This reverts commit 44a491f36b.
This makes packaging OpenSim slightly easier but then generates a FriendsConnector error if naively used after compilation.
So let's revert to stick with the "DataStore" error for now - at least that's a known error.
I must have had it cached last night.
This now makes completely fresh avatar setups appear properly (remember to wipe viewer cache and server cache as well as database).
As a bonus, eyes now appear by default too!
This is controlled by the CreateDefaultAvatarEntries switch in the [UserAccountService] section of config-include/Stnadalone.ini
This is left as false in grid mode
Viewer 2 no longer contains the default avatar assets (i.e. "Ruth") that would appear if the user had insufficient body part/clothing entries.
Instead, avatars always appear as a cloud, which is a very bad experience for out-of-the-box OpenSim.
Default is currently off. My intention is to switch it on for standalone shortly.
This is not particularly flexible as "Ruth" is hardcoded, but this can change in the future, in co-ordination with the existing RemoteAdmin capabilities.
Need to fix creation of suitable entries for users created as estate owners on standalone.
Avatars still appear with spooky empty eyes, need to see if we can address this.
This commit adds a "Default Iris" to the library (thanks to Eirynne Sieyes from http://opensimulator.org/mantis/view.php?id=1461) which can be used.
This matches the GridCommon setting and is the best tested setting. It appears to work fine on standalone.
Also, add information that the flotsam asset cache is the recommended cache, since it is most used and actively maintained.
Had to stop using AvatarService for now since it doesn't store baked texture IDs (which is why this was failing).
Also failing because cloning appearance was also cloning the AvatarApperance.Owner field, which we weren't then changing.
Extended TestCreate() to check this.
When this setting is false and other defaults haven't been changed, scripts entering the simulator from another sim are automatically restarted but their state is not preserved.
When this setting is true, state that accompanies the script is reloaded (e.g. if a script had a variable i = 1 when leaving the source region, it will still have i = 1 on the destination region).
This setting does not affect crossings when regions are run in the same simulator. In this case, state is already preserved.
Some items had completely wrong permissions - this is easier than correcting them all.
The ability to set permissions in xml is retained since there are use cases for this (e.g. to create no-mod library scripts)
Library items always need the same permissions, so it doesn't make sense to load them from the xml files. This just opens the door to permissions mistakes.
This matches the ability to disable the memory part
This is controlled through the FileCacheEnabled parameter in FlotsamCache.ini
Default is true, so existing installations are not affected.
Improved fcache command feedback when various caches are disabled.
Re-enabled test for flotsam cache with file caching disabled.
Instead of performing the 4096 check when the region is linked (and subsequently removing the link), leave the link in place and perform the check in the entity transfer module
This allows us to explicitly tell the hypergridder why the teleport failed (region out of range).
It also allows people on regions that are within range (on a large source grid) to teleport.
The Check4096 config parameter in the [GridService] section is replaced by a max_distance paramter in a new [EntityTransfer] section in OpenSimDefaults.ini
Since the parameter is in OpenSimDefaults.ini no action needs to be taken unless you want to increase this limit. It could also be decreased.
The check is being made in the base entity transfer module, since I believe the viewer problem occurs both on extremely large grids and while hypergridding.
Changed the experimental capability introduced a couple of commits ago: now sending that extra information as part of the response in the SimulatorFeatures cap.
For the most part, defaults are made to match those already in FlotsamCache.ini.example.
The one exception is that the 48 hour file timeout from the code is used instead of the 0 hours that was in the example file. This can be tweaked if necessary.
Most importantly, the default cache directory is now ./assetcache (as in FlotsamCache.ini.example) rather than ./FlotsamAssetCache (as was the internal code default).
Therefore, if you were using flotasm without using the config file, then please rename your cache directory or start using the ini file and change the default there if you want to keep using your existing cache.
this is a prerequisite to fixing llDialog issues for the latest Linden viewers, since they are now making use of a new OwnerData field in the ScriptDialog message
It appears that if the viewer requests a folder containing links, we must also send the folders that contain the link targets first.
This was tested with Kokua 0.1.0 WIP though I predict it will also work with other viewer 2s
This is to facilitate running Freeswitch in a separate ROBUST shell, as on grid setups there can be quite a lot of traffic flowing between FreeSWITCH and ROBUST. However, running in a spearate shell is not mandatory.
limits because the only ones used now are the defaults (which are overwritten
by the client throttles anyway). Updated the default rates to correspond to
about 350kbps.
Also added a configuration to disable adaptive throttle. The default
is the previous behavior (no adaptation).
This brings it into line with the Windows and Linux libraries.
