iniFile/binFile::GetCachedKey used RegCreateKeyExA for every access, so a plain
GetStr/GetInt on a missing section materialised an empty subkey. CScreenManager's
constructor calls LoadQualityProfiles() unconditionally, which reads profile0..
profile5 x4 keys, so every client start left six zero-valued subkeys behind:
126 of them on the dev box, all with zero values. The server's HKCU\SoftwareYAMA\quality key is the same artefact.
Split the helper by intent: writes go through RegCreateKeyExA, reads through
RegOpenKeyExA. Both still ask for KEY_READ|KEY_WRITE, because the handle cache is
shared and a read-only cached handle would make a later SetStr on the same
section fail.
Also fixes a second bug in SaveQualityProfiles: when a level matched the default
it simply continued, so a level the operator reset to default kept the old
custom values in the registry and the reset never took effect. Add
SectionExists() (open-only, never creates) and write the defaults back when the
section already exists.
A host whose DXGI initialisation hangs never opens the remote desktop window at
all. WorkThreadProc runs InitScreenSpy() and SendBitMapInfo() back to back, so a
capturer that never finishes constructing means TOKEN_BITMAPINFO is never sent,
and the master - which only opens the window on that token - waits forever. On
the affected Server 2022 box the screen thread stopped right after logging its
type and algorithm, printing neither the DXGI init summary nor the existing
"DXGI SPY init failed" fallback, so it was stuck inside the capturer
constructor; a healthy machine on the same build reaches the summary in under a
second. Six clicks left six stuck threads, each holding a D3D device.
DXGI desktop duplication needs a display to duplicate. With no interactive
desktop it can return no frames at best, and on some drivers it never returns
from the initialisation calls at all. BitBlt has no such dependency and keeps
working, which is why switching the screen type to GDI made the same host
connect. Detect that up front and pick GDI instead of relying on a hang to be
survivable:
- IsWorkstationLocked() - the input desktop is Winlogon. Moved out of
CKernelManager into Manager.cpp so the capturer can use it too; the manager
keeps its private wrapper for the "Locked:"/"Inactive:" activity string.
- IsCurrentSessionDisconnected() - the process's own session is
WTSDisconnected. That is the state the reported host was actually in: its
client runs in session 2, which was disconnected, and the desktop only came
back after the master's "RDP session reset" ran tscon. Session 0 has no
disconnected state and is left alone.
Either condition means the same thing for capture, so both route to the plain
GDI capturer. An active session keeps DXGI untouched, so hosts that work today
are unaffected; the new log line names which condition fired.
Co-Authored-By: deepseek-flash
Add a third view to the web UI, the screen wall, which shows several
devices at once in one tab with 1x1 / 2x2 / 3x3 / 4x4 / 5x5 layouts and a
device sidebar on the left. Tiles are monitoring-only; clicking one
enlarges that device into the existing single-screen page for control, and
Back returns to the wall. The same device can be added several times: all
tiles of one device share a single decoder and a single stream.
The wall decodes at most 9 streams on desktop and 4 on touch devices.
Tiles beyond the budget keep receiving frames but have their decoder
closed and show the last frame plus a Paused badge; they are promoted
automatically when a live tile goes away. Subscriptions are reference
counted on the client (streams map with owner sets), so a stream is only
connected on the 0 -> 1 transition and disconnected on 1 -> 0, and the
wall and the single-screen page can share one device without tearing the
stream down.
Server side, WebClient now holds a set of watched devices instead of a
single id. A new multiWatch command flag selects set semantics, while
clients without it keep the old replace-all behaviour, so an old frontend
against the new server is unchanged. connect and disconnect responses echo
id and req so the frontend can route them, input commands carry an id and
are resolved against the watch set (dropped when the set has several
devices and no id), and the per-client watch set is capped at 32.
For fast recovery after a pause or when joining late, the server caches
each device's raw IDR payload and replays it on connect and on a new
keyframe command, then asks the client to force a new IDR. The client gets
COMMAND_SCREEN_KEYFRAME_REQ = 169 and an atomic request flag that is only
cleared after an IDR actually came out of the encoder, so a failed capture
or EAGAIN does not lose the request. The frontend retries the keyframe
request up to three times; an old client that ignores the command simply
recovers on the next natural GOP.
Also fix the keyframe cache double wrapping: BroadcastH264Frame used to
cache the full packet header and HandleConnect wrapped it again. The cache
now stores data + 9, len - 9 and every sender wraps it through
BuildFramePacket, and the cache is cleared on cold start and on resolution
change. StopRemoteDesktop is always called outside m_ClientsMutex, with
the ids collected under the lock first.
