0
Fork 0
mirror of https://github.com/zeronsh/comet.git synced 2026-11-16 18:57:19 +00:00
Commit graph

396 commits

Author SHA1 Message Date
Wing
2d480e9fb7 Zeron 0.3.5: provider accounts, tool result pictures, steadier composer
0.3.4 plus main since (#1331–#1440):
- Per-provider accounts for Pi, OpenCode and Hermes, API keys and Claude
  for Pi; every model in the picker says its provider (#1425).
- A Codex chat's / lists Codex's commands and skills (#1403).
- A tool's result pictures show in its row (#1103, #1186).
- The composer's layout and focus hold steady through moves and resizes
  (#1420); Settings reopens on the page shown last (#1416).
- A prompt the opening tail brings shows once: no frozen transcript from
  the duplicate-key panic (#1433).
- Windows: messages have their time and Copy again (#1440); microphones
  listed by name (#1334).
- macOS: chimes no longer trigger a kernel deadlock that hangs the Mac
  (#1333).
- Edge: chat rooms can offer the relay layer (Wing's account only;
  #1331, #1335).

Version 0.3.5.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-14 17:05:45 -07:00
Wing
82502ace6c Zeron 0.3.4: Windows terminal tabs work; a switch stands right with motion reduced
0.3.3 plus:
- Windows terminal tabs (#1219's three from Daniel, and three more): the
  bridge gets a console (cmd.exe) and runs it raw (one echo, no `>>`,
  Ctrl+C reaches the program), follows the tab's size, opens PowerShell
  without its banner; `~` is the user's profile on Windows (zgx pinned
  at 0ca78b2cb, zeronsh/zgx zeron/pin-0.3.3); the engine lets Ctrl+C
  through to a terminal's shell.
- With Reduce motion on, a switch built on stands its knob on the right.

Version 0.3.4.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-12 23:53:06 -07:00
Wing
464377bbd5 Zeron 0.3.3: setup finished without an account opens the main window
0.3.2 plus 56c3be088: whoever finished setup without signing in no
longer gets the intro and setup on every launch.

Version 0.3.3.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-12 21:05:22 -07:00
Wing
703500ee1e Zeron 0.3.2: Linux windows draw on the display's GPU and cover it at its size
zgx pinned at ac935c6f8 (zeronsh/zgx zeron/pin-0.3.2): 0.3.1's pin
plus, cherry-picked:
- zeronsh/zgx#2107: a window on Linux draws on the GPU its display
  server composites with (linux-dmabuf main device / DRI3, matched by
  PCI address), else the low-power adapter as before. Zeron 0.3.0
  opened black on Hyprland with an RTX card beside a Ryzen iGPU, and
  panicked under Xwayland.
- zeronsh/zgx#2108: a window covering its display on Wayland asks for
  the output's logical size (xdg-output), so the intro fits a scaled
  display instead of opening half as large again.
- #2104's rustfmt-only follow-up.

Version 0.3.2.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-12 14:00:16 -07:00
Wing
0733d73dca Zeron 0.3.1: a view scrolled while a Wayland frame waits is painted whole
zgx pinned at 3651e8d27 (zeronsh/zgx zeron/pin-0.3.1): 0.3.0's
06c6d56ac with only zeronsh/zgx#2104 cherry-picked on top. On
Hyprland at fractional scales, a transcript scrolled while a frame
waited kept its old pixels, torn against its freshly drawn fades.

Version 0.3.1.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-12 12:48:43 -07:00
Wing
d0c77e9c14 macOS: the engine's heap no longer reads as GPU memory
mimalloc v2 maps its regions with the VM tag in its os_tag option,
100 by default (options.c:147, prim/unix/prim.c:307), and on macOS tag
100 is VM_MEMORY_IOACCELERATOR. footprint, vmmap and Activity Monitor
showed the whole Rust heap of the app and the background engine as
GPU memory: a live engine read "819 MB IOAccelerator" for what was its
own heap, and its purged pages as 4.5 GB of reclaimable GPU memory.
An A/B on an isolated engine with a known 256 MiB allocation moved all
of it from tag 100 to tag 254 with the same footprint.

The global allocator now sets os_tag to 254 (Apple's application range,
mimalloc's own fallback) before its first allocation, so the heap reads
as "Memory Tag 254". Accounting only: no allocation changes. A test
reads the region tag of a small and a large allocation back through
mach_vm_region (100 without the fix, 254 with it).

Refs #1369

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-12 02:09:11 -07:00
Wing
8f75cc28f7 Version 0.3.0: the zgx app's first stable release
Wing: "this release is big so it should be 0.3.0". The workspace and both
lockfiles' own crates move to 0.3.0. REMOTE_APPROVALS_SINCE, which only
names the release to update to, says 0.3.0 (there's no 0.2.139 stable).

Audited for a minor bump: 0.2.138's updater and the shipped betas compare
versions numerically with no major/minor or minimum-version rule (0.3.0 is
newer than 0.2.138 and every 0.2.139-beta.N, and a test says so); Windows'
version check is digits and dots; Linux's app/<version> and the beta\<v>
folder take any version; the macOS bundle version 0.3.0.9999 sorts above
0.2.139.N; templates compare by equality, as at every release; the edge,
relay and mesh carry no app-version gates.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-11 14:40:29 -07:00
Wing
e1525b286a Zeron MCP: device_stats reads a device's machine stats, here or on another of the account's devices
A new DeviceStats engine RPC (OS, CPU, memory, storage, GPU, load,
battery, Zeron version; every field optional) and an MCP device_stats
{device?} tool. Another device answers through its relay within 5 s,
or the error says it's offline (last seen), too old, or not one of
this account's devices. sysinfo (already in the tree) with only its
system and disk parts; reports cached for 5 s.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-10 12:34:14 -07:00
Wing
40e1bde977 Merge #1322 (Cloud, relay and GitHub work) into zeron/zgx
Conflicts:
- locales (en, zh-Hans): both new groups kept, zgx's dictation strings
  and #1322's Cloud machine notices and turn states.
- plugins/init.luau: both sides fixed a plugin's own Open on GitHub going
  back into its tab. Kept #1322's (plugin_links.unclaimed around the app's
  opener), which also keeps a comment's link out of the tab; and its test,
  which covers zgx's test's case and checks a chat's link still opens the tab.
- v2's .base.json: re-recorded after the merge.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-10 00:54:33 -07:00
Wing
47300ada2e Merge commit 'refs/pull/1223/head' of github-winglee:zeronsh/zeron into zeron/laurent-chips
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

