reitit VS missionary

I would build the CLJS for your game logic. Then I would build the backend server separately with API endpoints your front end to talk to. https://github.com/metosin/reitit is a pretty good option.

A significant feature of reitit is that routes are composed of simple vectors and maps. That means you can pre-process that data structure however you want before building the router.

You might look at the examples on the Reitit page. Reitit seems to be the emerging library for routing, and it does get you most of the way there, no matter which direction you are looking to go..

for runtime validation & coercion, you can use the pedestal-utilities from reitit which support all of spec, schema and malli. For dev-time var instrumentation (static & dynamic), there are several ways to add those to existing or 3rd party codebases.

I think the reactive query problem is more of a spectrum of tradeoffs, there's a middle ground between "full page refresh on nav" and "refresh all query subscriptions per user per tx". Truly realtime things like chat come from a streaming event source (not a database) and even in a chat app, most of the information coordinates on a page is slow moving. So really this is about regaining control over concurrent data flow so we can sample different views at different speeds. See technologies like https://github.com/leonoel/missionary.

I wonder if anyone can compare these to missionary?

related: https://github.com/leonoel/missionary

- https://github.com/leonoel/missionary

https://github.com/leonoel/missionary has excellent syntax, it uses a macro to extend regular clojure syntax with monadic join operator – basically turning clojure sexprs into do-notation

I recommend having a look at missionary - it is probably the most essential library in our application. But it depends on what type of application you are building.

Process supervision is what Missionary implements: https://github.com/leonoel/missionary

Agree - mostly the opinions are just dated, in 2005-2012 the pure FP world (especially Scala) was in rough shape but then in 2018 Maybe Not (the really controversial talk about "Maybe Sheep") it didn't seem like he had taken the time to understand haskell. I would love to see how his opinions have evolved since then, in the 2017 interview with fogus he said "If I had more free time, I’d spend it with Haskell". Haskell has especially come a long way in 2018-2022 with the popularization of functional effect systems which are extraordinarily powerful; for example https://github.com/leonoel/missionary (2020-2022) is the best Clojure effect system and is a leap forward over core.async (2013), but that 9 year difference is a lifetime in CS