| Age | Commit message (Collapse) | Author |
|---|
|
this commit includes some optimisations related to the llvertextbuffer and is closer to LL code in that regard. It also removes the alternative vertexbuffer mode.
|
|
#5815 Rename Lua runtime dropdown entries
|
|
Signed-off-by: Luan Nutels - VolkSec <luan.nutels@owasp.org>
|
|
mResutCodes gets modified from different threads, mutex it.
|
|
|
|
No user facing editor UI, cycle, or randomise yet.
Sittings cannot be separately disabled yet.
Sitting On Ground doesn't show Stand button (and cannot stand right away
from it either unless sitting, but not on ground, on some object, and
then stand).
Pre-jumping and landing can get you stuck (unless you disable
pre-jumping and landing, whose code might be the cause of this
stuckness).
Standing can get glitchy if idle which might be because of the random
cycle of the default stands.
The #AO folder can be anywhere, doesn't have to be under some folder
with a specific name.
AOs are loaded lazily, and per action, not all at once.
Reset animation changes by turning AO off and on again.
All these limitations will be fixed later.
|
|
on x86-64 as it breaks web media. 151.3.18 is the latest version known
to work.
|
|
|
|
Combo box doesn't translate bools right and isn't meant to. Switch to U32.
|
|
|
|
|
|
|
|
|
|
Rider/runtime msging
|
|
|
|
|
|
|
|
|
|
|
|
We use a non member static associative array here cause it needs to
be accessed in sendChatFromViewer that is static. A public function
for modifying it is also just static.
As per spec, "It does not apply to emotes, and will not trigger any
animation", hence some extra efforts there.
|
|
like in RLvCommand constructor in rlvhelper.cpp.
|
|
There is a race condition where voice can start up before the voice server
type is retrieved from the region. Since the voice server type defaults to null,
and that wasn't allowed, a popup message happened.
The default voice type is webrtc, so we connect to that in cases where the voice
server type is empty.
|
|
Fix shutdown crash and cut WebRTC teardown time
|
|
Key already is an LLSD::String and needs no further parsing.
This parsing was added for command line handling and should be done
explicitly in command line's parser. Also extended event listeners,
there should be no ambiguity in what kind of data we are passing.
|
|
Seems to be needed now on aarch64, but doesn't hurt to have it too
on x86-64.
|
|
|
|
message passing into unified format.
|
|
|
|
|
|
this commit is a fix for the scaling issue happening randomly on macs with a retina display.
|
|
|
|
#5574 Add Lua and LSL-on-Luau options to Build - Scripts - Recompile Scripts.
|
|
and that two threads won't race t delete same request
|
|
|
|
|
|
|
|
|
|
geenz/linux-to-develop to lua_editor
|
|
Rider/explore selection
|
|
|
|
LLGridManager is performance critical and low level enough that it should
just be turned into LLSimpleton. Without capture_dependency there won't
be a 'deadlock'.
|
|
out of bounds
|
|
# Conflicts:
# .github/workflows/build.yaml
# indra/llinventory/llinventory.cpp
# indra/newview/CMakeLists.txt
# indra/newview/llappviewer.cpp
# indra/newview/llappviewerwin32.cpp
# indra/newview/llviewermenu.cpp
# indra/newview/llvoiceclient.cpp
# indra/newview/skins/default/xui/en/floater_preferences_graphics_advanced.xml
|
|
|
|
|
|
|
|
objects to be saved back to their rezzer's inventory.
|
|
terminate() released the thread unique_ptrs on the timeout path while the
detached shutdown thread was still running the lambda that reads them, so a
slow shutdown became a null deref at the next BlockingCall. Leave the members
alone and report the timeout instead, so the caller leaks this object rather
than deleting it out from under the detached thread.
Reduce how often that timeout is reached at all:
- Post the audio device stop instead of blocking on it. StopRecording and
StopPlayout each wait up to 2s on a WASAPI thread, spending much of the 10s
budget before the connections were touched. The device stop has to stay on
the worker thread (AudioDeviceBuffer is guarded by a sequence checker bound
to it), but the queue is FIFO, so ForceTerminate still runs after it while
the signaling close gets on with its network-thread work.
- Drain the peer connections in LLWebRTCVoiceClient::terminate() before handing
off to llwebrtc::terminate(), bounded to 3s. An estate session can hold ten
live connections -- the current region plus up to eight neighbours -- and
closing them inline and serially under the shutdown budget is what made it
tight. Draining also quiets the stats poll, so PeerConnection::Close() no
longer blocks in WaitForPendingRequest() on an in-flight GetStats.
Connections reaped by the drain are destroyed while webrtc is still fully
alive, so they must unregister and release their peer connections normally;
key the destructor's early-out off a new sWebRTCTerminated instead of
isShuttingDown(). Notify signaling observers even when there was no peer
connection to close, or a connection shut down before it finished initializing
parks in VOICE_STATE_WAIT_FOR_CLOSE, which has no timeout of its own.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
|
|
|
|
