feat(editor): Settings > Appearance gets "Show timeline on hover"

Off by default. On, the timeline sits as a strip — its header, so the title,
transport and Composer stay reachable — and opens to full height while the
pointer is on it or something inside it has focus.

It also opens for the whole of a DRAG. During one the pointer is in drag mode
and a strip that stays shut is a drop target you cannot hit, so dragenter holds
it open and drop/dragend closes it. The tracker counts depth rather than
holding a boolean: dragleave for the element being left fires AFTER dragenter
for the one being entered, so a flag closes the strip mid-drag as the pointer
crosses from one clip to the next.

The strip height is MEASURED from the header rather than fixed: .zone-head
wraps when the zone is narrow (34px min-height, 138px on a normal window), so
any constant clips the very controls the strip exists to keep visible. It is
re-measured on resize. flex: 0 0 auto matters — .zone is a shrinkable column
child, and a height with min-height: 0 let the workspace squeeze the strip to
1px. The body collapses rather than being display:none so the lanes are not
torn down and rebuilt on every open.

A browser preference like the colour scheme, not a project one, and registered
in the same Settings section for that reason.

Also in this commit:
- vf: train wherever there is a GPU, not only when the sample sits on one. The
  x0 value function's snapshot loads with map_location="cpu", so keying on the
  sample left it training on the CPU — the run that named it measured
  "x0 vf: train 10.1s" against 46ms for the same net locally.
- The stage line counted nested sub-stages twice and reported 153% of a run.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

value_function.py

@@ -139,10 +139,19 @@ class OnlineValueFunction(nn.Module):
 
     @staticmethod
     def _training_device(conditioning, home):
-        """Where to run the twenty training steps: the sample's GPU if it is on one."""
+        """Where to run the twenty training steps: a GPU if there is one at all.
+
+        Not "the sample's device". The x0 value function's sample is a snapshot loaded with
+        map_location="cpu", so keying on the sample left it training on the CPU — measured at
+        10.1s a run against 46ms for the same net locally. What has to move is the ensemble
+        and one stacked batch out of the buffer, and both are small; where the sample happens
+        to have been read from says nothing about where the arithmetic belongs.
+        """
         try:
             if isinstance(conditioning, torch.Tensor) and conditioning.is_cuda:
                 return conditioning.device
+            if torch.cuda.is_available():
+                return torch.device("cuda")
         except Exception:  # noqa: BLE001
             pass
         return home