aboutsummaryrefslogtreecommitdiffstatshomepage
Commit message (Collapse)AuthorAgeFilesLines
* OOOPPPSSSUbitUmarov2019-03-031-2/+2
|
* direct encode lludp terse object update, let contain texture entryUbitUmarov2019-03-023-41/+78
|
* ll rez objects: silent ginore null object id; remove redundante llSetColor codeUbitUmarov2019-03-022-42/+3
|
* put back assetViewer cap code, to allow viewers testing. It is disabled at ↵UbitUmarov2019-03-012-8/+7
| | | | OpenSimDefaults, and should not be enabled until all major viewers do it right on opensim
* reduce the chance of using a invalid avatar physics actorUbitUmarov2019-03-011-15/+28
|
* Add Thread.ResetAbort() to various thread loops to clean up errors onRobert Adams2019-02-275-32/+63
| | | | | | | | shutdown. Fixes Mantis #8494. Threads are aborted when shutting down and ThreadAbortException is odd in that it is rethrown at the end of the catch unless the abort is reset. No functional changes but fewer error messages on shutdown.
* combat module: those parcel changes are now handled elsewhereUbitUmarov2019-02-272-35/+2
|
* direct encode terseupdatesUbitUmarov2019-02-274-37/+207
|
* avoid packet split on terseupdatesUbitUmarov2019-02-261-30/+76
|
* issues with udp buffers pool on heavy loadUbitUmarov2019-02-264-42/+53
|
* mantis 6569: restore full health on invulnerable set to true; combat module ↵UbitUmarov2019-02-261-1/+6
| | | | is outdated and needs work. By then the proposal on this mantis should be reviewed
* buffers in pool have no data, better mk that clearUbitUmarov2019-02-261-0/+1
|
* fix UDPBuffersPoolCount value (readable with stats show clientstackUbitUmarov2019-02-261-2/+1
|
* zero decode using a buffer from the poolUbitUmarov2019-02-251-6/+14
|
* removed a redundant function; try to make a particular vi coder happy about ↵UbitUmarov2019-02-252-23/+7
| | | | removed comments
* change UDPPacketBuffer pools (does waste a bit of memory)UbitUmarov2019-02-256-323/+156
|
* make the options visible on OpenSimDefaultsUbitUmarov2019-02-241-0/+19
|
* also throttle llhttprequest by owner, options: PrimOwnerRequestsBurst = 5, ↵UbitUmarov2019-02-243-70/+86
| | | | PrimOwnerRequestsPerSec = 25; increase concurrency to 8
* several changes to llHttpRequest processing: options section renamed ↵UbitUmarov2019-02-244-81/+117
| | | | ScriptsHttpRequestModule; throttle by prim (PrimRequestsBurst = 3, PrimRequestsPerSec = 1) and max concurrent connections per instance (MaxPoolThreads = 5), llhttprequest returns if error, nullkey if throttled, reqid otherwise
* update script sintaxeUbitUmarov2019-02-222-33/+59
|
* remove redundant osInventoryDescription() and improve osInventoryDesc() to ↵UbitUmarov2019-02-223-34/+4
| | | | do the same
* add LSL_Key osGetLastChangedEventKey() ( it is a alias for lldetectedkey(0) ↵UbitUmarov2019-02-213-0/+15
| | | | but don't tell anyone
* let Xengine also have that detectedkeyUbitUmarov2019-02-212-6/+19
|
* add LSL_String osGetInventoryName(LSL_Key itemId) and LSL_String ↵UbitUmarov2019-02-213-3/+46
| | | | osGetInventoryDescription(LSL_String itemNameOrId)
* let changed allowed drop have lldetectedkey[0] return the dropped item id, ↵UbitUmarov2019-02-215-9/+22
| | | | Yengine and still testing"
* Yengine changed position and shape events can bt throttled by mineventdelayUbitUmarov2019-02-211-0/+15
|
* add LSL_Key osGetInventoryLastOwner(string itemNameorid)UbitUmarov2019-02-213-2/+26
|
* fix last owner on add to object inventoryUbitUmarov2019-02-212-5/+5
|
* Yengine:save minEventDelay in script stateUbitUmarov2019-02-202-0/+18
|
* Yengine: let llminEventDelay do something: it just ignores a more limited ↵UbitUmarov2019-02-206-17/+62
| | | | set of events than Xengine (neither do as SL)
* Xengine: exclude a few more events from mineventdelayUbitUmarov2019-02-201-1/+3
|
* missing file on the kick commitUbitUmarov2019-02-201-1/+1
|
* Xengine: don't lose state change eventsUbitUmarov2019-02-201-1/+1
|
* estate kick now just kill user connection if on same region. The teleport ↵UbitUmarov2019-02-203-22/+39
| | | | home we did, does not make much sense, and would need more work anyways
* add osKickAvatar(LSL_Key agentKey, string alert)UbitUmarov2019-02-203-22/+53
|
* oops do show the object deleteUbitUmarov2019-02-151-1/+1
|
* (almost) useless changeUbitUmarov2019-02-151-23/+42
|
* minor cleanupUbitUmarov2019-02-152-29/+0
|
* don't break permitions on god object return or deleteUbitUmarov2019-02-151-4/+11
|
* mantis 8479: deep change DeRezObjects(..) doing independent permitions ↵UbitUmarov2019-02-152-103/+71
| | | | checks per action. m_useTrashOnDelete should now work except if god deletes, but still not recomended
* Remove some more cut-and-paste typos in 6dof constraint setup in XNARobert Adams2019-02-051-3/+3
| | | | | | version of BulletSim. Could only effect operation of flexible linksets when using the C# version of the Bullet physics engine. Found by Tampa and AliciaRaven and reported in Mantis 7612.
* Remove cut-and-paste typo in 6dof contstrain setup in XNA version ofRobert Adams2019-02-051-1/+1
| | | | | | BulletSim. Could only effect operation of flexible linksets when using the C# version of the Bullet physics engine. Found by Tampa and reported in Mantis 7612.
* allow more suicidal NonPhysicalPrimMaxUbitUmarov2019-02-051-2/+2
|
* enforce prim size restrictions on physical state change; Place restrictions ↵UbitUmarov2019-02-052-2/+32
| | | | on ini values
* prevent freswitch from messing global cert validation, more work is need on ↵UbitUmarov2019-02-041-10/+1
| | | | rest requests ( also for other things)
* remove references to obsolete CertificatePolicyUbitUmarov2019-02-042-34/+2
|
* reduce the odds of MTU violationUbitUmarov2019-02-041-1/+1
|
* optimizers are sillyUbitUmarov2019-02-034-0/+0
|
* typos, thxUbitUmarov2019-02-012-2/+2
|
* place a global cooldown of 2 min on god map generationUbitUmarov2019-02-011-4/+23
|