages, small, highly specialised languages that allow experts to write exactly what they mean, without needing to involve developers for each new tweak.
These DSLs can be a lot of fun to use and write, but are also easy to get wrong. The purpose of this talk is to walk through a slightly simplified version of a language currently used in production to process millions of documents, show the most common problems that need to be solved, and offer reasonable, easy to understand solutions.
At the end of the talk, attendants will have a solid understanding of how to design their own DSLs, and of the strengths of weaknesses of their standard encoding.
Conor (on behalf of the LL organisers)