diff --git a/docs/guide/performance-on-a-phone.md b/docs/guide/performance-on-a-phone.md index 0604a9a..cbcd795 100644 --- a/docs/guide/performance-on-a-phone.md +++ b/docs/guide/performance-on-a-phone.md @@ -86,11 +86,6 @@ lines through exactly the same FFI: (recur (inc i))))) ``` -Spirograph, after the fix - -*Spirograph after the loop was rewritten: the curve draws itself at 55 fps -where the lazy-sequence version managed 14 at the same point count.* - | spirograph, points | 1008 | 1288 | 1552 | 1768 | | --- | --- | --- | --- | --- | | lazy sequence | 14 | - | - | - | @@ -132,13 +127,6 @@ On this runtime, in a loop that runs hundreds of times per frame: None of this is exotic Clojure advice. What is different is the magnitude: on a JIT these habits cost a few percent, and here they cost three to four times. -Penrose P3 tiling -Kaleidoscope - -*Left: penrose, 340 triangles and about 2400 FFI calls a frame, at 57 fps. It -is the scene that disproved the draw-call theory. Right: kaleidoscope, 708 -lines a frame at 58, after the allocation came out of its draw loop.* - ## Where the budget lands With the loops fixed, the remaining limit really is line count. Numbers to diff --git a/docs/guide/porting-an-example.md b/docs/guide/porting-an-example.md index 065fbe6..5e2e481 100644 --- a/docs/guide/porting-an-example.md +++ b/docs/guide/porting-an-example.md @@ -4,10 +4,10 @@ examples. 48 of them need no input at all, and those port almost mechanically. This is what "almost" means, worked through with the five that are done. -![Spirograph](../images/spirograph.png) - -*spirograph, the first port. Same maths as the original, a screen 1206x2622 -instead of 800x450, and a loop it no longer owns.* +Same maths as the original, a screen 1206x2622 instead of 800x450, and a loop +it no longer owns. (No screenshot: every capture in this guide was +raylib-ios's, the sibling build this project was ported from, not this one's, +and they're dropped until there's a real Android capture to show.) ## They are desktop-shaped in two ways @@ -87,14 +87,6 @@ nothing new at all; only `analog_clock` does, wanting `DrawLineEx`, `DrawRing` and a local-time call. Penrose needed rlgl immediate mode only because it fills polygons and raylib's shapes API has no call for that. -Kaleidoscope -Spirograph -Boids -Fireworks - -*Three of the five ports, running on the phone. None of them needed a single -new raylib binding.* - ## Then measure it, because the port is the easy half Two of the five did not hold 60 fps on first run, and neither for the reason diff --git a/docs/site.edn b/docs/site.edn index f744f4e..8482d91 100644 --- a/docs/site.edn +++ b/docs/site.edn @@ -15,9 +15,12 @@ :title "raylib-android" :description "raylib on an Android phone, driven from Clojure on Chez Scheme via Jolt. Seventeen scenes as native arm64 code in a NativeActivity, with no JVM, no Kotlin and no Java in the app." - ;; docs/images holds the GIFs and stills, referenced from the homepage and - ;; the guides. Without this the pages render with every image broken, and the - ;; engine only warns about a missing asset dir, so docs/check-site.sh turns + ;; docs/images holds the raylib-ios captures this project ported from. + ;; Nothing in the homepage or the guides shows them any more (they were + ;; screenshots of a different build, not this one, and got pulled once that + ;; stopped being clearly labelled), but the dir stays configured so an + ;; Android capture dropped in here later just works. Without this line the + ;; engine only warns about a missing asset dir; docs/check-site.sh turns ;; that into an error. :asset-dirs ["images"]