| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
| |
script wait event (llSleep(), etc.)
This makes use of EventWaitHandles since various web references indicate that Thread.Interrupt() can also cause runtime instability.
If co-op termination is enabled, then termination sets the wait handle instead of waiting for a timeout before possibly aborting the thread.
This allows the script to cleanly terminate if it's in a llSleep/LL function delay or the next time it enters such a wait without any timeout period.
Co-op termination is not yet testable since checking for termination request within loops that never trigger a wait is not yet implemented.
This commit, unlike 1b5c41c, passes the wait handle as an extra parameter through IScript.Initialize() instead of passing IScriptInstance itself.
|
|
|
|
|
|
|
|
| |
during a script wait event (llSleep(), etc.)"
Doing this as a favour to Melanie. This will be back with passing the wait handles directly to the api.
This reverts commit 1b5c41c14ad11325be249ea1cce3c65d4d6a89be.
|
|\ |
|
| |
| |
| |
| | |
other than the object currently sat on
|
|/
|
|
|
|
|
|
|
| |
script wait event (llSleep(), etc.)
This makes use of EventWaitHandles since various web references indicate that Thread.Interrupt() can also cause runtime instability.
If co-op termination is enabled, then termination sets the wait handle instead of waiting for a timeout before possibly aborting the thread.
This allows the script to cleanly terminate if it's in a llSleep/LL function delay or the next time it enters such a wait without any timeout period.
Co-op termination is not yet testable since checking for termination request within loops that never trigger a wait is not yet implemented.
|
|
|
|
| |
rather than just saying it happened.
|
|
|
|
|
|
|
| |
This involves making Attachments module listen for start/stop script changes.
It also involves removing the script from the region on detach in the same manner as every other DeleteSceneObject() call rather than simply stopping it
This is necessary tue to the bad assymetry of start and stop script triggers but it appears to be the correct behaviour anyway, as detached objects are completely gone from the sim.
Not just in a state where their scripts have been stopped.
|
|
|
|
| |
scene <key> <value>" to "debug scene set <key> <value>" to accomodate future settings
|
|
|
|
|
|
| |
periodic backup
This is rather than making unnecessary duplicate checks that the SOG later performs again.
|
|
|
|
|
|
| |
disable periodic scene backup. For debug purposes.
If false, scene is still saved on shutdown.
|
|
|
|
| |
This reverts commit 5561333668f61f043cdfc0733a4eb50a1bcfb14e.
|
|
|
|
| |
This reverts commit a775931a0cec9f65748c6e20dd2695edcbe21b7f.
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
As per http://wiki.secondlife.com/wiki/LlGetLinkKey
This is done by keeping a scene-object wide list of sitters.
This also fixes bugs in this function where linknums 0 and 1 weren't treated properly if there were sitting avatars on a single prim.
This also fixes a minor race condition for multiple concurrent sitters on a prim with no current sitters by locking on the object-wide list rather than individual sop lists
Addresses http://opensimulator.org/mantis/view.php?id=6477
|
|
|
|
| |
inconsistently used. - The sun position is always calculated by combining the sun settings in the Region and Estate. This fixes the problem that 'UseEstateSun' didn't work. - To remove ambiguity, the EstateToolsSunUpdate event no longer accepts the sun's position as parameters. That's because the position is always calculated from the Region and Estate settings. - Use only the 'FixedSun' flag to determine whether the sun is fixed; not the 'UseGlobalTime' flag. - Don't change the region's 'SunPosition' field according to the sun's position: this field is used only to set the position when using a FixedSun. (The 'SunVector' field does get updated according to the sun's position in the sky)
|
|
|
|
| |
will be a no-op for ODE but enables asymmetrical avatars for BulletSim.
|
|
|
|
| |
velocities of zero.
|
|
|
|
| |
Also made it so llSetStatus() can individually enable disable rotation axi using the bitmask of flags.
|
|
|
|
| |
test. This seems to cause border crossing to be sensed either a little early or a little late depending on which directin the object is moving. If border crossings become totally borked or someone remembers why this was displacement was done, revert this change.
|
|
|
|
| |
more helpfully log the name, id, position and region of that object.
|
|
|
|
| |
use (such as logging anim pack contents sent to clients)
|
|
|
|
|
|
|
| |
causing subsequent issues if physics was re-enabled.
Added regression tests
Addresses http://opensimulator.org/mantis/view.php?id=6365
|
|
|
|
| |
instead of HandleSitRequest. This eliminates m_nextSitAnimation, an unneeded state-saving variable in ScenePresence
|
|
|
|
| |
in, to avoid race conditions on the client.
|
|
|
|
| |
Inventory transfers: don't do async on asset transfers or now.
|
|
|
|
|
|
|
|
|
|
| |
aim of avoiding OpenSimulator problems due to script thread aborts.
When an object is removed, its scripts are stopped and then the thread running them is aborted if stop takes too long.
However, it appears that aborting a thread at just the wrong moment when it is obtaining a ReaderWriterLockSlim lock can leave this lock in an inconsistent state.
One symptom of this is that mono leaps to 100% cpu and a vm thread dump reveals lots of threads waiting for a ReaderWriterLockSlim lock without any thread actually holding it.
This is probably the same problem as encountered originally in commit 12cebb12
This commit looks to plaster this problem by putting lock obtaining methods inside finally blocks which should be uninterruptible by thread aborts.
|
| |
|
|
|
|
|
|
| |
functionally identical.
Move TestDeleteSceneObjectAsync() and TestDeleteSceneObjectAsyncToUserInventory() from SceneObjectBasicTests -> SceneObjectDeRezTests
|
|
|
|
| |
BasicInventoryAccessModule() and re-enable in test suite.
|
|
|
|
| |
automatically turns off any logging enabled between tests
|
|
|
|
| |
physics error
|
|
|
|
|
|
| |
various console commands (e.g. "show users") from working.
This was because the "currently selected" scene reference was being left as the dead scene instead of the restarted Scene object.
|
| |
|
|
|
|
|
| |
If we don't do this, then these scenes don't get deregistered from the grid, amongst other things.
Regression was introduced in commit 8c130bc (Mon Nov 12 22:50:28 2012 +0000)
|
|
|
|
| |
code is there
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
| |
and make it compile again
Not currently in test suite since not yet working.
Also add method doc to Scene.DeRezObjects()
|
| |
|
|
|
|
| |
No change to existing functions.
|
|
|
|
|
| |
Uses new IEntityInventory.TryGetScriptInstanceRunning()
Makes it clearer that TaskInventoryItem.ScriptRunning cannot be used as it is temporary and not updated.
|
|\ |
|
| |
| |
| |
| | |
actually are any targets.
|
|/ |
|
|
|
|
| |
specifies a UUID rather than a string
|
| |
|
|
|
|
| |
from the root prim rather than the source prim
|
|
|
|
| |
SceneObjectPart.SendSound as linksets are only meant to have a single owner
|