studio-engine 0.65.0 → 0.65.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8914b74430f105820bbc5a41f44d6c573ff88492bcd882534f769fa2c2a11608
-  data.tar.gz: f3d1ffbec8dcb09ef354d21ab3786d2d5b9b0590a665ac95a949e7282668a66b
+  metadata.gz: 3aa4604073fec8349d92e58a72bfb14d5c87fbaf03081f4bcb3e430a0ec35bf6
+  data.tar.gz: 3653491831ad04c4cb7aad8f1950297655b42a0582864173c998cbf8ea005a06
 SHA512:
-  metadata.gz: d52952922f19f864f97fd09297d8711c246ba54415bd40d60682569cbdfe0d51ccf7fb7673e7d921a361ab3657b35871f4999aad0baa80374dd3098cd3c267b2
-  data.tar.gz: 82afd75d6a4e57bfbe3708b59c1c1e7eec188cf2bcfd5921e3eeb50e006090d1f0839e5fc280be8ee60c8d6dc999bb07bb148e7e2220b7d061559048e14b6225
+  metadata.gz: 38cd8f8eedcf34cb0cac87d4c7737bb3fd29af673651e4593a1a93f0e91e822e582500e62ffdb7172c6de1d6f8673fe6502da4717376b789ee91f0e9b9ed2fc4
+  data.tar.gz: 2ac51eaf964ad5ef2d9e026298dd0565466540de68ae68677c77eb2b802f17bc392691d63594794bd537282b2a9d315fc64c01e52800a72a82cc87cf5a1ed329

data/app/assets/tailwind/studio_engine/engine.css

@@ -395,6 +395,28 @@ body.modal-open .studio-app-banner {
    x-data="navCollapse()" on it, and override only the endpoints that differ
    from the defaults below.
 
+   ADOPTING IT IN A FORK COSTS MORE THAN THE OPT-IN, corrected 2026-08-28 after
+   checking against a real forked header. Two attributes is what it takes to opt
+   IN; it is not the job. mcritchie-studio's inlined header needs, at minimum:
+
+     REMOVE the bindings that fight the primitive —
+       :class="scrolled ? 'py-2' : 'py-6'"        on the row
+       x-bind:class="scrolled ? 'w-5 h-5' : …"    on the logo
+       x-bind:class="scrolled ? 'text-base' : …"  on the h1, which is a DISCRETE
+         swap and therefore the exact 1px reverse lurch this primitive exists to
+         remove — a fork that keeps it reproduces the bug it is adopting to fix
+       transition-all duration-300                in three places
+     ADD the hooks the band table actually keys on —
+       nav-row, nav-logo, nav-title, nav-logo-link
+       (nav-logo-link is MOBILE-ONLY: omit it and a desktop check still passes,
+        which is how an adoption ships half-done)
+     WRAP the h1's bare text node in a span so there is an element to size.
+
+   That is roughly fourteen edits across five elements. The SIZING LAYER does
+   reach every app — that is the win this block delivers and it is unaffected.
+   Only the cost sentence was wrong, and an understated cost is how an adoption
+   gets started and abandoned half-done.
+
    LAYERED ON PURPOSE. These are DEFAULTS, so a consuming app has to be able to
    beat them, and unlayered CSS beats every layered rule regardless of
    specificity — shipping them bare would make the engine's numbers

data/app/views/layouts/_navbar.html.erb

@@ -74,8 +74,17 @@
        rendering THIS partial got it. Three of six apps fork the navbar, so each
        was left to hand-write its own — the way four independent copies happened
        in the first place. It ships from engine.css now (every consuming app
-       imports it), so a forking app adopts the collapse with `nav-shell` +
-       x-data="navCollapse()" and overrides only the endpoints that differ. */
+       imports it), so a forking app OPTS IN with `nav-shell` +
+       x-data="navCollapse()" and overrides only the endpoints that differ.
+
+       THE OPT-IN IS NOT THE WHOLE JOB — this is the partial a forker actually
+       reads, so the caveat belongs here too rather than only beside the band
+       table. A hand-written header must also REMOVE the bindings that fight the
+       primitive (a :class size swap on the row, x-bind:class on the logo AND on
+       the title, transition-all alongside) and ADD the hooks the table keys on
+       (nav-row, nav-logo, nav-title, and the mobile-only nav-logo-link). The
+       title swap matters most: it is DISCRETE, which is the 1px reverse lurch
+       the primitive exists to delete. Full accounting in engine.css. */
   </style>
   <div class="nav-row flex items-center">
     <%# Left side: logo + nav %>

data/app/views/layouts/studio/_head.html.erb

@@ -541,9 +541,10 @@
           // KEEP FRAMES COMING WHILE CATCHING UP. Nothing else will schedule
           // one: the finger is off, so no more scroll events arrive, and
           // without this the collapse freezes wherever the clamp left it. It
-          // converges LINEARLY and lands exactly — about 4 frames (~67ms) from
-          // a hard flick — rather than crawling asymptotically the way the
-          // 300ms ease this replaced did.
+          // converges LINEARLY and lands exactly — about 8 frames (~133ms) at
+          // the mobile --nav-ramp of 120, and about 10 (~167ms) at the default
+          // 144, from a hard flick — rather than crawling asymptotically the
+          // way the 300ms ease this replaced did.
           if (!snap && p !== target) {
             queued = true;
             requestAnimationFrame(apply);

data/lib/studio/version.rb

@@ -1,3 +1,3 @@
 module Studio
-  VERSION = "0.65.0"
+  VERSION = "0.65.1"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: studio-engine
 version: !ruby/object:Gem::Version
-  version: 0.65.0
+  version: 0.65.1
 platform: ruby
 authors:
 - Alex McRitchie