# Conflicts:
#	apps/zgx/Cargo.lock
#	apps/zgx/ui/ds/composer.luau
#	apps/zgx/ui/locales/en.json
#	apps/zgx/ui/locales/zh-Hans.json
#	apps/zgx/ui/templates/zeron-v2/.base.json
2026-11-09 23:57:57 -07:00
Wing
51b0e1dc43 Merge zeron/zgx into the Cloud, relay and GitHub work
Conflicts resolved:
- Engine checkpoints: zgx's upload in parts now carries the relay layer's
  atFloor/heal flags and hands the edge's refusal back to the relay state.
  The engine's own size cap follows the room's (256 MB in parts), so a long
  chat's checkpoint isn't held back before it's sent.
- Attachments: zgx's local copies of another device's uploads, rewritten with
  the app-capture attribute escaping.
- zgx app: zgx's composer, mention sections, link chips and remote harness
  retries, with the Cloud machine notices, held messages and turn states on
  top; v2 template base re-recorded.
- Linux packaging: zgx's beta packaging, keeping the fat-LTO engine build.
- apps/zgx Cargo pins: zgx's.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 23:21:27 -07:00
Wing
4361fb2200 zgx: one branch: the beta line folded into zeron/zgx
zgx/beta-all-platforms's own commits (the beta workflow for Mac, Windows
and Linux, its version bumps, the engine from the same commit, the
Windows zgx-desktop dependency and WebView2 tabs, the Windows build fix)
join zeron/zgx, so betas are tagged straight from here and all 14 zgx
pin lines move together. The Windows zgx-desktop line is at e875c5e0
with the rest.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-09 01:59:30 -07:00
Wing
8c1c966c79 Merge Zeron main into zeron/laurent-chips
edge/: main's landed checkpoint-in-parts and its size check, in place of
this branch's earlier draft of the same change (the engine half of that
draft stays: it speaks the same parts).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-09 00:55:49 -07:00
Wing
76ae0225fb Merge main into zeron/cloud-relay-v3
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-08 23:21:43 -07:00
Daniel Katulevskiy
533e89d63f
Linux: a dropped image no longer crashes the window with "texture must exist" (#1297)
* Linux: keep dropped images on screen until the next frame, not a crash

Pins zui 15ee708, which keeps a dropped image's atlas tiles until the next
draw and lets the wgpu renderer skip, then redraw, a scene that names a
freed atlas texture. Before, re-presenting a frame after an image was
dropped (an evicted tool image or visual, a resized browser frame) could
abort the window with "texture must exist" in wgpu_atlas.rs.

* Pin gpui-base cab50b0 so it shares zui 15ee708 with Zeron

zeronsh/gpui-component#7 moves gpui-base to the same zui; without it the
lock carried two copies of zui.
2026-11-08 17:46:53 -07:00
José Gurruchaga
a166e1bfbe Merge main into feat/zgx-dictation-plugin
Preserve the current plugin network and preview guards, reconcile the voice dependencies, and integrate protected dictation with the updated side-chat attachment and queue-editing flows in both templates. Regenerate the Zeron v2 baseline hashes.

Validation: 12 dictation/plugin/template/split unit tests and 14 editor/paste/drop/tray integration tests passed. The diff against main passes whitespace checks.
2026-11-08 17:44:48 -03:00
Wing
960a21734b Merge zeron/cloud-attachments-r2
# Conflicts:
#	crates/client/src/live/mod.rs
#	crates/client/src/session/mod.rs
#	crates/client/tests/support/mock_edge.rs
#	crates/sync/src/store.rs
#	crates/ui/src/state.rs
2026-11-07 19:37:42 -07:00
Wing
6095e58cb6 Merge main into zeron/cloud-attachments-r2
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-07 18:14:01 -07:00
Wing
279a985854 Merge main into zeron/relay-layer
The room's checkpoint upload now takes parts (main) and applies the relay
layer's rules (stale refusal, floor coverage, heal) to both the whole and
the staged commit, before reading the body.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-07 17:38:02 -07:00
Wing
53fa172085 zgx beta: 0.2.139, the version after stable 0.2.138
The Windows zgx-desktop dependency moves to the same zgx pin as the rest
(2b6f48e8), and the beta-only icon build dependency goes: zeron/zgx's
build.rs now embeds the icon for every Windows build.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-07 14:09:33 -07:00
Wing
3182a97026 chat2: a checkpoint past one request's most goes in parts
The edge takes at most 32 MiB of checkpoint in one request; a long chat's
snapshot outgrew it (139d8682: 35 MB, da5bd3cc: 39 MB, its state alone
~20 MB), so every post was refused 413 and those rooms kept no checkpoint.