Co-Authored-By: deepseek-flash
Add a novice-friendly "File Transfer Speed" dialog under the Advanced menu
that controls file-transfer speed in both directions without touching
SimplePlugins. Speed is passed through InitFileUpload's existing
chunkSizeKb/sendDurationMs arguments, so the precompiled library stays
untouched. Seven presets run from ~0.3 MB/s up to a new 10 MB/s tier plus an
unlimited tier; the chosen value is stored in [transfer] ChunkKb/DurationMs.
Upload reads THIS_CFG at the master's InitFileUpload (2015RemoteDlg.cpp:2225)
and takes effect on restart. Download is pushed to the client over a new
CMD_FILE_SPEED (252) command, sent from OnOnlineFileManager and
OnOnlineDesktopManager before each session opens, and read by all four client
InitFileUpload call sites via GetTransferSpeed so the process-global refcount
cannot latch a stale speed. All Chinese strings go through LangManager, with
en_US/zh_TW/de_DE entries added for the tier names, format strings and hint
text.
Co-Authored-By: deepseek-v4-pro
ProcessCommand chose the MSG record size by testing "ulLength % 28 == 0"
before "% 48 == 0". The two sizes' least common multiple is 336 (7*48 =
12*28), so a 48-byte batch whose record count is a multiple of 7 was
misclassified as 28-byte MSG32 records, shifting every field and garbling
(or silently dropping) injected input.
Check "% 48 == 0" first. The modern controller always emits 48-byte MSG64;
the 28-byte MSG32 path is legacy 32-bit-controller compatibility and is only
reached when 48 does not divide evenly. This is the first feature (MCP
remote_keyboard "type", and the upcoming remote_mouse drag) to emit
multi-record batches, which is what made the latent bug reachable.
Co-Authored-By: deepseek-v4-pro
Replacing the old connection on a duplicate login made the client
reconnect in a tight loop. Instead, park the new connection in a standby
list keyed by clientID and promote it into the main host list only after
the old connection is removed by heartbeat timeout or disconnect.
Fix heartbeat attribution so a parked standby refreshes its own
last-heartbeat time rather than the main connection's, otherwise the main
never times out and the standby can never take over. Gate the offline
notification on an actual removal with no promoted standby to avoid a
spurious "host offline" when takeover happens. Skip promoting a standby
already marked removed, and reuse the normal login sign/settings path when
a standby is promoted.
Co-Authored-By: deepseek-v4-pro
Two changes to reduce unnecessary CScreenSpy restarts when connecting:
1. Client (ScreenManager.cpp): Initialize CScreenSpy with bitrate from
the locally-saved quality profile, so CMD_QUALITY_LEVEL arriving from
the server (same bitrate as default) hits SetBitRate(3000)==3000 and
skips the restart instead of comparing against the hard-coded 0.
Also fixes QualityLevel init to use the already-computed `quality`
variable (which honours the QUALITY_DISABLED override when algo!=NUL)
rather than re-reading the cfg key a second time.
2. Server (ScreenSpyDlg.cpp): Only send CMD_SCREEN_SIZE strategy=2 when
the session is in QUALITY_ADAPTIVE mode and a cached maxWidth exists.
Fixed quality levels already carry resolution via CMD_QUALITY_PROFILES,
so unconditionally sending CMD_SCREEN_SIZE caused a second restart when
the screen spy was still rebuilding from the first one.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Server: clamp web session adaptive quality to H264-only levels (>=Good) in EvaluateQuality and ApplyQualityLevel; Ultra/High (DIFF/RGB565) caused the browser to freeze ~1 min into a session
- Server: move session-type detection to the top of ScreenSpyDlg::OnInitDialog and skip SetWindowPlacement/EnterFullScreen for hidden web sessions, eliminating the MFC dialog flash on web-triggered opens
- Linux client: default QualityLevel from QUALITY_ADAPTIVE to QUALITY_GOOD to match Windows/macOS so the server's adaptive controller doesn't auto-upgrade to non-H264 algorithms
- Web: clear the floating quick-action toolbar on fullscreen exit so its row of buttons (RDP reset / Mouse / Close) doesn't stay pinned to the top of the page
- Web: route F11 to the remote in control mode instead of toggling local fullscreen
- Web: route Esc to the remote in control mode via the Keyboard Lock API instead of exiting native fullscreen