4e3fba3cfd
Apparently some recent server side change causes kill object messages for attachments sometimes being sent after the TP has technically finished already. Currently the experimental fix doesn't cover that case because the TP state would already be TELEPORT_NONE. To handle this new failure case, resort to introducing a timer (yuck!) that will prevent attachments from being detached temporarily after a TP has finished. This basically has the same effect of enabling TP progress screens, as it delays the finalization of TP for 2 seconds, which often seem to be enough time prevent attachments from being detached. For now, set the timer to prevent attachments being detached after TP to 3 seconds to test how it works out. This not only affects attachments being falsely detached by the region, but also prevents manual detaching during this period immediately after a TP. The delay of this new timer can be configured via FSExperimentalLostAttachmentsFixKillDelay debug setting. |
||
|---|---|---|
| .. | ||
| cmake | ||
| copy_win_scripts | ||
| doxygen | ||
| integration_tests | ||
| lib/python/indra | ||
| linux_crash_logger | ||
| llappearance | ||
| llaudio | ||
| llcharacter | ||
| llcommon | ||
| llcorehttp | ||
| llcrashlogger | ||
| llimage | ||
| llimagej2coj | ||
| llinventory | ||
| llkdu | ||
| llmath | ||
| llmessage | ||
| llplugin | ||
| llprimitive | ||
| llrender | ||
| llui | ||
| llvfs | ||
| llwindow | ||
| llxml | ||
| mac_crash_logger | ||
| media_plugins | ||
| newview | ||
| test | ||
| tools | ||
| viewer_components | ||
| win_crash_logger | ||
| CMakeLists.txt | ||
| Version | ||
| edit-me-to-trigger-new-build.txt | ||
| fix-incredibuild.py | ||