- Edge: `POST /chat2/{id}/checkpoint-part?offset=` appends a piece (at most
  8 MiB) to the room's staged checkpoint, kept as the blob store's chunk
  rows; `POST /checkpoint?seqCovered=&staged=<size>` commits it by renaming
  those rows, with the same floor and head guards (and `incomplete` when
  it isn't all there). GET /checkpoint streams it a row at a time (Range as
  before), and the nightly backup writes a large one as an object of its
  own, streamed, so the room never holds one whole. A single POST, its
  32 MiB cap and its 413 are as they were.
- Engine (src/chat2_checkpoint.rs): up to 32 MiB, one POST as always; past
  it, parts then the commit. An edge from before parts answers 404 and the
  room stays as it was.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-07 00:54:48 -07:00
Wing
f847389dcf Merge zeron/shared-project-look: one project name and icon on every device
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

# Conflicts:
#	crates/doc/src/registry/tests.rs
2026-11-06 21:20:56 -07:00
Wing
c713580b27 Bump version to v0.2.138
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-06 19:54:20 -07:00
Wing
446e4343f8 A project's name and icon belong to its repository
Each device had its own project row for one repository, so a rename or an
icon stayed on the copy it was made on. A repositoryLooks registry row, keyed
by a hash of the shared repository id, now holds them. RegistryDoc reads each
field as the newest clock among that row and the copies' own fields, so copies
that disagree settle with no migration, and what older engines and Cloud
runners write to their own row still reaches current readers. Local ids and
plain folders keep their own fields.

A set writes the look row and the one Space it was made on, so older engines
still see it on that row. No edge change: the room accepts the new kind, and
its runner policy already refuses and hides it.

Not yet compiled or run: this machine was below its free-disk limit for builds.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-06 21:33:41 +00:00
Wing
a9883e5866 Add the relay layer behind flags that default off
Gen-3 chats relay their host's ops instead of storing them. Delivery over the
room or a direct WebRTC link is a hint, and version exchange finds and fetches
whatever was missed. A separate offline queue holds messages for a host that
isn't there until it has saved them. Cloud hosts also post checkpoints.

With CHAT_RELAY_LAYER and ZERON_RELAY_LAYER unset, hello is byte-identical, the
room answers new frames as unknown types, no queue table exists and no engine
state or stream service is created.

Nothing deletes existing rows: the queue is its own table, and relay-era
checkpoints prune nothing. Old clients get a current checkpoint and stored rows
while they're around, and anything queued moves into rows they can read.

Mobile isn't part of this, so a phone seen in the last 14 days keeps its
account's chats on stored rows.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-06 13:32:39 -07:00
Wing
ae03b84a23 Send attachments for Cloud chats through R2 first
Attachment bytes only travelled over a direct peer link to the chat's host.
A Cloud machine may be asleep, so a slow wake or a sender going offline lost
the message after 15 minutes with "attachments never arrived".

A send to a Cloud-hosted chat now waits in a durable outbox while its bytes go
up to /blob/{chatId}/attachments/{uploadId}. Only then does the message enter
the chat. The machine fetches the bytes when it wakes, re-drains as an
UploadCommit does, and deletes the object. The 15-minute cap now applies only
when the store has nothing either.

Chats on a user's own computers, hosts without attachment-store-v1, and edges
without the route keep the peer path. The desktop engine and the mobile core
share the outbox. A permanent failure shows in the chat with a retry. Objects
are deleted on delivery, with a 7-day lifecycle rule as the fallback; see
docs/design/cloud-attachments.md.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-06 12:57:03 -07:00
Wing
f1cb76bff2 Bump version to v0.2.137
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-06 01:02:48 -07:00
Wing
a8d1dc6ecd
Updates that can't strand a device, on every platform, from any device (#1241)
* update: say how a device is installed and whether it can update from afar

UpdateStatus now carries `install`, `canApplyRemotely`, `cannotApplyReason`
and `manualCommand`, so a client no longer has to guess by platform. Older
engines omit them; a method this build doesn't know reads as Unknown rather
than losing the report. `update-ability-v1` advertises it.

ApplyUpdate refuses with the same reason the report gives. That includes a
managed install started by hand, which used to swap `current` and then fail
to restart a service it isn't running under.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* providers: decide what to run at the last moment, and only claim what happened

An update judged its plan before the wait for idle and never looked again, so
a CLI reinstalled meanwhile got the old owner's updater. Under the lease and
owner gate it now re-reads the executable and what's installed and rebuilds the
plan, refusing and checking again if that isn't what the check offered.

With no target version (Hermes, Cursor), exit 0 used to report Updated. It now
takes a moved version, a rewritten executable, or Hermes's own check no longer
finding commits. Otherwise the row stays where it was, with the reason.

Homebrew:
- Only Homebrew's published version is offered. A release Homebrew doesn't have
  yet was an Update button that always failed, and would have failed Update all.
- A formula keeping an npm CLI in its keg is Homebrew's, not npm's. `npm install
  --prefix` was changing the keg behind brew's back. A Node keg's globals stay npm's.
- A Cellar or Caskroom path with no brew at that prefix can't apply, and the
  manual step names brew rather than a self-updater that writes a second copy.

Also: a Chocolatey Claude stays manual, and updaters get 30 minutes rather than
15, since stopping a silent download can leave a truncated CLI.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* update: Linux updates that can't strand a box

`app/current` now moves exactly once, when a new version has proved it's ready.
Until then every start of the service runs the old version, so if the new binary
exits, hangs, or the power goes, the next start lands on a binary that works and
knows how to put things back.

- Refuses before stopping anything if the release and a copy of the databases
  won't fit.
- Records the update durably before answering, with an id; `UpdateStatus.update`
  says how it went (inProgress, updated, rolledBack, failed), since reconnecting
  can't tell an update from a rollback.
- The old version copies the SQLite stores once nothing has them open, leaves a
  watchdog behind, and replaces itself with the new binary. It's the same process,
  so the service manager sees nothing change.
- The new version holds back work it can't take back (resuming chats, queued
  messages), proves it reads its data, serves its port and, when signed in, reaches
  Zeron's servers, then commits and only then resumes chats.
- A failed or timed-out trial goes back: the old version restores the databases
  before anything opens them. An interrupted restore finishes first.

