| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |\ |
|
| | |
| |
| |
| |
| |
| |
| | |
(A,B,C) where the A root agent is still closed, terminating the connection.
This was occuring because teleport to B did not set DoNotCloseAfterTeleport on A as it was a neighbour (where it isn't set to avoid the issue where the source region doesn't send Close() to regions that are still neighbours (hence not resetting DoNotCloseAfterTeleport).
Fix here is to still set DoNotCloseAfterTeleport if scene presence is still registered as in transit from A
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
an line from A->B->C would not close region A when reaching C
The root cause was that v2 was only closing neighbour agents if the root connection also needed a close.
However, fixing this requires the neighbour regions also detect when they should not close due to re-teleports re-establishing the child connection.
This involves restructuring the code to introduce a scene presence state machine that can serialize the different add and remove client calls that are now possible with the late close of the
This commit appears to fix these issues and improve teleport, but still has holes on at least quick reteleporting (and possibly occasionally on ordinary teleports).
Also, has not been completely tested yet in scenarios where regions are running on different simulators
|
| |\ \
| |/ |
|
| | |
| |
| |
| |
| |
| | |
add minor details to some log messages, rename a misleading local variable name.
No functional changes.
|
| |/ |
|
| |
|
|
| |
estate bans also, and delete the obsolete config var.
|
| | |
|
| |
|
|
|
|
|
|
|
| |
as needed. (Although sometimes Justin's state machine kicks in and doesn't let you) The EventQueues are a hairy mess, and it's very easy to mess things up. But it looks like this commit makes them work right. Here's what's going on:
- Child and root agents are only closed after 15 sec, maybe
- If the user comes back, they aren't closed, and everything is reused
- On the receiving side, clients and scene presences are reused if they already exist
- Caps are always recreated (this is where I spent most of my time!). It turns out that, because the agents carry the seeds around, the seed gets the same URL, except for the root agent coming back to a far away region, which gets a new seed (because we don't know what was its seed in the departing region, and we can't send it back to the client when the agent returns there).
|
| |
|
|
| |
DoNotClose to DoNotCloseAfterTeleport
|
| |
|
|
|
|
|
|
| |
was a previous root agent, do not close that child agent at the end of the 15 sec teleport timer.
This prevents an issue if the user teleports back to the neighbour simulator of a source before 15 seconds have elapsed.
This more closely emulates observed linden behaviour, though the timeout there is 50 secs and applies to all the pre-teleport agents.
Currently sticks a DoNotClose flag on ScenePresence though this may be temporary as possibly it could be incorporated into the ETM state machine
|
| | |
|
| |
|
|
|
|
|
|
| |
the destination is being checked)
In this new protocol, and as committed before, the viewer is not sent EnableSimulator/EstablishChildCommunication for the destination. Instead, it is sent TeleportFinish directly. TeleportFinish, in turn, makes the viewer send a UserCircuitCode packet followed by CompleteMovementIntoRegion packet. These 2 packets tend to occur one after the other almost immediately to the point that when CMIR arrives the client is not even connected yet and that packet is ignored (there might have been some race conditions here before); then the viewer sends CMIR again within 5-8 secs. But the delay between them may be higher in busier regions, which may lead to race conditions.
This commit improves the process so there are are no race conditions at the destination. CompleteMovement (triggered by the viewer) waits until Update has been sent from the origin. Update, in turn, waits until there is a *root* scene presence -- so making sure CompleteMovement has run MakeRoot. In other words, there are two threadlets at the destination, one from the viewer and one from the origin region, waiting for each other to do the right thing. That makes it safe to close the agent at the origin upon return of the Update call without having to wait for callback, because we are absolutely sure that the viewer knows it is in th new region.
Note also that in the V1 protocol, the destination was getting UseCircuitCode from the viewer twice -- once on EstablishAgentCommunication and then again on TeleportFinish. The second UCC was being ignored, but it shows how we were not following the expected steps...
|
| |
|
|
|
|
| |
15sec before closing the agent. This seems to be working fairly well. The viewer seems to have an 8 sec delay between UseCircuitCode and CompleteMovement.
Also added back the position on UpdateAgent, because it's needed for TPing between neighboring regions.
|
| | |
|
| |
|
|
| |
(Note: different from the defunct see_into_neighboring_sim, which used to control the process from the other end). This enables child agents in neighbors for which the root agent doesn't have permission to be in.
|
| |
|
|
| |
chaotic while ppl are using different versions of opensim. Warning only, but no enforcement.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
test.
|
| |
|
|
| |
requesting this for all landmarks in inventory. Not sure why. But this seems to be the root cause of the login freeze mentioned before. This commit adds a blocking queue / process thread pattern.
|
| | |
|
| |
|
|
| |
may yield unexpected results in some cases. No database persistence yet,
|
| |
|
|
| |
This allows object modification before the usual heartbeat operation.
|
| |
|
|
|
|
|
|
|
|
| |
establishing a connection, to see if this helps with "Unknown UserUMMTGUN" issues.
