From 2012e8644273a36c19510a45501fc3d802dd6269 Mon Sep 17 00:00:00 2001 From: Joseph Humfrey Date: Wed, 10 May 2017 16:12:50 +0100 Subject: [PATCH] Remove mention of enums feature from "probably coming" --- README.md | 1 - 1 file changed, 1 deletion(-) diff --git a/README.md b/README.md index 8b460040..93d1f6dc 100644 --- a/README.md +++ b/README.md @@ -156,7 +156,6 @@ Internally we've been thinking about the following. We can't guarantee we'll imp - A scheme to split up the monolithic JSON output into smaller files that can be loaded on the fly in the runtime. This was necessary on Sorcery! and 80 Days as the quantity of content increased substantially, and we were running our games on low-end iOS devices at the time. - Our implementation for the previous version of ink was to have two files: one huge text file with lots of JSON snippets that never got loaded in one go, and an index file, which contained byte offsets and lengths for all the compiled knots in the game. This worked pretty well, although it meant that the compiled JSON was still in one huge file. - A possible alternative we could consider for ink2 is to (optionally?) be able to have a one-to-one mapping between source `.ink` files and output `.json` files, so that the size and arrangement is predictable and controllable. - - Some kind of enum / flag system. Design yet to be determined. - General refactoring and improvements to code structure and optimisation of the compiler. - Structured JSON-like data objects within ink format. Exact design still to be determined, but goals are for it to be a superset of JSON, so that it’s compatible, but can be simpler (a bit like YAML, though not YAML for various reasons). Would allow more complex hierarchical game state to be stored within the ink engine.