2 ms·
For those wondering if there's actually something relevant to the Epic's approach to the metaverse: there's not beyond slide 3. Everything after slide 3 is a de
by tinco 4y ago
For those wondering if there's actually something relevant to the Epic's approach to the metaverse: there's not beyond slide 3. Everything after slide 3 is a description of the semantics of a novel functional logic programming language, the sort of which you'd expect you'd get if you paid the world's premier Haskell core contributor a large sum to design a new language. Of course its intended audience is Haskell academics, so perhaps it would be weird to expect anything else.
Slide 61 and 63 hint at things that might actually be of consequence to real programmers making real software for a potentially real metaverse, such as the effect system for I/O, transactional memory and code organisation through classes and inheritance, but it gives no detail.
If you're interested how elegantly functions like head, tail, cons, snoc, append and map might be implemented, this presentation is for you. If you're more interested in how you might unpack a message from bytestring, model a state machine or coordinate a group of actors in a distributed system, then I guess you'll have to wait for the next presentation.
- yazzku 4y agoSimon my boy. The "View from 100,000 feet" slide does highlight the kind of experience SPJ would have in identifying where Haskell lacks a bit in pragmatism as a basis for a new, but not super different language. Hopefully the tooling is much better though.
- helsontaveras18 4y agoI was expecting an entire section devoted to how this language would deliver on its promise on powering the metaverse. This was a nice post about a new language, but I was ready for a bitcoin paper.
- tome 4y ago> its intended audience is Haskell academics It was presented at Haskell eXchange so the audience is more like Haskell users in industry and those who would like to be.