`zeron update` on a box with a running service takes the same path and reports the
outcome. Automatic updates don't retry a release that went back; a person can.
ZERON_SERVICE_UNIT names another unit, so a test install can never restart a real
one.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* providers: an update someone asks for doesn't wait for running chats

It used to wait behind an exclusive lease with no bound, which with agents always
running can look stuck for ever. Where the installer keeps the old version's files,
it now goes ahead: running chats keep the version they started on, new ones get
the new one, and nothing is drained or held up. That's Claude's native install,
Codex standalone, Grok standalone, Cursor (chats use its SDK, not this CLI),
Antigravity and agents from the ACP registry.

Installers that replace or delete files keep waiting, because a chat's process
lives for its whole session and goes on loading files from where it started: npm
and the other package managers, Homebrew, Muse, OpenCode, Pi and Hermes. Their row
says why, in the new `waitReason`. Automatic updates still wait for a quiet moment.

The registry's cleanup used to delete any version older than the last, now that
updates don't wait for chats it keeps a version something is running from.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* update: Mac and Windows devices can update from afar

The same rule as Linux, with the engine doing all of it, so the app doesn't have
to be open: the installed app changes only at the commit, and anything before that
lands back on the old version.

macOS: the new bundle is copied beside the installed one before anything stops, so
what can fail, fails with Zeron still running. The engine drains, backs up and
replaces itself in place, the same process, so it never starts a successor. Once
proved, one atomic swap makes the new bundle the app. An app still open carries on
from the retired bundle until it restarts.

Windows: the old process can't be replaced, so it starts the new exe from a folder
inside the installation and stays on as the watchdog. A running exe can be renamed,
so moving it into place is the commit.

A bundle something runs from is no longer deleted, for updates from the app too:
that's #1037. Retired bundles and old stages go once a process scan finds nothing
running from them.

A second start during a trial (the app's reconnect seeing the port go quiet) now
bows out rather than rolling back a trial that's still running, and zombies no
longer count as running on macOS. The failure tests give each fake install its own
copy of the binary: hard links made the process checks see through to other
installs, which was the cause of their intermittent failures.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* update: nothing but the trial opens the databases mid-update

An app that can't start its helper embeds the engine, which skips the update's
start-up step, and could have opened databases a trial had changed or a restore
was part way through. It now refuses until the update has settled. The app's own
install on quit also refuses during an update: going back from a trial needs the
installed app to stay as it was.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* update: compile on Windows

The watchdog's environment variables are only read on Unix, the test's
version lookup no longer ends in a Unix-only expression, and the Grok CLI
fixture and the test that uses it are Unix-only like their callers.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* update: check free disk space on Windows

Without it every Windows update refused with "couldn't check the free disk
space". The store-list test compares paths with forward slashes.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* update: a brief reader can't fail the update record on Windows

Replacing a file fails with "Access is denied" while anything has it open
without delete sharing (antivirus, an indexer, a script). Seen once on a real
trial: the watchdog couldn't record the trial's pid. Retry the rename for up
to a second.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* update: give the second-start test a deadline a slow machine can't miss

It failed once in 80 runs on a loaded box, when the second start landed after a
3-second trial deadline.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* update: another device's status follows it through its restart

One device asking another's `UpdateStatus` got a subscription that ended the moment
the update restarted the engine, so whoever asked had to know to ask again. It now
carries on until the subscriber goes, asking again if nothing has come for a while,
which also covers a device that died without a word.

Asking a device that can't be reached now says that plainly, and that the update
didn't start, rather than `transport: peer …: readiness check failed`.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* update: a beta Mac stays on beta, and says what's on offer

Tested with a bundle shaped like a beta: a beta interface carrying stable's engine.

An update asked from another device already followed the app's recorded channel,
but with no record, as when the app hasn't run since it was installed, the default
was stable: it installed the stable app and left beta. An app with no recorded
choice now stays on the channel it was built for. A choice someone made at that Mac
still stands.

The engine's report compared releases with its own version, which on a beta is
stable's, so a newer beta read as nothing new and no update was offered. On a macOS
app install it now compares with the app's version, and says what that is in
`appVersion`. It looks again once a commit has changed the answer.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-05 16:59:05 -07:00
José Gurruchaga
86105fa594 test(dictation): verify editor safety, plugin integration and document backends
Cover native editor revisions and IME, Unicode selection and undo, cancellation, explicit Send and Queue, shortcut holds and release outside the microphone. Keep dictation scoped to its compositor, preserve the collapsed editor geometry, support Zeron v2 and avoid polling while idle. Reuse the device model cache and pin the shared ONNX runtime.

Validation: voice tests, engine manifest and dictation permission tests, ZGX dictation integration, split/template tests and locked cargo check pass. Five unrelated ZGX unit failures also reproduce on the base revision; pre-existing debug-leftover checks fail too. Physical microphone and macOS/Windows package checks remain to be run on those platforms.
2026-11-04 20:34:15 -03:00
Wing
cda71d1ed7 Merge main into zeron/zgx
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

# Conflicts:
#	crates/rpc/src/lib.rs
2026-11-03 18:13:25 -07:00
Wing
0a6aef0279
Theme tools for agents: list, read, create, edit, delete and use themes (#1188)
* theme: one catalogue of themes for the window and the engine

zeron_theme::catalog holds what the window's Themes page and an agent's
theme tools both need: the built-in, template and imported themes, the
palette shape ds/palettes wears (read with every problem said, written
back the same), themes made from palettes, update and delete for yours,
the light and dark choice, and which inks the window moves to keep a
palette legible (checked against the window's own mapping in zgx's
tests). zeron_theme::luau_data reads a template's theme.luau as data.

