Workaround to avoid full hang (#589)

As the scale factor becomes unreasonably large, the number of
subdivisions of an Euler spiral into lines grows, making the flatten
stage take too long.

This patch just bounds the number of subdivisions, so it will eventually
make progress. It will be slow, but not hang.

A better solution would be to aggressively cull so it only generates
geometry inside the viewport, but that is considerably more complicated.

Workaround for #548

crates/shaders/src/cpu/flatten.rs

@@ -333,7 +333,7 @@ fn flatten_euler(
                 let n_frac = scaled_int;
                 (n_frac, EspcRobust::Normal)
             };
-            let n = (n_frac * scale_multiplier).ceil().max(1.0);
+            let n = (n_frac * scale_multiplier).ceil().clamp(1.0, 100.0);
 
             // Flatten line segments
             log!("@@@   loop: lines: {n}");

shader/flatten.wgsl

@@ -434,7 +434,10 @@ fn flatten_euler(
                 let integrand_peak = sqrt(abs(k_peak * (k_peak * dist_scaled + 1.0)));
                 n_frac = integral * integrand_peak / a;
             }
-            let n = max(ceil(n_frac * scale_multiplier), 1.0);
+            // Bound number of subdivisions to a reasonable number when the scale is huge.
+            // This may give slightly incorrect rendering but avoids hangs.
+            // TODO: aggressively cull to viewport
+            let n = clamp(ceil(n_frac * scale_multiplier), 1.0, 100.0);
             for (var i = 0u; i < u32(n); i++) {
                 var lp1: vec2f;
                 if i + 1u == u32(n) && t1 == 1.0 {