The UMMTGUN form of Unknown User seems to appear because a viewer sometimes sends a UUIDNameRequest UDP request that fails to find a binding.
However, in theory the incoming agent should have made that binding before any such request is triggered.
So moving this binding to an earlier point in the process to see if this makes a difference.
Unknown user name is also updated to UserUMMTGUN2 - if you see the old name then you need to clear your viewer cache.
This relates to http://opensimulator.org/mantis/view.php?id=6625
|
| |
|
|
|
|
|
| |
neighbour rather than sending all closes concurrently on a separate thread.
This is to reduce race conditions where neighbours may be responding erratically, thus mixing up create and close agent requests in time.
This mirrors OpenSimulator behaviour on enabling child agents where each region is contacted separately.
|
| |
|
|
|
|
| |
user is not prevented from relogging if many neighbours are present but not responsive.
The symptom here is that previous user connections are still present but are labelled active == false
|
| |
|
|
| |
about who has permission to set home where.
|
| |\ |
|
| | | |
|
| |/
|
|
| |
Warning - contains a small migration.
|
| |
|
|
| |
(complete module) as changes are too extensive to apply manually
|
| |\ |
|
| | |
| |
| |
| |
| |
| |
| | |
starting scenes before others have been created.
This aims to avoid a race condition where scenes could look to inform neighbours that they were up before those neighbours had been created.
http://opensimulator.org/mantis/view.php?id=6618
|
| | |
| |
| |
| | |
parcel in the sim, keep them out entirely.
|
| |/
|
|
| |
to be added earlier for some of the checks to work correctly.
|
| | |
|
| |
|
|
|
|
|
|
|
| |
megaregions.
This fixes the combined bounding box location for regions bigger than 256x256.
It also fixes the position on taking coalesced objects in the non-root regions, where position checks are properly done on rez instead.
It also fixes the megaregion land channel to return null if the land does not exist, which should probably also be done for the ordinary land channels rather than returning a dummy region.
Inspiration from Garmin's commit in http://opensimulator.org/mantis/view.php?id=6595. Thanks!
|
| |
|
|
|
|
|
|
| |
scripts in attachments.
This was because the script resumption in AttachmentsModule was firing the attach event instead.
Had to reinstate the code in 285bd3a do we can resume the scripts there instead, though the bug existed before its removal.
This is to resolve http://opensimulator.org/mantis/view.php?id=6578
|
| | |
|
| |
|
|
|
|
|
|
| |
with the other attachments code, using TeleportFlags.ViaLogin check to fire if necessary.
This is to simplify the code (no tricky 'wasChild' signalling required)
and to reduce the risk of a thread clash between simulator-side attaching (necessary for v1 viewers)
and the viewer-side attaching the v3 viewers perform.
|
| |
|
|
| |
with our own and add export permissions as well as a new definition for "All" as meaning "all conventional permissions" rather than "all possible permissions"
|
| |
|
|
|
| |
This is relevant to mega-regions where the borders are very different to a regular region.
Also adds some method doc and other code comments.
|
| |
|
|
|
|
| |
running regression tests (fire and forget calls launched on the same thread).
Also adds code comments as to why this pause exists.
|
| |
|
|
|
|
|
|
|
| |
to fix viewer 3 issues if its own rezzing actions on login collide with the simulator side actions.
This resolves issues (at least in my tests with LL 3.3.4) where this can make attachments invisible until one zooms in on the avatar.
This doesn't affect version 1 viewers since this delay is shorter than the login delay.
This doesn't increase the login time since this part of the process was already being performed asynchronously.
This may be a temporary solution.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Conflicts:
OpenSim/Framework/AvatarAppearance.cs
OpenSim/Region/CoreModules/Avatar/Attachments/AttachmentsModule.cs
OpenSim/Region/CoreModules/Avatar/Attachments/Tests/AttachmentsModuleTests.cs
OpenSim/Region/Framework/Interfaces/IAttachmentsModule.cs
OpenSim/Region/Framework/Scenes/Scene.cs
OpenSim/Region/OptionalModules/Avatar/Attachments/TempAttachmentsModule.cs
OpenSim/Region/ScriptEngine/Shared/Api/Implementation/LSL_Api.cs
|
| |
|
|
|
|
|
| |
in OpenSim.ini from [Startup]
This change also corrects the setting names - they were actually wrong (though the text in {} was correct).
If there are settings in [Startup] they will continue to be used and anything there will override settings in [AccessControl]
|