The zgx app's themes host now uses it, and reads the files again when
the engine says its tools changed them (WatchThemes).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* engine, mcp: theme tools for agents

list_themes, get_theme, create_theme, update_theme, delete_theme and
use_theme, over the engine's ListThemes … UseTheme on this device's
theme files. A theme is data, so nothing asks the person; only yours
can be changed or deleted, a built-in or template theme is copied
instead. Palettes are checked with clear errors, a VS Code theme's JSON
comes in through the importer, and each reply says which colours the
window wears lighter or darker to stay legible.

The methods are local-only (a peer never reads or writes another
device's themes) and open to the agent token, but not the window's
WatchThemes. docs/themes.md is served to agents as the other guides are.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* zgx: a theme changed outside the window restyles it; docs for the tools

The Themes fixture can change the library as an agent's tools would,
and a test checks the page lists the new theme and the theme in use
restyles. README and templates.md point to the theme tools.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-03 16:17:36 -07:00
Daniel Katulevskiy
a2113ac886
A command's output keeps a short tail in the chat, with its exit code (#1168)
* A command's output keeps a short tail in the chat, with its exit code

Tool rows have been one-liners since the 2026-08-10 call: the doc fold
dropped every tool's output, and the Claude Code and Codex adapters never
sent any. A command's row could not say how it ended.

Now a command (Exec: Claude Code's Bash, Codex's command execution, an ACP
agent's shell call) keeps the end of its output in its tool part: its last
50 lines or 4 KB, whichever is less, with `outputBytes` (the size it came
at), `outputTruncated` when earlier output was let go, and `exitCode`.
Every other tool's output (file reads, searches, fetches, MCP results) is
still dropped, as before.

- Claude Code: a Bash result's text is its output; a leading `Exit code N`
  line (`Exit code 137 [killed]` too) becomes the exit code, and a result
  without one exited 0, unless the command went to the background. Calls
  are recognised by id from their tool_use (parent, subagent and typed-fork
  transcripts).
- Codex: a command's `aggregatedOutput` and `exitCode` ride its result.
- Both send only the output's last 32 KB, bounding the run journal.
- The tail is redacted before it is cut (tokens, keys, credentials in urls,
  terminal escapes), with the redactor the sign-in flows use, moved from
  zeron-harness to zeron-proto so the doc fold can call it; zeron-harness
  re-exports it.

The output is not streamed while a command runs: the doc is a Loro doc
that keeps its history, so rewriting a 4 KB tail a few times a second would
grow a chat by megabytes over one long build.

* Command output tails: hardened redaction, no output from secret printers

A command's output tail persists in the chat's doc, which syncs to the
person's other devices and the cloud, so its redaction goes further than
the sign-in redactor's (`redact_command_output`, beside `redact_output`):

- private key blocks (PEM and OpenSSH, BEGIN through END, or to the end
  when cut) are dropped for one `[private key redacted]` line;
- an env-style assignment (`AWS_SECRET_ACCESS_KEY=…`, `export X=…`) hides
  its value whatever the key;
- keys naming a password, private key or webhook (`DB_PASS`, `pwd`,
  `private_key`, `slack_webhook`) are secret labels too, while `passed`
  and `compass` stay readable.

Everywhere, a webhook url (Slack, Discord, Telegram bot, Teams, Zapier, …)
now keeps only its host.

A command that prints secrets by nature keeps none of its output, only
"(output not kept: may contain secrets)": env, printenv, export -p, set,
cat/head/tail of .env*, *credentials*, *.pem, *.key, id_*, .netrc,
base64 -d, gpg -d, security find-*-password, pass show, op read,
aws configure get, gcloud auth print-access-token, gh auth token, az
account get-access-token. Pipelines, `&&`/`;` chains, sudo and shell
wrappers (Codex's `zsh -lc '…'`) are looked through.

* Command output tails: 10 lines / 1 KB in the chat

The doc keeps a command's last 10 lines or 1 KB, whichever is less (was
50 lines / 4 KB), with its exit code, size and truncated flag as before.
Redaction before the cut, line-boundary cuts, the single-long-line case
and the secret-command withholding are unchanged. The parts.rs decision
comment and docs/chat2-sync.md (A1) now describe this exception to the
2026-08-10 strip.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: Wing <44917547+wingleeio@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-03 15:00:21 -07:00
Daniel Katulevskiy
c83937d386
zgx: Ports list settles, keeps closed ports a while, and names each process (#1167)
* zgx: Ports list settles, keeps closed ports a while, and names each process

The panel home's Ports listed every listener the engine's 2 s scan found,
so short-lived test servers started by agents appeared and vanished every
few seconds, and every row read "HTTP server".

- Previews.hold (ui/screens/panel/previews.luau): a port (service id +
  port) shows only after it has stayed open 5 s (SETTLE_MS), and stays
  10 s (GRACE_MS) after it closes, faded, with "Closed" in its tooltip;
  back within that time it is listed on. Ports already open when the
  home first gets its answer, their process older than 5 s, show at once.
  The clock is a ticker that runs only while something is settling or
  closing.
- Each row is titled by the process serving it (`cargo test
  zeron_engine`, `node server.js`, `python3 -m http.server`, `vite`), its
  full command line in a tooltip; without one, the engine's name as
  before. The "can't open" line shows only while a port is open.
- Host (src/previews.rs): passes pid, startedAt, and `process`/`command`:
  from the engine's new `command` field, else read from /proc when the
  process runs on this machine (same pid in the same cwd), else empty.
  Values that look like credentials read "…".
- Engine (zeron-proto, zeron-preview): PreviewService gains `command`,
  the listener's argv with credential-like values redacted (optional,
  skipped when empty, so older and newer engines interoperate).

