Repository navigation
Commit 90b9344
authored
chore(deps): Architect at tag v0.14.0 (#124)
Keyflow's Architect pin moves from `main` `9d1f2f48` to the tag
`v0.14.0` (`a236142c`), as studio pins it. This is the bottom of a
stack; the rough-edge fixes from the walkthrough sit on top.
## What moves
- **Entries:** all 29 Architect entries: the workspace's, the `[patch]`
table's (the `www.` spelling that lifts `daw`'s copy) and `keyflow-ui`'s
`architect-workspace`.
- **Lockfile:** it holds the one tag and no other Architect revision.
Besides the source lines, the only change is that one Architect crate
now depends on `data-url`, which was already in the lockfile.
- **What v0.14.0 brings past `9d1f2f48`:**
- #874: a remote view's images draw, so the palettes' glyphs show in a
process view (they were blank at `9d1f2f48`);
- the wire-8 tests (#879).
Nothing in Keyflow needs to change for it.
## Checked
- **Build:** `cargo check` of the extension builds against the tag with
no local patch.
- **`just ci` here was red, and neither failure is the pin's.** The
machine was at load 107–131:
- `edit_vault` (the whole dev vault through note input) hit nextest's 30
s timeout. It takes 21 s at load 40 on `main` too, so it's slow, not
broken; it got there once #122 made every bar editable. A nextest
override for it is the first commit of the next layer, as the screenshot
test already has.
- `session_host::tests::a_presence_only_leader_is_followed` timed out
("the leader never served"). Alone it passes three times in under a
second. It's the session tests' known flakiness under load, the session
thread's, and noted here.
- **keyflow#68's `bump-architect` bot:** expects a tag, so it can take
over from here.3 files changed
Lines changed: 119 additions & 118 deletions
0 commit comments