Notice that none of the critics or preservation groups are offering to buy the land themselves to preserve it, because it's clearly too valuable. Hence why building denser housing makes sense!

I'm planning to run a draw steel adventure for my home group, so very curious to hear your experience over time!

Yes the key question is, will new players bounce off that asymmetric class design, or will the depth lead to the "expert enjoyment" that you get from a game like Overwatch or League, mastering a character.

Difference between video games and ttrpgs, though, is that the player is the rules computer.

I do wonder if "enjoys xcom" is more of a prerequisite for someone enjoying Draw Steel than for 5e.

More rules and stuff to track hopefully means more meaningful choices during combat. So a player needs to value fun tactics more than the cost of noodling over a larger rulebook.

Fantastic. IMO this picture also highlights sound transits sub par interior design. No central poles to hold onto in the middle, forcing people out to the sides, where people are already sitting.

Who is working on this?

It might, I need to experiment more.

I think there are many app devs like me who are used to React Query/Mobx/Redux apps, ie big mutable client caches, and moving from that rich ecosystem to RSC feels hard.

Likely, the RSC ecosystem and idioms just aren't as settled yet.

I know there's react.dev/reference/re..., but it feels scoped to a component, and form-based.

What if I have a computed "friend count" state in the nav bar, and a friends list in the main dashboard, and I want the nav bar count to immediately update when I remove a friend from the dashboard?

I really appreciate the articles' pragmatic angle. How RSC follows from common problems.

With RSC, I've found the optimistic update story hardest to understand. I want snappy local updates between distant client components, but all I have is the `revalidatePath` hammer.

Replicache might just be what I’m looking for

Multiplayer video games must struggle with the same client state/server state issues. In my experience, client interactions have always felt instant, though. If my League spell had a loading spinner before triggering, I would think that’s a bug. Do video game frameworks make opt. updates easy?

Of course, there is the server first view that optimistic updates aren’t worth the DX pain. When an app has dozens of interactions, though, small latency delays build to crummy UX.

I am sick of writing optimistic updates, generally. One mutation on the client in JS, another on the server in SQL. Maybe I need to try out a local first framework. Sync instead of fetch. Although, I think I’d still end up duplicating mutations in JS and SQL.