* Previews: advertise a short process name only, never the command line

PreviewService no longer carries the listener's argv (`command`): argv
can hold credentials, and the service is advertised to other devices and
the coordinator. It carries `process` instead, computed on the device:
the program plus a script's, module's or package's name, or a runner's
subcommand and script (`node server.js`, `python3 -m http.server`,
`npm run dev`, `cargo test zeron_engine`), each kept only while it is a
plain name; no other arguments.

The app reads the full command line for its tooltip only from /proc on
this machine (same pid, same cwd), with credential-looking values, URL
userinfo and query strings removed; it never comes from the engine.

* Previews: a flag's value is never read as the script or subcommand in a process name

A flag without '=' takes the next argument (--token abc, -p 3000),
unless it is an interpreter switch known to stand alone (-u, --inspect);
one whose name looks like a credential's always does. Python's -m
module may not start with '-'.

* Preview labels: one copy of process_label and shown_command, in zeron-proto

The engine (zeron-preview) advertises the label and the app labels local
processes and redacts their command lines the same way; one shared
implementation keeps the credential filtering from drifting. Tests move
with them.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* Preview labels: the shared code in its own dependency-free crate

zeron-preview-label holds process_label and shown_command for the engine
(zeron-preview) and the zgx app, so the app doesn't pull zeron-proto's
dependencies in for two string functions.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* Cargo.lock: zeron-preview-label at the workspace's version

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

---------

Co-authored-by: Wing <44917547+wingleeio@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-03 14:40:14 -07:00
Wing
8c8535b7c1
Plugins and templates: consent bound to capabilities, engine-held decisions, hardened runtime (#1139)
* Plugins and templates: consent bound to capabilities, engine-held decisions, hardened runtime

- Consent binds a capability digest (programs, MCP servers, tools, network hosts, services, module replacement); it re-asks only when that grows. UI targets never ask.
- "You" is this device's key; code-running powers are never auto-granted; keys are revoked on sign-out and rotation, listed and revocable in Settings, and new keys are announced.
- Decisions and signing need a one-time review the consent window holds; agents get a scoped IPC token; agent-created chats can't exceed their creator's access; peers can't allow; relayed calls carry the edge-verified device.
- Runtime: text-only Luau, one command policy (env allow-list, hardened git/gh), plugin network blocks private ranges unless declared, safe path joins, verified snapshots, hardened built-in Git service.
- App: ui.change can't override modules, Settings sealed, decision controls protected, plugins get ids and resolve data through grants, consent sheet hardened, dev switches debug-only.
- Templates: unapproved templates run as previews with Keep/Revert, a host-owned way back on every platform, safe mode, no builtin-theme spoofing, increasing versions.
- Edge: step-up for key registration and account deletion, key list/revoke, reserved names, invitation/image/admin limits, relay device stamp.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* Signing keys register on the first sign-in, revocations act at once, setup and layout fixes

- A sign-in registers this device's key with its fresh ticket, also on a local-mode engine; new keys are noticed at start; a revoked own key rotates at once (peers are told), drops code-running grants and shows a notice.
- Calls during unfinished account setup answer setup_incomplete; plugins and templates load once an organization is chosen.
- The host strip's wrapper fills the window.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* Engine: only plugin and template calls answer setup_incomplete during onboarding; everything else waits as before

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-03 11:33:36 -07:00
Wing
5325585f75 Bump version to v0.2.136
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-03 03:24:21 -07:00
Wing
e4292b568f Bump version to v0.2.135
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-03 03:23:49 -07:00
Wing
d898f936ed
zgx: Zeron's interface on zgx (app, design system, plugins, beta channel) (#1137)
* zgx: Zeron's interface on zgx (app, design system, plugins, onboarding, Zeron Preview)

A new desktop app in apps/zgx: Zeron's interface written in Luau on zgx,
its design system, built-in plugins (Git, GitHub, Parallax scenes) and
theme templates, plus the engine and edge work it needs (plugin services
and agent tools, the agent registry, project icons, profile, organisation
admin, onboarding, plugin signing keys) and a separate Zeron Preview
package and workflow (preview-v* tags, off until enabled).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* Fix Linux and Windows builds: Listener's signing_in, Command::as_std on Windows

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* Engine: plugin watchers ignore reads (inotify loop), per-component snapshot paths, Windows create-over-folder is DestinationExists

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* zgx as a beta channel of Zeron.app; shared CLI, launch and notify crates; update channels

- Update channels (stable/beta) with semver prereleases, cross-channel installs, ZeronEngineVersion and a macOS 13 floor for beta; "Get beta updates" in the current app's Check for Updates.
- The command line moves into crates/cli and engine start-up into crates/launch, shared by the current app, zeron-engine and zgx; zgx hands engine commands to the bundle's Zeron Engine helper.
- zgx beta packaging and .github/workflows/zgx-beta.yml (beta-v* tags, gated by ZGX_BETA_ENABLED); the separate Preview packaging is removed.
- Templates install from links with updates, an Unverified badge for unknown authors, the Archived settings page, the Cloud repository picker, palette sub-steps without scope tabs.
- Git plugin refreshes on file changes; GitHub plugin batches, tiers and throttles its polling; failed PR watches retry with backoff.
- zgx f81d5e99.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* zgx: registry agents' sign-in state, methods and icons; PR badges isolated per session

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* Harness: registry sign-in uses the crate's Child (Windows build)

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-03 03:21:42 -07:00
Wing
9ef868762d Bump version to v0.2.134
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-03 03:59:54 +00:00
Wing
c2c117663e Bump version to v0.2.133
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-03 03:14:18 +00:00
Wing
a7947b7213 Bump version to v0.2.132
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-11-02 02:41:36 -07:00
Wing
0a153df5ad Bump version to v0.2.131
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-10-31 11:57:37 -07:00
Chance
6973500bc0
Sync: update loro to 1.17 so streaming chats stop leaking memory (#1116)
* Sync: update loro to 1.17 so streaming chats stop leaking memory

loro before 1.17 left another copy of a doc's text in its string arena on
every snapshot export (fixed in loro-dev/loro 089a4e6). chat_persistence
exports a dirty chat each second, so a chat with a streaming agent grew
quadratically until the arena passed 4 GiB and export panicked (#905).