This is a universal dylib with x86_64, i386 and ppc parts. However, even on a 64 bit Intel machine Mono can only P/INVOKE the i386 version right now. ppc is untested.
The configuration used to compile was
CFLAGS="-g -O2 -isysroot /Developer/SDKs/MacOSX10.6.sdk -arch i386 -arch x86_64 -arch ppc" CXXFLAGS="-g -O2 -isysroot /Developer/SDKs/MacOSX10.6.sdk -arch i386 -arch x86_64 -arch ppc" LDFLAGS="-arch i386 -arch x86_64 -arch ppc" ./configure --enable-old-trimesh --disable-asserts --enable-shared --disable-dependency-tracking --disable-demos --without-x
--disable-demos --without-x is required to build ODE on Mac OS X
CFLAGS, CXXFLAGS and --disable-dependency-tracking are necessary to build the universal dylib (some compilation lines use CFLAGS instead of CXXFLAGS)
The other settings are tweaks for using ODE with OpenSim
This means that Mac OS X users can now use the standard sqlite adaptor instead of the legacy one.
This is SQLite 3.7.5. I configured the build with the line
CFLAGS="-Os -g -isysroot /Developer/SDKs/MacOSX10.6.sdk -arch i386 -arch x86_64 -arch ppc7400 -DSQLITE_ENABLE_COLUMN_METADATA" LDFLAGS="-arch i386 -arch x86_64 -arch ppc7400" ./configure --disable-dependency-tracking
The SQLITE_ENABLE_COLUMN_METADATA flag exports the metadata symbols that modern Mono SQLite adaptors need.
The -Os removes debugging symbols (as per the sqlite3 source pkg README)
The other parts are to cross compile for x86_64, i386 and ppc. On Mac OS X, Mono can actually only P/INVOKE 32-bit libraries even on 64 bit platforms so i386 is being used. The dylib has not been tested on ppc.
Apply the same change to both the 32-bit and 64-bit DLL target paths. This makes the previous addition to the instructions in the README for running on Linux unnecessary.
This version fixes a bug so that MySqlCommand.CommandTimeout = 0 now correctly sets an infinite command timeout. This might help with failures on long-running migrations.
The mysql.data.dll v2 is used from the mysql package (as opposed to v4). The v4 version (probably standing for NET 4.0) does not work under mono 2.6.7, which is still commonly used.
child agents in neighbor regions. This means you can extend
the view on a simulator beyond the default 3x3 regions.
This uses a region default draw distance and should be
replaced at some point by the avatar specified draw distance.
That will require more careful, dynamic recomputation of child
agents every time the draw distance changes.
WARNING: this is experimental and has known instabilities. specifically
all regions "within site" should be running the same default draw distance
or agents will not be closed correctly.
This matches the behaviour for other deletions and stops this operation failing on Windows .NET (though mono is fine with the duplication).
Probably a holdover from copying code from the older legacy adaptor.
Resolves http://opensimulator.org/mantis/view.php?id=5361
Made FireAndForgetWrapper a singleton class to allow us to drop
dependancy on the BclExtras35 library. BclExtras is broken in
Mono 2.8.2 and we used the library in only one function.
Frontend into a separate region module. Configuration
settings added to the defaults ini file. Others may find
this useful for pushing map tiles to a separate server.
Change the remote Grid region module so that it takes the network connector as a config variable. This allows the region plugin to be reused for both Robust and Simian network connectors.
Configuration option AssetService in section [GridService] must be set to enable this functionality.
Map tiles do currently not show for hyperlinks set in simulators that are connected to grids, see Mantis #5270.
the same thing was actually done for ode.dll a couple of commits ago but I got the commit msg wrong (should have been r1755, not r1754). Damn these fat fingers.
This revision dates from 02 Sep 2010 and has received some testing. This code will be updated in the opensim-libs repository
Updates for libode-x86_64, ode.dll and libode.dylib will follow
WARNING!!! This changes a default. The old default is to allow multiple
presences, the new default disables this. As the feature currently has no users,
this should not present any difficulty and will alleviate the presence
issues somewhat.
This will cause visual params to be persisted along with worn items. With
this, alpha and tattoo laters will be saved. Multiple layers MAY work, but
not tested because I don't use Viewer 2.
Setting this to true avoids a 500ms or so client freeze when the LLUDP server thread is taken up with processing a UseCircuitCode packet synchronously.
Extensive testing on Wright Plaza appeared to show no bad effects and this seems to reduce login lag considerably.
Of course, a lot of login lag is still coming from other sources.
Main reason for upgrade is to revert API change that was introduced in order to support headerless binary LLSD decoding needed for deserialization of the mesh assets, which turned out to be unnecessary as we can transparently decode both kinds.
Justin Clark-Casey (justincc)