classifyProjectTransition now checks the stored GUID first, then the pointer,
fixing the w10 regression where a recycled ReaProject* address stopped the bank
reloading. poll()'s fork re-GUID gate is bound to !sameProjectObject. Full
transition matrix pinned in tests.
Register toggle/activate/tag/untag/show-both actions driving the D2
shell and D3 model; reapply saved active mode on project load via a
persist load-signal seam. Shared guidString helper; pure nextModeId
unit-tested.
M4's GUID-only classifier read a tab-switch between two Save-As forks (shared
copied GUID, different paths) as a Save-As and clobbered a bank. Thread
sameProjectObject into classifyProjectTransition: a different object always
Loads, never relocates; a forked sibling gets re-GUID'd to diverge.
Add forever-stable "view_state" key under the "reasampler" namespace.
saveToActiveProject additionally serializes the ViewModeModel;
loadFromProject restores it on every load path, defaulting on an
absent/malformed key. Model state only — no visibility apply (D4).
After capture, saveToActiveProject wrote ext state but left the project
clean, so Ctrl+S was a no-op and the bank was silently lost on close.
Also flushes the Save-As re-GUID on the normal save flow.
Serialize BankIndex to project ext state ('reasampler'), reload on project
load, resolve paths project-relative. Save-As copies the bank to the new .rpp;
identity keyed off a minted GUID (not the recycled ReaProject*) so project
switches don't clobber banks. Pure classifyProjectTransition tested.
daniel