* Sync: upgrade edge Loro and guard snapshot memory growth

Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

---------

Co-authored-by: gratenes <claude@storage.gay>
Co-authored-by: Wing <44917547+wingleeio@users.noreply.github.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-10-31 11:55:18 -07:00
Wing
369d428151 Bump version to v0.2.130
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-10-29 12:37:26 -07:00
Wing
b1d4927f84
Direct links: let WebRTC packets fragment, so links over Tailscale carry data (#1106)
Sign-ins from another device failed with "can't connect" once the browser
came back: the callback rides the direct P2P link, and a link that ICE put
on Tailscale paired but never carried data.

webrtc's tokio runtime wraps its UDP sockets with quinn-udp, which forbids
fragmentation (IP_DONTFRAG on macOS, IP_PMTUDISC_PROBE on Linux,
IP_DONTFRAGMENT on Windows). A full SCTP packet is ~1293 bytes on the wire;
Tailscale carries 1280. Every full packet failed with EMSGSIZE ("Message too
long (os error 40)") and was retransmitted forever, stalling the ordered
channel, so the sign-in's stream open timed out and the browser was dropped.
Terminals hid it by falling back to the relay; sign-ins have no fallback.

The peer connection now uses a runtime that delegates to webrtc's tokio
runtime and clears those options after each UDP socket is wrapped, so the
kernel fragments the oversized packets as a plain UDP socket would.

Verified:
- unit test: the stock runtime forbids fragmentation, ours doesn't
  (macOS + Linux)
- the 1 MiB P2P streaming test in a Linux netns with loopback MTU 1280:
  main stalls every run ("P2P stream stalled"); this branch passes in 0.5s
- full zeron-preview suite passes on Linux

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-10-29 12:36:30 -07:00
wing-anara
8b7afb306d Bump version to v0.2.129
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-28 09:25:50 +00:00
Wing
ab189159f1 Bump version to v0.2.128
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-10-27 16:37:07 -07:00
wing-anara
31d26eeb6b Bump version to v0.2.127 2026-10-27 17:55:23 +00:00
Alexandre V.L
3d2f1e64af
Smooth scrolling: mouse-wheel notches glide like Chrome's, with a setting (#1005)
* Smooth scrolling: mouse-wheel notches glide like Chrome's, with a setting

A mouse-wheel notch moved content in one jump: GPUI hands non-precise
wheels to containers as `ScrollDelta::Lines`, applied at once. The window
root now takes each notch at the end of the capture phase and replays it
over the following frames as fractional `Lines` slices, on Chrome's wheel
curve (`cc::ScrollOffsetAnimationCurve` for `kMouseWheel`): ease-in-out,
cubic-bezier(0.42, 0, 0.58, 1), over 200ms, shortening to 100ms for long
moves. A notch in flight retargets at the current speed, so a spin is one
continuous glide. Slices stay in lines, so every container keeps its own
line height and a notch travels exactly as far as before.

Precise input, notches with Ctrl, Alt, Cmd or Fn held, and content that
eases notches itself (WebView2) stay raw; a press ends a glide.

The in-app browser on macOS glides too. AppKit hands a page its wheel
directly and WKWebView applies a tick at once, so a local event monitor
holds the page's ticks and replays the same glide as precise pixel deltas,
which WebKit routes like a trackpad's: to the scroller under the pointer,
clamped at its edges.

Settings → Appearance → Motion → Smooth scrolling, on by default. Reduced
motion turns it off.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* Transcript: upward input breaks the bottom pin as it arrives

The pin broke only when one wheel event moved the view more than 1px from
the bottom, decided in a deferred check. A glided notch starts and ends
with sub-pixel slices, and its slices arrive on frames, where the spring's
step runs before that check: it pulled the view back and recorded the
distance the check compared against. A notch up from the bottom of a chat
drifted back down. Upward input over a scrollable transcript now breaks
the pin on the spot, as use-stick-to-bottom's wheel handler does.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* docs: smooth scrolling media

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* Transcript: only a wheel the list took breaks the bottom pin

`on_scroll_wheel` broke the pin for any upward wheel that bubbled to the
transcript root, including one an overlay above the list let through
(an image viewer at its pan limit, say) while the list never moved. The
list's scroll handler now marks the event it took, and only that event
can escape the pin.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* Smooth scrolling: the AppKit glide flag is macOS-only

Only `browser::macos` reads it; elsewhere `gliding` was dead code.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* Smooth scrolling: leave touchpads alone on X11 and Windows

X11 reports touchpads (and high-resolution wheels) as fractional
`ScrollDelta::Lines`, so every touchpad event was held back and eased
over 200ms. Outside macOS, where trackpads are flagged precise, only a
whole-line delta is a wheel notch; fractional lines now pass through.

The transcript's arrival-time pin escape is for wheel notches only:
precise input keeps the distance test it had before.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* Smooth scrolling: drop the browser's unused wheel fixture hook

No fixture calls fixture_wheel; it served a throwaway capture that was never
committed.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* Smooth scrolling: findable in settings search

The Motion row joins the settings index under Appearance → Motion, with the
words people reach for: mouse, wheel, glide, ease.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* Settings search: the Appshots pick test runs where Appshots shows

Appshots is hidden off macOS and Linux, so search can't find its rows there
and the test failed on Windows since #1018.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Wing <44917547+wingleeio@users.noreply.github.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-10-27 06:12:52 +00:00
Alexandre V.L
c31124aa91
Linux browser: WPE WebKit at the display's rate, frames shared with GPUI as dma-bufs (#1039)
* Linux browser: WPE WebKit at the display's rate, frames shared with GPUI as dma-bufs

The sidebar browser on Linux ran WebKitGTK in an offscreen window: every
frame went through the CPU, over a pipe, into a new GPUI image, so pages
topped out near 60 fps whatever the display, and large ones fell well below.

Zeron now carries its own WPE WebKit platform (browser/linux/wpe.c) and runs
it in the helper when the build and the system have WPE WebKit 2.50+:

- WebKit renders on the GPU GPUI draws with, into dma-bufs in a DRM format
  modifier GPUI's Vulkan device imports. GPUI copies only each frame's
  damaged regions into the page's atlas tile, on the GPU (zeronsh/zui#18).
- Where the driver renders no layout Vulkan imports, the helper reads frames
  back with OpenGL ES into a sealed memfd. Without a GPU, WebKit renders on
  the CPU into shared memory. Both update the tile in place.
- GPUI's frame callbacks are the pages' vblank: WebKit hears a frame is on
  screen at most once per display refresh, so pages run at 60, 144 or 240 Hz
  and never render frames the display won't show. Hidden pages, and pages
  replaced by their error, get none.
- WebKit runs copy, cut and paste itself, sets the pointer's cursor, and sees
  click counts and the pointer leaving the page.

If the WPE helper can't start, the session uses WebKitGTK. Its helper now
renders at the output's scale (it rendered 4x the pixels at 1.5x on Wayland
and downscaled them) and sends Cairo's surface as-is instead of converting it
through GdkPixbuf, which took most of its CPU.

A helper that crashes, sends a message Zeron can't read or is still being
written by a forked child (ETXTBSY) is stopped, replaced or retried; the
tests run the real helper on every render node, in every layout its driver
renders, read back and in shared memory, at 60, 144 and 240 Hz.

* Linux browser bench: shown frames, pacing, copies, press latency, memory

`ZERON_BROWSER_BENCH` measured one page's frame rate and CPU. It now loads
a fresh document for each of a 2D canvas, WebGL 2, scrolling, a CSS spinner
and, with `ZERON_BROWSER_BENCH_VIDEO`, a 60 fps video, and records for each:
the page's frame rate, when GPUI painted every new frame, the bytes copied
into the page's tile from CPU memory and between GPU buffers, CPU, PSS and
NVIDIA GPU memory for Zeron and for WebKit's processes, and a capture. It
then times 40 presses from GPUI's input to the frame showing them, and
measures an idle page. `ZERON_BROWSER_BENCH_WIDTH` sizes the browser.

The counts are kept only in fixture builds (`browser-fixture`), for both
engines, so WebKitGTK and every WPE path are measured the same way.

`ZERON_BROWSER_DMABUF=readback` keeps WebKit on the GPU but reads its frames
back, the path GPUs that render no importable layout take.

* Linux browser bench: WebGL page checks a WebGL 1 context too

The WebGL page's benchInfo() clears a WebGL 1 context and reads a pixel back, so a bench run records that both WebGL versions render.

* Linux browser: rustfmt the WPE engine and its tests

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* Linux browser: a paste without text clears WPE's clipboard copy

The WPE helper kept the last clipboard text Zeron shared, and Zeron shared
only text. Pasting with an image or nothing on the system clipboard then
inserted that earlier text, possibly into another page. Zeron now shares
the clipboard on every paste, and empty text clears the helper's copy.

build.rs embeds an empty WPE helper, as when WPE is missing, if wpe.c does
not compile against the installed development files (an older GBM), and
detects them again when PKG_CONFIG_PATH changes. probe_gpu releases the GBM
device and render node it opened when the GPU is unusable.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* Linux browser fixture: rustfmt

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

* Linux browser docs: the WPE vblank check needs a compositor that answers idle frame callbacks

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

---------

Co-authored-by: Wing <44917547+wingleeio@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-10-27 06:08:25 +00:00
Alexandre V.L
5d2b91901b
Agents can show local images in chat: show_image, and Markdown images in replies (#1036)
* Agents can show local images in chat: show_image, and Markdown images in replies

show_image {path} (Zeron MCP) copies a PNG/JPEG/WebP/GIF into uploads and
publishes the image part Codex's generated images use, so every client
already renders it and it outlives the file.

An agent reply's ![alt](path) now draws the image in place, read live by the
chat's host through ReadChatImageChunk: strictly inside that chat's folder,
no link followed, no URL fetched, extension and bytes checked, decoded within
the shared raster limits. Anything refused keeps today's link.

* Chat image folder tests: refuse the filesystem root portably

On Windows `/` isn't absolute, so it was refused as "not an absolute
path" instead of as the root. Test the root of the temp dir's own path
(`/`, or a drive like `C:\`) instead.

* Retrigger CI: cursor_shim frame timeout on the Windows runner

* Reply images: right-click copies the original, and narrow columns scale them whole

Rebased onto main (Origin's new side_chats_only in the show_image test).

- A Markdown image in a reply now opens the shared Copy path / Copy image
  menu, like generated images and tool previews. Copy image reads the
  original through ReadChatImageChunk under that chat's jail, not the wider
  ReadAttachmentChunk, and not the downscaled preview.
- The image row keeps its aspect ratio when the column is narrower than the
  image, instead of keeping its height and letterboxing.
- docs/image-preview.md: the menu, the sizing, and what the cache means for a
  Markdown image whose file changes.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>

---------

Co-authored-by: Wing <44917547+wingleeio@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: zeronsh[bot] <346184242+zeronsh[bot]@users.noreply.github.com>
2026-10-27 02:55:52 +00:00